gmic - Man Page

Perform image processing operations using the G'MIC framework.

Help

 gmic: GREYC's Magic for Image Computing: command-line interface
       Version 2.9.9
       (https://gmic.eu)

       Copyright (c) Since 2008, David Tschumperlé / GREYC / CNRS.
       (https://www.greyc.fr)

 1. Usage
    -----

   gmic [command1 [arg1_1,arg1_2,..]] .. [commandN [argN_1,argN_2,..]]

 'gmic' is the open-source interpreter of the G'MIC language, a  
  script-based programming language dedicated to the design of possibly complex  
  image processing pipelines and operators.  
 It can be used to convert, manipulate, filter and visualize image datasets made  
  of one or several 1D/2D or 3D multi-spectral images.  

 This reference documentation describes all the technical aspects of the G'MIC  
  framework, in its current version 2.9.9.  

 As a starting point, you may want to visit our detailed tutorial pages, at:  
  https://gmic.eu/tutorial/

 2. Overall Context
    ---------------

  * At any time, G'MIC manages one list of numbered (and optionally named)  
  pixel-based images, entirely stored in computer memory (uncompressed).  
  * The first image of the list has index '0' and is denoted by  
  '[0]'. The second image of the list is denoted by '[1]', the third  
  by '[2]' and so on.  
  * Negative indices are treated in a periodic way: '[-1]' refers to the  
  last image of the list, '[-2]' to the penultimate one, etc. Thus, if the  
  list has 4 images, '[1]' and '[-3]' both designate the second image  
  of the list.  
  * A named image may be also indicated by '[name]', if 'name' uses  
  the character set '[a-zA-Z0-9_]' and does not start with a number. Image  
  names can be set or reassigned at any moment during the processing pipeline  
  (see command name for this purpose).  
  * G'MIC defines a set of various commands and substitution mechanisms to allow  
  the design of complex pipelines and operators managing this list of images, in  
  a very flexible way: You can insert or remove images in the list, rearrange  
  image order, process images (individually or grouped), merge image data  
  together, display and output image files, etc.  
  * Such a pipeline can define a new custom G'MIC command (stored in a user  
  command file), and re-used afterwards as a regular command, in a larger  
  pipeline if necessary.  

 3. Image Definition and Terminology
    --------------------------------

  * In G'MIC, each image is modeled as a 1D, 2D, 3D or 4D array of scalar  
  values, uniformly discretized on a rectangular/parallelepipedic domain.  
  * The four dimensions of this array are respectively denoted by:  
    -  'width', the number of image columns (size along the  
  'x-axis').  
    -  'height', the number of image rows (size along the 'y-axis').  
    -  'depth', the number of image slices (size along the  
  'z-axis'). The depth is equal to '1' for usual color or grayscale  
  2D images.  
    -  'spectrum', the number of image channels (size along the  
  'c-axis'). The spectrum is respectively equal to '3' and '4'  
  for usual 'RGB' and 'RGBA' color images.  

  * There are no hard limitations on the size of the image along each dimension.  
  For instance, the number of image slices or channels can be of arbitrary size  
  within the limits of the available memory.  
  * The 'width', 'height' and 'depth' of an image are  
  considered as spatial dimensions, while the 'spectrum' has a  
  multi-spectral meaning. Thus, a 4D image in G'MIC should be most often regarded  
  as a 3D dataset of multi-spectral voxels. Most of the G'MIC commands will stick  
  with this idea (e.g. command blur blurs images only along the spatial  
  'xyz'-axes).  
  * G'MIC stores all the image data as buffers of 'float' values (32 bits,  
  value range '[-3.4E38,+3.4E38]'. It performs all its image processing  
  operations with floating point numbers. Each image pixel takes then  
  32bits/channel (except if double-precision buffers have been enabled during the  
  compilation of the software, in which case 64bits/channel can be the default).  
  * Considering 'float'-valued pixels ensure to keep the numerical  
  precision when executing image processing pipelines. For image input/output  
  operations, you may want to prescribe the image datatype to be different than  
  'float' (like 'bool', 'char', 'int', etc.). This is  
  possible by specifying it as a file option when using I/O commands. (see  
  section Input/Output Properties to learn more about file options).  

 4. Items of a Processing Pipeline
    ------------------------------

  * In G'MIC, an image processing pipeline is described as a sequence of  
  items separated by the space character. Such items are interpreted and executed  
  from the left to the right. For instance, the expression:  

   filename.jpg blur 3,0 sharpen 10 resize 200%,200% output file_out.jpg  

 defines a valid pipeline composed of nine G'MIC items.  

  * Each G'MIC item is a string that is either a command, a list of  
  command arguments, a filename or a special input string.  
  * Escape characters '' and double quotes '"' can be used to define items  
  containing spaces or other special characters. For instance, the two strings  
  'single item' and '"single item"' both define the same single item,  
   with a space in it.  

 5. Input Data Items
    ----------------

  * If a specified G'MIC item appears to be an existing filename, the  
  corresponding image data are loaded and inserted at the end of the image list  
  (which is equivalent to the use of 'input filename').  
  * Special filenames '-' and '-.ext' stand for the standard  
  input/output streams, optionally forced to be in a specific 'ext' file  
  format (e.g. '-.jpg' or '-.png').  
  * The following special input strings may be used as G'MIC items to create and  
  insert new images with prescribed values, at the end of the image list:  
    -  '[selection]' or '[selection]xN': Insert 1 or N copies of  
  already existing images. 'selection' may represent one or several images  
  (see section Command Items and Selections to learn more about selections).  
    -  'width[%],_height[%],_depth[%],_spectrum[%],_values[xN]': Insert  
  one or N images with specified size and values (adding '%' to a dimension  
  means "percentage of the size along the same axis", taken from the last  
  image '[-1]'). Any specified dimension can be also written as  
  '[image]', and is then set to the size (along the same axis) of the  
  existing specified image '[image]'. 'values' can be either a  
  sequence of numbers separated by commas ',', or a mathematical expression,  
   as e.g. in input item '256,256,1,3,[x,y,128]' which creates a  
  '256x256' RGB color image with a spatial shading on the red and green  
  channels. (see section Mathematical Expressions to learn more about  
  mathematical expressions).  
    -  '(v1,v2,..)[xN]': Insert one or 'N' new images from specified  
  prescribed values. Value separator inside parentheses can be ',' (column  
  separator), ';' (row separator), '/' (slice separator) or '^'  
  (channel separator). For instance, expression '(1,2,3;4,5,6;7,8,9)'  
  creates a 3x3 matrix (scalar image), with values running from 1 to 9.  
    -  '('string'[:delimiter])[xN]': Insert one or N new images from  
  specified string, by filling the images with the character codes composing the  
  string. When specified, 'delimiter' tells about the main orientation of  
  the image. Delimiter can be 'x' (eq. to ',' which is the default),  
  'y' (eq. to ';'), 'z' (eq. to '/') or 'c' (eq. to  
  '^'). When specified delimiter is ',', ';', '/' or  
  '^', the expression is actually equivalent to  
  '({'string'[:delimiter]})[xN]' (see section Substitution Rules for  
  more information on the syntax).  
    -  '0[xN]': Insert one or N new 'empty' images, containing no  
  pixel data. Empty images are used only in rare occasions.  

  * Input item 'name=value' declares a new variable 'name', or  
  assign a new string value to an existing variable. Variable names must use the  
  character set '[a-zA-Z0-9_]' and cannot start with a number.  
  * A variable definition is always local to the current command except when it  
  starts by the underscore character '_'. In that case, it becomes also  
  accessible by any command invoked outside the current command scope (global  
  variable).  
  * If a variable name starts with two underscores '__', the global  
  variable is also shared among different threads and can be read/set by commands  
  running in parallel (see command parallel for this purpose). Otherwise,  
  it remains local to the thread that defined it.  
  * Numerical variables can be updated with the use of these special operators:  
  '+=' (addition), '-=' (subtraction), '*=' (multiplication),  
  '/=' (division), '%=' (modulo), '&=' (bitwise and),  
  '|=' (bitwise or), '^=' (power), '<<=' and '>>'  
  (bitwise left and right shifts). For instance, 'foo=1' 'foo+=3'.  
  * Input item 'name.=string' appends specified 'string' at the end  
  of variable 'name'.  
  * Input item 'name..=string' prepends specified 'string' at the  
  beginning of variable 'name'.  
  * Multiple variable assignments and updates are allowed, with expressions:  
  'name1,name2,...,nameN=value' or 'name1,name2,...,nameN=value1,value2,  
  ...,valueN' where assignment operator '=' can be replaced by one of  
  the allowed operators (e.g. '+=').  
  * Variables usually store numbers or strings. Use command store to  
  assign variables from image data (and syntax 'input $variable' to bring  
  them back on the image list afterwards).  

 6. Command Items and Selections
    ----------------------------

  * A G'MIC item that is not a filename nor a special input string  
  designates a 'command' most of the time. Generally, commands perform  
  image processing operations on one or several available images of the list.  
  * Reccurent commands have two equivalent names ('regular' and  
  'short'). For instance, command names 'resize' and 'r' refer  
  to the same image resizing action.  
  * A G'MIC command may have mandatory or optional arguments. Command  
  arguments must be specified in the next item on the command line. Commas ',  
  ' are used to separate multiple arguments of a single command, when required.  
  * The execution of a G'MIC command may be restricted only to a subset of  
  the image list, by appending '[selection]' to the command name. Examples  
  of valid syntaxes for 'selection' are:  
    -  'command[-2]': Apply command only on the penultimate image  
  '[-2]' of the list.  
    -  'command[0,1,3]': Apply command only on images '[0]',  
  '[1]' and '[3]'.  
    -  'command[3-6]': Apply command only on images '[3]' to  
  '[6]' (i.e, '[3]', '[4]', '[5]' and '[6]').  
    -  'command[50%-100%]': Apply command only on the second half of the  
  image list.  
    -  'command[0,-4--1]': Apply command only on the first image and the  
  last four images.  
    -  'command[0-9:3]': Apply command only on images '[0]' to  
  '[9]', with a step of 3 (i.e. on images '[0]', '[3]',  
  '[6]' and '[9]').  
    -  'command[0-9:25%]': Apply command only on images '[0]' to  
  '[9]', with a step of 25% (i.e. on images '[0]', '[3]',  
  '[6]' and '[9]').  
    -  'command[0--1:2]': Apply command only on images of the list with  
  even indices.  
    -  'command[0,2-4,50%--1]': Apply command on images '[0]',  
  '[2]', '[3]', '[4]' and on the second half of the image list.  
    -  'command[^0,1]': Apply command on all images except the first two.  
    -  'command[name1,name2]': Apply command on named images 'name1'  
  and 'name2'.  

  * Indices in selections are always sorted in increasing order, and duplicate  
  indices are discarded. For instance, selections '[3-1,1-3]' and '[1,1,  
  1,3,2]' are both equivalent to '[1-3]'. If you want to repeat a single  
  command multiple times on an image, use a 'repeat..done' loop instead.  
  Inverting the order of images for a command is achieved by explicitly inverting  
  the order of the images in the list, with command 'reverse[selection]'.  
  * Command selections '[-1]', '[-2]' and '[-3]' are so often  
  used they have their own shortcuts, respectively '.', '..' and  
  '...'. For instance, command 'blur..' is equivalent to  
  'blur[-2]'. These shortcuts work also when specifying command arguments.  
  * G'MIC commands invoked without '[selection]' are applied on all images  
  of the list, i.e. the default selection is '[0--1]' (except for command  
  input whose default selection is '[-1]'').  
  * Prepending a single hyphen '-' to a G'MIC command is allowed. This may  
  be useful to recognize command items more easily in a one-liner pipeline  
  (typically invoked from a shell).  
  * A G'MIC command prepended with a plus sign '+' does not act  
  in-place but inserts its result as one or several new images at the end  
  of the image list.  
  * There are two different types of commands that can be run by the G'MIC  
  interpreter:  
    -  Built-in commands are the hard-coded functionalities in the  
  interpreter core. They are thus compiled as binary code and run fast, most of  
  the time. Omitting an argument when invoking a built-in command is not  
  permitted, except if all following arguments are also omitted. For instance,  
  invoking 'plasma 10,,5' is invalid but 'plasma 10' is correct.  
    -  Custom commands, are defined as G'MIC pipelines of built-in or  
  other custom commands. They are parsed by the G'MIC interpreter, and thus run a  
  bit slower than built-in commands. Omitting arguments when invoking a custom  
  command is permitted. For instance, expressions 'flower ,,,100,,2' or  
  'flower ,' are correct.  

  * Most of the existing commands in G'MIC are actually defined as custom  
  commands.  
  * A user can easily add its own custom commands to the G'MIC interpreter (see  
  section  Adding Custom Commands for more details). New built-in commands  
  cannot be added (unless you modify the G'MIC interpreter source code and  
  recompile it).  

 7. Input/Output Properties
    -----------------------

  * G'MIC is able to read/write most of the classical image file formats,  
  including:  
    -  2D grayscale/color files: '.png', '.jpeg', '.gif',  
  '.pnm', '.tif', '.bmp', ...  
    -  3D volumetric files: '.dcm', '.hdr', '.nii',  
  '.cube', '.pan', '.inr', '.pnk', ...  
    -  Video files: '.mpeg', '.avi', '.mp4', '.mov',  
  '.ogg', '.flv', ...  
    -  Generic text or binary data files: '.gmz', '.cimg',  
  '.cimgz', 'flo', 'ggr', 'gpl', '.dlm',  
  '.asc', '.pfm', '.raw', '.txt', '.h'.  
    -  3D mesh files: '.off', '.obj' (output only)  

  * When dealing with color images, G'MIC generally reads, writes and displays  
  data using the usual sRGB color space.  
  * When loading a '.png' and '.tiff' file, the bit-depth of the  
  input image(s) is returned to the status.  
  * G'MIC is able to manage 3D objects that may be read from files or  
  generated by G'MIC commands. A 3D object is stored as a one-column scalar image  
  containing the object data, in the following order: { magic_number; sizes;  
  vertices; primitives; colors; opacities }. These 3D representations can be  
  then processed as regular images (see command split3d for accessing each  
  of these 3D object data separately).  
  * Be aware that usual file formats may be sometimes not adapted to store all  
  the available image data, since G'MIC uses float-valued image buffers. For  
  instance, saving an image that was initially loaded as a 16bits/channel image,  
  as a '.jpg' file will result in a loss of information. Use the  
  G'MIC-specific file extension '.gmz' to ensure that all data precision is  
  preserved when saving images.  
  * Sometimes, file options may/must be set for file formats:  
    -  Video files: Only sub-frames of an image sequence may be loaded,  
  using the input expression 'filename.ext,[first_frame[,last_frame[,  
  step]]]'. Set 'last_frame==-1' to tell it must be the last frame of  
  the video. Set 'step' to '0' to force an opened video file to be  
  opened/closed. Output framerate and codec can be also set by using the output  
  expression 'filename.avi,_fps,_codec,_keep_open' where 'keep_open'  
  can be { 0 | 1 }. 'codec' is a 4-char string (see  
  http://www.fourcc.org/codecs.php ) or '0' for the default codec.  
  'keep_open' tells if the output video file must be kept open for  
  appending new frames afterwards.  
    -  '.cimg[z]' files: Only crops and sub-images of .cimg files  
  can be loaded, using the input expressions 'filename.cimg,N0,N1',  
  'filename.cimg,N0,N1,x0,x1', 'filename.cimg,N0,N1,x0,y0,x1,y1',  
  'filename.cimg,N0,N1,x0,y0,z0,x1,y1,z1' or 'filename.cimg,N0,N1,x0,y0,  
  z0,c0,x1,y1,z1,c1'. Specifying '-1' for one coordinates stands for the  
  maximum possible value. Output expression 'filename.cimg[z][,datatype]'  
  can be used to force the output pixel type. 'datatype' can be { auto |  
  bool | uchar | char | ushort | short | uint | int | uint64 | int64 | float |  
  double }.  
    -  '.raw' binary files: Image dimensions and input pixel type  
  may be specified when loading '.raw' files with input expression  
  'filename.raw[,datatype][,width][,height[,depth[,dim[,offset]]]]]'. If no  
  dimensions are specified, the resulting image is a one-column vector with  
  maximum possible height. Pixel type can also be specified with the output  
  expression 'filename.raw[,datatype]'. 'datatype' can be the same as  
  for '.cimg[z]' files.  
    -  '.yuv' files: Image dimensions must be specified when loading,  
   and only sub-frames of an image sequence may be loaded, using the input  
  expression 'filename.yuv,width,height[,chroma_subsampling[,first_frame[,  
  last_frame[,step]]]'. 'chroma_subsampling' can be { 420 | 422 | 444  
  }. When saving, chroma subsampling mode can be specified with output  
  expression 'filename.yuv[,chroma_subsampling]'.  
    -  '.tiff' files: Only sub-images of multi-pages tiff files can  
  be loaded, using the input expression 'filename.tif,_first_frame,_last_frame,  
  _step'. Output expression 'filename.tiff,_datatype,_compression,  
  _force_multipage,_use_bigtiff' can be used to specify the output pixel type,  
  as well as the compression method. 'datatype' can be the same as for  
  '.cimg[z]' files. 'compression' can be  { none (default) | lzw |  
  jpeg }. 'force_multipage' can be { 0=no (default) | 1=yes }.  
  'use_bigtiff' can be { 0=no | 1=yes (default) }.  
    -  '.pdf' files: When loading a file, the rendering resolution  
  can be specified using the input expression 'filename.pdf,resolution',  
  where 'resolution' is an unsigned integer value.  
    -  '.gif' files: Animated gif files can be saved, using the  
  input expression 'filename.gif,fps>0,nb_loops'. Specify  
  'nb_loops=0' to get an infinite number of animation loops (this is the  
  default behavior).  
    -  '.jpeg' files: The output quality may be specified (in %),  
  using the output expression 'filename.jpg,30' (here, to get a 30% quality  
  output). '100' is the default.  
    -  '.mnc' files: The output header can set from another file,  
  using the output expression 'filename.mnc,header_template.mnc'.  
    -  '.pan', '.cpp', '.hpp', '.c' and '.h'  
  files: The output datatype can be selected with output expression  
  'filename[,datatype]'. 'datatype' can be the same as for  
  '.cimg[z]' files.  
    -  '.gmic' files: These filenames are assumed to be G'MIC custom  
  commands files. Loading such a file will add the commands it defines to the  
  interpreter. Debug information can be enabled/disabled by the input expression  
  'filename.gmic[,add_debug_info' where 'debug_info' can be {  
  0=false | 1=true }.  
    -  Inserting 'ext:' on the beginning of a filename (e.g.  
  'jpg:filename') forces G'MIC to read/write the file as it would have been  
  done if it had the specified extension '.ext'.  

  * Some input/output formats and options may not be supported, depending on the  
  configuration flags that have been set during the build of the G'MIC software.  

 8. Substitution Rules
    ------------------

  * G'MIC items containing '$' or '{}' are substituted before  
  being interpreted. Use these substituting expressions to access various data  
  from the interpreter environment.  
  * '$name' and '${name}' are both substituted by the value of the  
  specified named variable (set previously by the item 'name=value'). If  
  this variable has not been already set, the expression is substituted by the  
  highest positive index of the named image '[name]'. If no image has this  
  name, the expression is substituted by the value of the OS environment variable  
  with same name (it may be thus an empty string if it is not defined).  
  * The following reserved variables are predefined by the G'MIC interpreter:  
    -  '$!': The current number of images in the list.  
    -  '$>' and '$<': The increasing/decreasing index of the latest  
  (currently running) 'repeat...done' loop. '$>' goes from '0'  
  (first loop iteration) to 'nb_iterations - 1' (last iteration).  
  '$<' does the opposite.  
    -  '$/': The current call stack. Stack items are separated by slashes  
  '/'.  
    -  '$|': The current value (expressed in seconds) of a millisecond  
  precision timer.  
    -  '$^': The current verbosity level.  
    -  '$_cpus': The number of computation cores available on your machine.  
    -  '$_flags': The list of enabled flags when G'MIC interpreter has  
  been compiled.  
    -  '$_host': A string telling about the host running the G'MIC  
  interpreter (e.g. 'cli' or 'gimp').  
    -  '$_os': A string describing the running operating system.  
    -  '$_path_rc': The path to the G'MIC folder used to store  
  configuration files (its value is OS-dependent).  
    -  '$_path_user': The path to the G'MIC user file '.gmic' or  
  'user.gmic' (its value is OS-dependent).  
    -  '$_path_commands': A list of all imported command files (stored as  
  a list-valued variable).  
    -  '$_pid': The current process identifier, as an integer.  
    -  '$_pixeltype': The type of image pixels (default: 'float').  
    -  '$_prerelease': For pre-releases, the date of the pre-release as  
  'yymmdd'. For stable releases, this variable is set to '0'.  
    -  '$_version': A 3-digits number telling about the current version of  
  the G'MIC interpreter  (e.g. '299').  
    -  '$_vt100': Set to '1' if colored text output is allowed on  
  the console. Otherwise, set to '0'.  

  * '$$name' and '$${name}' are both substituted by the G'MIC script  
  code of the specified named 'custom command', or by an empty string if no  
  custom command with specified name exists.  
  * '${"-pipeline"}' is substituted by the status value after the  
  execution of the specified G'MIC pipeline (see command status).  
  Expression '${}' thus stands for the current status value.  
  * '{``string}' (starting with two backquotes) is substituted by a  
  double-quoted version of the specified string.  
  * '{/string}' is substituted by the escaped version of the specified  
  string.  
  * '{'string'[:delimiter]}' (between single quotes) is substituted by the  
  sequence of character codes that composes the specified string, separated by  
  specified delimiter. Possible delimiters are ',' (default), ';',  
  '/', '^' or ' '. For instance, item '{'foo'}' is substituted  
  by '102,111,111' and '{'foo':;}' by '102;111;111'.  
  * '{image,feature[:delimiter]}' is substituted by a specific feature of  
  the image '[image]'. 'image' can be either an image number or an  
  image name. It can be also eluded, in which case, the last image '[-1]'  
  of the list is considered for the requested feature. Specified 'feature'  
  can be one of:  
    -  'b': The image basename (i.e. filename without the folder path nor  
  extension).  
    -  'f': The image folder name.  
    -  'n': The image name or filename (if the image has been read from a  
  file).  
    -  't': The text string from the image values regarded as character  
  codes.  
    -  'x': The image extension (i.e the characters after the last  
  '.' in the image name).  
    -  '^': The sequence of all image values, separated by commas ',  
  '.  
    -  '@subset': The sequence of image values corresponding to the  
  specified subset, and separated by commas ','.  
    -  Any other 'feature' is considered as a mathematical  
  expression associated to the image '[image]' and is substituted by the  
  result of its evaluation (float value). For instance, expression '{0,  
  w+h}' is substituted by the sum of the width and height of the first image  
  (see section Mathematical Expressions for more details). If a  
  mathematical expression starts with an underscore '_', the resulting  
  value is truncated to a readable format. For instance, item '{_pi}' is  
  substituted by '3.14159' (while '{pi}' is substituted by  
  '3.141592653589793').  
    -  A 'feature' delimited by backquotes is replaced by a string whose  
  character codes correspond to the list of values resulting from the evaluation  
  of the specified mathematical expression. For instance, item '{`[102,111,  
  111]`}' is substituted by 'foo' and item '{`vector8(65)`}' by  
  'AAAAAAAA'.  

  * '{*}' is substituted by the visibility state of the instant display  
  window '#0' (can be { 0=closed | 1=visible }.  
  * '{*[index],feature1,...,featureN[:delimiter]}' is substituted by a  
  specific set of features of the instant display window '#0' (or  
  '#index', if specified). Requested 'features' can be:  
    -  'w': display width (i.e. width of the display area managed by the  
  window).  
    -  'h': display height (i.e. height of the display area managed by the  
  window).  
    -  'wh': display width x display height.  
    -  'd': window width (i.e. width of the window widget).  
    -  'e': window height (i.e. height of the window widget).  
    -  'de': window width x window height.  
    -  'u': screen width (actually independent on the window size).  
    -  'v': screen height (actually independent on the window size).  
    -  'uv': screen width x screen height.  
    -  'n': current normalization type of the instant display.  
    -  't': window title of the instant display.  
    -  'x': X-coordinate of the mouse position (or -1, if outside the  
  display area).  
    -  'y': Y-coordinate of the mouse position (or -1, if outside the  
  display area).  
    -  'b': state of the mouse buttons { 1=left-but. | 2=right-but. |  
  4=middle-but. }.  
    -  'o': state of the mouse wheel.  
    -  'k': decimal code of the pressed key if any, 0 otherwise.  
    -  'c': boolean (0 or 1) telling if the instant display has been  
  closed recently.  
    -  'r': boolean telling if the instant display has been resized  
  recently.  
    -  'm': boolean telling if the instant display has been moved recently.  
    -  Any other 'feature' stands for a keycode name (in capital letters),  
  and is substituted by a boolean describing the current key state { 0=pressed  
  | 1=released }.  
    -  You can also prepend a hyphen '-' to a 'feature' (that  
  supports it) to flush the corresponding event immediately after reading its  
  state (works for keys, mouse and window events).  

  * Item substitution is never performed in items between double quotes.  
  One must break the quotes to enable substitution if needed, as in '"3+8 kg =  
  "{3+8}" kg"'. Using double quotes is then a convenient way to disable the  
  substitutions mechanism in items, when necessary.  
  * One can also disable the substitution mechanism on items outside double  
  quotes, by escaping the '{', '}' or '$' characters, as in  
  '3+4 doesn't evaluate'.  

 9. Mathematical Expressions
    ------------------------

  * G'MIC has an embedded mathematical parser, used to evaluate  
  (possibly complex) math expressions specified inside braces '{}', or  
  formulas in commands that may take one as an argument (e.g. fill or  
  eval).  
  * When the context allows it, a formula is evaluated for each pixel of  
  the selected images (e.g. fill or eval).  
  * A math expression may return a scalar or a vector-valued result  
  (with a fixed number of components).  
 The mathematical parser understands the following set of functions, operators  
  and variables:  

 ## Usual operators:  

 '||' (logical or), '&&' (logical and), '|' (bitwise or),  
  '&' (bitwise and), '!=', '==', '<=', '>=',  
  '<', '>', '<<' (left bitwise shift), '>>' (right  
  bitwise shift), '-', '+', '*', '/', '%' (modulo),  
  '^' (power), '!' (logical not), '~' (bitwise not), '++',  
   '--', '+=', '-=', '*=', '/=', '%=',  
  '&=', '|=', '^=', '>>', '<<=' (in-place  
  operators).  

 ## Usual math functions:  

 'abs()', 'acos()', 'acosh()', 'arg()', 'arg0()',  
  'argkth()', 'argmax()', 'argmaxabs()', 'argmin()',  
  'argminabs()', 'asin()', 'asinh()', 'atan()',  
  'atan2()', 'atanh()', 'avg()', 'bool()', 'cbrt()',  
   'ceil()', 'cos()', 'cosh()', 'cut()',  
  'deg2rad()', 'erf()', 'erfinv()', 'exp()',  
  'fact()', 'fibo()', 'floor()', 'gauss()', 'gcd()',  
   'int()', 'isnan()', 'isnum()', 'isinf()',  
  'isint()', 'isbool()', 'isexpr()', 'isfile()',  
  'isdir()', 'isin()', 'kth()', 'log()', 'log2()',  
  'log10()', 'max()', 'maxabs()', 'med()', 'min()',  
  'minabs()', 'narg()', 'prod()', 'rad2deg()',  
  'rol()' (left bit rotation), 'ror()' (right bit rotation),  
  'round()', 'sign()', 'sin()', 'sinc()', 'sinh()',  
  'sqrt()', 'std()', 'srand(_seed)', 'sum()',  
  'tan()', 'tanh()', 'var()', 'xor()'.  

  * 'atan2(y,x)' is the version of 'atan()' with two arguments  
  'y' and 'x' (as in C/C++).  
  * 'permut(k,n,with_order)' computes the number of permutations of  
  'k' objects from a set of 'n' objects.  
  * 'gauss(x,_sigma,_is_normalized)' returns  
  'exp(-x^2/(2*s^2))/(is_normalized?sqrt(2*pi*sigma^2):1)'.  
  * 'cut(value,min,max)' returns 'value' if it is in range '[min,  
  max]', or 'min' or 'max' otherwise.  
  * 'narg(a_1,...,a_N)' returns the number of specified arguments (here,  
  'N').  
  * 'arg(i,a_1,..,a_N)' returns the 'i'-th argument 'a_i'.  
  * 'isnum()', 'isnan()', 'isinf()', 'isint()',  
  'isbool()' test the type of the given number or expression, and return  
  '0' (false) or '1' (true).  
  * 'isfile('path')' (resp. 'isdir('path')') returns '0'  
  (false) or '1' (true) whether its string argument is a path to an  
  existing file (resp. to a directory) or not.  
  * 'isin(v,a_1,...,a_n)' returns '0' (false) or '1' (true)  
  whether the first value 'v' appears in the set of other values  
  'a_i'.  
  * 'inrange(value,m,M,include_m,include_M)' returns '0' (false) or  
  '1' (true) whether the specified value lies in range '[m,M]' or not  
  ('include_m' and 'includeM' tells how boundaries 'm' and  
  'M' are considered).  
  * 'argkth()', 'argmin()', 'argmax()', 'argminabs()',  
  'argmaxabs()'', 'avg()', 'kth()', 'min()',  
  'max()', 'minabs()', 'maxabs()', 'med()',  
  'prod()', 'std()', 'sum()' and 'var()' can be called  
  with an arbitrary number of scalar/vector arguments.  
  * 'vargkth()', 'vargmin()', 'vargmax()',  
  'vargminabs()', 'vargmaxabs()', 'vavg()', 'vkth()',  
  'vmin()', 'vmax()', 'vminabs()', 'vmaxabs()',  
  'vmed()', 'vprod()', 'vstd()', 'vsum()' and  
  'vvar()' are the versions of the previous function with vector-valued  
  arguments.  
  * 'round(value,rounding_value,direction)' returns a rounded value.  
  'direction' can be { -1=to-lowest | 0=to-nearest | 1=to-highest }.  
  * 'lerp(a,b,t)' returns 'a*(1-t)+b*t'.  
  * 'swap(a,b)' swaps the values of the given arguments.  

 ## Variable names:  

 Variable names below are pre-defined. They can be overridden.  
  * 'l': length of the associated list of images.  
  * 'k': index of the associated image, in '[0,l-1]'.  
  * 'w': width of the associated image, if any ('0' otherwise).  
  * 'h': height of the associated image, if any ('0' otherwise).  
  * 'd': depth of the associated image, if any ('0' otherwise).  
  * 's': spectrum of the associated image, if any ('0' otherwise).  
  * 'r': shared state of the associated image, if any ('0'  
  otherwise).  
  * 'wh': shortcut for width x height.  
  * 'whd': shortcut for width x height x depth.  
  * 'whds': shortcut for width x height x depth x spectrum (i.e. number of  
  image values).  
  * 'im', 'iM', 'ia', 'iv', 'is', 'ip',  
  'ic', 'in': Respectively the minimum, maximum, average, variance,  
  sum, product, median value and L2-norm of the associated image, if any  
  ('0' otherwise).  
  * 'xm', 'ym', 'zm', 'cm': The pixel coordinates of the  
  minimum value in the associated image, if any ('0' otherwise).  
  * 'xM', 'yM', 'zM', 'cM': The pixel coordinates of the  
  maximum value in the associated image, if any ('0' otherwise).  
  * All these variables are considered as constant values by the math  
  parser (for optimization purposes) which is indeed the case most of the time.  
  Anyway, this might not be the case, if function 'resize(#ind,..)' is used  
  in the math expression. If so, it is safer to invoke functions 'l()',  
  'w(_#ind)', 'h(_#ind)', ... 's(_#ind)' and 'in(_#ind)'  
  instead of the corresponding named variables.  
  * 'i': current processed pixel value (i.e. value located at '(x,y,z,  
  c)') in the associated image, if any ('0' otherwise).  
  * 'iN': N-th channel value of current processed pixel (i.e. value  
  located at '(x,y,z,N)' in the associated image, if any ('0'  
  otherwise). 'N' must be an integer in range '[0,9]'.  
  * 'R', 'G', 'B' and 'A' are equivalent to 'i0',  
  'i1', 'i2' and 'i3' respectively.  
  * 'I': current vector-valued processed pixel in the associated image, if  
  any ('0' otherwise). The number of vector components is equal to the  
  number of image channels (e.g. 'I' = '[ R,G,B ]' for a 'RGB'  
  image).  
  * You may add '#ind' to any of the variable name above to retrieve the  
  information for any numbered image '[ind]' of the list (when this makes  
  sense). For instance 'ia#0' denotes the average value of the first image  
  of the list).  
  * 'x': current processed column of the associated image, if any  
  ('0' otherwise).  
  * 'y': current processed row of the associated image, if any ('0'  
  otherwise).  
  * 'z': current processed slice of the associated image, if any  
  ('0' otherwise).  
  * 'c': current processed channel of the associated image, if any  
  ('0' otherwise).  
  * 't': thread id when an expression is evaluated with multiple threads  
  ('0' means master thread).  
  * 'n': maximum number of threads when expression is evaluated in  
  parallel (so that 't' goes from '0' to 'n-1').  
  * 'e': value of e, i.e. '2.71828...'.  
  * 'pi': value of pi, i.e. '3.1415926...'.  
  * 'u': a random value between '[0,1]', following a uniform  
  distribution.  
  * 'g': a random value, following a gaussian distribution of variance 1  
  (roughly in '[-6,6]').  
  * 'interpolation': value of the default interpolation mode used when  
  reading pixel values with the pixel access operators (i.e. when the  
  interpolation argument is not explicitly specified, see below for more details  
  on pixel access operators). Its initial default value is '0'.  
  * 'boundary': value of the default boundary conditions used when reading  
  pixel values with the pixel access operators (i.e. when the boundary condition  
  argument is not explicitly specified, see below for more details on pixel  
  access operators). Its initial default value is '0'.  
  * The last image of the list is always associated to the evaluations of  
  'expressions', e.g. G'MIC sequence  

   256,128 fill {w}  

  will create a 256x128 image filled with value 256.  

 ## Vector calculus:  

 Most operators are also able to work with vector-valued elements.  
  * '[a0,a1,...,aN-1]' defines a 'N'-dimensional vector with scalar  
  coefficients 'ak'.  
  * 'vectorN(a0,a1,,...,aN-1)' does the same, with the 'ak' being  
  repeated periodically if only a few are specified.  
  * 'vector(#N,a0,a1,,...,aN-1)' does the same, and can be used for any  
  constant expression 'N'.  
  * In previous expressions, the 'ak' can be vectors themselves, to be  
  concatenated into a single vector.  
  * The scalar element 'ak' of a vector 'X' is retrieved by  
  'X[k]'.  
  * The sub-vector '[X[p],X[p+s]...X[p+s*(q-1)]]' (of size 'q') of a  
  vector 'X' is retrieved by 'X[p,q,s]'.  
  * 'expr(formula,_w,_h,_d,_s)' outputs a vector of size 'w*h*d*s'  
  with values generated from the specified formula, as if one were filling an  
  image with dimensions '(w,h,d,s)'.  
  * Equality/inequality comparisons between two vectors is done with operators  
  '==' and '!='.  
  * Some vector-specific functions can be used on vector values: 'cross(X,  
  Y)' (cross product), 'dot(X,Y)' (dot product), 'size(X)' (vector  
  dimension), 'sort(X,_is_increasing,_nb_elts,_size_elt)' (sorted values),  
  'reverse(A)' (reverse order of components), 'shift(A,_length,  
  _boundary_conditions)' and 'same(A,B,_nb_vals,_is_case_sensitive)'  
  (vector equality test).  
  * Function 'normP(u1,...,un)' computes the LP-norm of the specified  
  vector ('P' being an 'unsigned integer' constant or 'inf').  
  If 'P' is omitted, the L2 norm is calculated.  
  * Function 'resize(A,size,_interpolation,_boundary_conditions)' returns  
  a resized version of a vector 'A' with specified interpolation mode.  
  'interpolation' can be { -1=none (memory content) | 0=none | 1=nearest  
  | 2=average | 3=linear | 4=grid | 5=bicubic | 6=lanczos }, and  
  'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic |  
  3=mirror }.  
  * Function 'find(A,B,_starting_index,_search_step)' returns the index  
  where sub-vector 'B' appears in vector 'A', (or '-1' if  
  'B' is not contained in 'A'). Argument 'A' can be also  
  replaced by an image index '#ind'.  
  * A '2'-dimensional vector may be seen as a complex number and used in  
  those particular functions/operators: '**' (complex multiplication),  
  '//' (complex division), '^^' (complex exponentiation), '**='  
  (complex self-multiplication), '//=' (complex self-division), '^^='  
  (complex self-exponentiation), 'cabs()' (complex modulus), 'carg()'  
  (complex argument), 'cconj()' (complex conjugate), 'cexp()'  
  (complex exponential), 'clog()' (complex logarithm),  'ccos()'  
  (complex cosine), 'csin()' (complex sine), 'ctan()' (complex  
  tangent), 'ccosh()' (complex hyperpolic cosine), 'csinh()' (complex  
  hyperbolic sine) and 'ctanh()' (complex hyperbolic tangent).  
  * A 'MN'-dimensional vector may be seen as a 'M' x 'N'  
  matrix and used in those particular functions/operators: '*'  
  (matrix-vector multiplication), 'det(A)' (determinant), 'diag(V)'  
  (diagonal matrix from a vector), 'eig(A)' (eigenvalues/eigenvectors),  
  'eye(n)' (n x n identity matrix), 'invert(A,_solver)' (matrix  
  inverse), 'mul(A,B,_nb_colsB)' (matrix-matrix multiplication),  
  'pseudoinvert(A,_nb_colsA,_solver)', 'rot(u,v,w,angle)' (3D  
  rotation matrix), 'rot(angle)' (2D rotation matrix), 'solve(A,B,  
  _nb_colsB)' (solver of linear system A.X = B), 'svd(A,_nb_colsA)'  
  (singular value decomposition), 'trace(A)' (matrix trace) and  
  'transpose(A,nb_colsA)' (matrix transpose). Argument 'nb_colsB' may  
  be omitted if it is equal to '1'.  
  * 'mproj(S,nb_colsS,D,nb_colsD,method,max_iter,max_residual)' projects a  
  matrix 'S' onto a dictionary (matrix) 'D'. Equivalent to command  
  mproj but inside the math evaluator.  
  * Specifying a vector-valued math expression as an argument of a command that  
  operates on image values (e.g. 'fill') modifies the whole spectrum range  
  of the processed image(s), for each spatial coordinates '(x,y,z)'. The  
  command does not loop over the 'c'-axis in this case.  

 ## String manipulation:  

 Character strings are defined and managed as vectors objects. Dedicated  
  functions and initializers to manage strings are:  
  * '['string']' and ''string'' define a vector whose values are the  
  character codes of the specified 'character string' (e.g. ''foo''  
  is equal to '[ 102,111,111 ]').  
  * '_'character'' returns the (scalar) byte code of the specified  
  character (e.g. '_'A'' is equal to '65').  
  * A special case happens for empty strings: Values of both expressions  
  '['']' and '''' are '0'.  
  * Functions 'lowercase()' and 'uppercase()' return string with all  
  string characters lowercased or uppercased.  
  * Function 'stov(str,_starting_index,_is_strict)' parses specified  
  string 'str' and returns the value contained in it.  
  * Function 'vtos(expr,_nb_digits,_siz)' returns a vector of size  
  'siz' which contains the character representation of values described by  
  expression 'expr'. 'nb_digits' can be { -1=auto-reduced | 0=all  
  | >0=max number of digits }.  
  * Function 'echo(str1,str2,...,strN)' prints the concatenation of given  
  string arguments on the console.  
  * Function 'string(_#siz,str1,str2,...,strN)' generates a vector  
  corresponding to the concatenation of given string/number arguments.  

 ## Special operators:  

  * ';': expression separator. The returned value is always the last  
  encountered expression. For instance expression '1;2;pi' is evaluated as  
  'pi'.  
  * '=': variable assignment. Variables in mathematical parser can only  
  refer to numerical values (vectors or scalars). Variable names are  
  case-sensitive. Use this operator in conjunction with ';' to define more  
  complex evaluable expressions, such as  

   t = cos(x); 3*t^2 + 2*t + 1  

 These variables remain local to the mathematical parser and cannot be  
  accessed outside the evaluated expression.  
  * Variables defined in math parser may have a constant property, by  
  specifying keyword 'const' before the variable name (e.g. 'const foo =  
  pi/4;'). The value set to such a variable must be indeed a constant  
  scalar. Constant variables allows certain types of optimizations in the math  
  JIT compiler.  

 ## Specific functions:  

  * 'addr(expr)': return the pointer address to the specified expression  
  'expr'.  
  * 'fill(target,expr)' or 'fill(target,index_name,expr)' fill the  
  content of the specified target (often vector-valued) using a given expression,  
  e.g. 'V = vector16(); fill(V,k,k^2 + k + 1);'. For a vector-valued target,  
   it is basically equivalent to: 'for (index_name = 0,  
  index_name<size(target), ++index_name, target[index_name] = expr);'.  
  * 'u(max)' or 'u(min,max)': return a random value between '[0,  
  max]' or '[min,max]', following a uniform distribution.  
  * 'f2ui(value)' and 'ui2f(value)': Convert a large unsigned  
  integer as a negative floating point value (and vice-versa), so that 32bits  
  floats can be used to store large integers while keeping a unitary precision.  
  * 'i(_a,_b,_c,_d,_interpolation_type,_boundary_conditions)': return the  
  value of the pixel located at position '(a,b,c,d)' in the associated  
  image, if any ('0' otherwise). 'interpolation_type' can be {  
  0=nearest neighbor | 1=linear | 2=cubic }. 'boundary_conditions' can  
  be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }. Omitted  
  coordinates are replaced by their default values which are respectively  
  'x', 'y', 'z', 'c', 'interpolation' and  
  'boundary'. For instance command  

   fill 0.5*(i(x+1)-i(x-1))  

  will estimate the X-derivative of an image with a classical finite difference  
  scheme.  
  * 'j(_dx,_dy,_dz,_dc,_interpolation_type,_boundary_conditions)' does the  
  same for the pixel located at position '(x+dx,y+dy,z+dz,c+dc)' (pixel  
  access relative to the current coordinates).  
  * 'i[offset,_boundary_conditions]' returns the value of the pixel  
  located at specified 'offset' in the associated image buffer (or  
  '0' if offset is out-of-bounds).  
  * 'j[offset,_boundary_conditions]' does the same for an offset relative  
  to the current pixel coordinates '(x,y,z,c)'.  
  * 'i(#ind,_x,_y,_z,_c,_interpolation,_boundary_conditions)', 'j(#ind,  
  _dx,_dy,_dz,_dc,_interpolation,_boundary_conditions)', 'i[#ind,offset,  
  _boundary_conditions]' and 'i[offset,_boundary_conditions]' are  
  similar expressions used to access pixel values for any numbered image  
  '[ind]' of the list.  
  * 'I/J[offset,_boundary_conditions]' and 'I/J(#ind,_x,_y,_z,  
  _interpolation,_boundary_conditions)' do the same as 'i/j[offset,  
  _boundary_conditions]' and 'i/j(#ind,_x,_y,_z,_c,_interpolation,  
  _boundary_conditions)' but return a vector instead of a scalar (e.g. a  
  vector '[ R,G,B ]' for a pixel at '(a,b,c)' in a color image).  
  * 'crop(_#ind,_x,_y,_z,_c,_dx,_dy,_dz,_dc,_boundary_conditions)' returns  
  a vector whose values come from the cropped region of image '[ind]' (or  
  from default image selected if 'ind' is not specified). Cropped region  
  starts from point '(x,y,z,c)' and has a size of 'dx x dy x dz x  
  dc'. Arguments for coordinates and sizes can be omitted if they are not  
  ambiguous (e.g. 'crop(#ind,x,y,dx,dy)' is a valid invocation of this  
  function).  
  * 'draw(_#ind,S,x,y,z,c,dx,_dy,_dz,_dc,_opacity,_M,_max_M)' draws a  
  sprite 'S' in image '[ind]' (or in default image selected if  
  'ind' is not specified) at coordinates '(x,y,z,c)'. The size of the  
  sprite 'dx x dy x dz x dc' must be specified. You can also specify a  
  corresponding opacity mask 'M' if its size matches 'S'.  
  * 'polygon(_#ind,nb_vertices,coords,_opacity,_color)' draws a filled  
  polygon in image '[ind]' (or in default image selected if 'ind' is  
  not specified) at specified coordinates. It draws a single line if  
  'nb_vertices' is set to 2.  
  * 'polygon(_#ind,-nb_vertices,coords,_opacity,_pattern,_color)' draws a  
  outlined polygon in image '[ind]' (or in default image selected if  
  'ind' is not specified) at specified coordinates and with specified line  
  pattern. It draws a single line if 'nb_vertices' is set to 2.  
  * 'ellipse(_#ind,xc,yc,radius1,_radius2,_angle,_opacity,_color)' draws a  
  filled ellipse in image '[ind]' (or in default image selected if  
  'ind' is not specified) with specified coordinates.  
  * 'ellipse(_#ind,xc,yc,-radius1,-_radius2,_angle,_opacity,_pattern,  
  _color)' draws an outlined ellipse in image '[ind]' (or in default  
  image selected if 'ind' is not specified).  
  * 'resize(#ind,w,_h,_d,_s,_interp,_boundary_conditions,_cx,_cy,_cz,_cc)'  
  resizes an image of the associated list with specified dimension and  
  interpolation method. When using this function, you should consider retrieving  
  the (non-constant) image dimensions using the dynamic functions  
  'w(_#ind)', 'h(_#ind)', 'd(_#ind)', 's(_#ind)',  
  'wh(_#ind)', 'whd(_#ind)' and 'whds(_#ind)' instead of the  
  corresponding constant variables.  
  * 'if(condition,expr_then,_expr_else)': return value of  
  'expr_then' or 'expr_else', depending on the value of  
  'condition' { 0=false | other=true }. 'expr_else' can be  
  omitted in which case '0' is returned if the condition does not hold.  
  Using the ternary operator 'condition?expr_then[:expr_else]' gives an  
  equivalent expression. For instance, G'MIC commands  

   fill if(x%10==0,255,i)  

  and  

   fill x%10?i:255  

  both draw blank vertical lines on every 10th column of an image.  
  * 'do(expression,_condition)' repeats the evaluation of  
  'expression' until 'condition' vanishes (or until  
  'expression' vanishes if no 'condition' is specified). For instance,  
   the expression:  

   if(N<2,N,n=N-1;F0=0;F1=1;do(F2=F0+F1;F0=F1;F1=F2,n=n-1))  

  returns the N-th value of the Fibonacci sequence, for 'N>=0' (e.g.,  
  '46368' for 'N=24'). 'do(expression,condition)' always  
  evaluates the specified expression at least once, then check for the loop  
  condition. When done, it returns the last value of 'expression'.  
  * 'for(init,condition,_procedure,body)' first evaluates the expression  
  'init', then iteratively evaluates 'body' (followed by  
  'procedure' if specified) while 'condition' holds (i.e. not zero).  
  It may happen that no iterations are done, in which case the function returns  
  'nan'. Otherwise, it returns the last value of 'body'. For instance,  
   the expression:  

   if(N<2,N,for(n=N;F0=0;F1=1,n=n-1,F2=F0+F1;F0=F1;F1=F2))  

  returns the 'N'-th value of the Fibonacci sequence, for 'N>=0'  
  (e.g., '46368' for 'N=24').  
  * 'while(condition,expression)' is exactly the same as 'for(init,  
  condition,expression)' without the specification of an initializing  
  expression.  
  * 'repeat(nb_iters,expr)' or 'fill(nb_iters,iter_name,expr)' run  
  'nb_iters' iterations of the specified expression 'expr', e.g.  
  'V = vector16(); repeat(16,k,V[k] = k^2 + k + 1);'. It is basically  
  equivalent to: 'for (iter_name = 0, iter_name<nb_iters, ++iter_name,  
  expr);'.  
  * 'break()' and 'continue()' respectively breaks and continues the  
  current running bloc (loop, init or main environment).  
  * 'fsize('filename')' returns the size of the specified 'filename'  
  (or '-1' if file does not exist).  
  * 'date(attr,'path')' returns the date attribute for the given  
  'path' (file or directory), with 'attr' being { 0=year | 1=month  
  | 2=day | 3=day of week | 4=hour | 5=minute | 6=second }, or a vector of  
  those values.  
  * 'date(_attr)' returns the specified attribute for the current (locale)  
  date (attributes being { 0...6=same meaning as above | 7=milliseconds }).  
  * 'print(expr1,expr2,...)' or 'print(#ind)' prints the value of  
  the specified expressions (or image information) on the console, and returns  
  the value of the last expression (or 'nan' in case of an image). Function  
  'prints(expr)' also prints the string composed of the character codes  
  defined by the vector-valued expression (e.g. 'prints('Hello')').  
  * 'debug(expression)' prints detailed debug info about the sequence of  
  operations done by the math parser to evaluate the expression (and returns its  
  value).  
  * 'display(_X,_w,_h,_d,_s)' or 'display(#ind)' display the  
  contents of the vector 'X' (or specified image) and wait for user events.  
  if no arguments are provided, a memory snapshot of the math parser environment  
  is displayed instead.  
  * 'begin(expression)' and 'end(expression)' evaluates the  
  specified expressions only once, respectively at the beginning and end of the  
  evaluation procedure, and this, even when multiple evaluations are required  
  (e.g. in 'fill ">begin(foo = 0); ++foo"').  
  * 'copy(dest,src,_nb_elts,_inc_d,_inc_s,_opacity)' copies an entire  
  memory block of 'nb_elts' elements starting from a source value  
  'src' to a specified destination 'dest', with increments defined by  
  'inc_d' and 'inc_s' respectively for the destination and source  
  pointers.  
  * 'stats(_#ind)' returns the statistics vector of the running image  
  '[ind]', i.e the vector '[ im,iM,ia,iv,xm,ym,zm,cm,xM,yM,zM,cM,is,ip  
  ]' (14 values).  
  * 'ref(expr,a)' references specified expression 'expr' as variable  
  name 'a'.  
  * 'unref(a,b,...)' destroys references to the named variable given as  
  arguments.  
  * 'breakpoint()' inserts a possible computation breakpoint (useless with  
  the cli interface).  
  * '_(comment) expr' just returns expression 'expr' (useful for inserting  
  inline comments in math expressions).  
  * 'run('pipeline')' executes the specified G'MIC pipeline as if it was  
  called outside the currently evaluated expression.  
  * 'set(A,'variable_name')' set the G'MIC variable '$variable_name'  
  with the value of expression 'A'. If 'A' is a vector-valued  
  variable, it is assumed to encode a string.  
  * 'store(A,'variable_name',_w,_h,_d,_s,_is_compressed)' transfers the  
  data of vector 'A' as a 'w x h x d x s' image to the G'MIC variable  
  '$variable_name'. Thus, the data becomes available outside the math  
  expression (that is equivalent to using the regular command store, but  
  directly in the math expression).  
  * 'get('variable_name',_size,_return_as_string)' returns the value of  
  the specified variable, as a vector of 'size' values, or as a scalar (if  
  'size' is zero or not specified).  
  * 'name(_#ind,size)' returns a vector of size 'size', whose values  
  are the characters codes of the name of image '[ind]' (or default image  
  selected if 'ind' is not specified).  
  * 'correlate(I,wI,hI,dI,sI,K,wK,hK,dK,sK,_boundary_conditions,  
  _is_normalized,_channel_mode,_xcenter,_ycenter,_zcenter,_xstart,_ystart,_zstart,  
  _xend,_yend,_zend,_xstride,_ystride,_zstride,_xdilation,_ydilation,_zdilation,  
  _interpolation_type)' returns the correlation, unrolled as a vector, of the  
  'wI x hI x dI x sI'-sized image 'I' with the 'wK x hK x dK x  
  sK'-sized kernel 'K' (the meaning of the other arguments are the same  
  as in command 'correlate'). Similar function 'convolve(...)' is  
  also defined for computing the convolution between 'I' and 'K'.  

 ## User-defined macros:  

  * Custom macro functions can be defined in a math expression, using the  
  assignment operator '=', e.g.  

   foo(x,y) = cos(x + y); result = foo(1,2) + foo(2,3)

  * Trying to override a built-in function (e.g. 'abs()') has no effect.  
  * Overloading macros with different number of arguments is possible.  
  Re-defining a previously defined macro with the same number of arguments  
  discards its previous definition.  
  * Macro functions are indeed processed as macros by the mathematical  
  evaluator. You should avoid invoking them with arguments that are themselves  
  results of assignments or self-operations. For instance,  

   foo(x) = x + x; z = 0; foo(++z)  

  returns '4' rather than expected value '2'.  
  * When substituted, macro arguments are placed inside parentheses, except if a  
  number sign '#' is located just before or after the argument name. For  
  instance, expression  

   foo(x,y) = x*y; foo(1+2,3)  

  returns '9' (being substituted as '(1+2)*(3)'), while expression  

   foo(x,y) = x#*y#; foo(1+2,3)  

  returns '7' (being substituted as '1+2*3').  
  * Number signs appearing between macro arguments function actually count for  
  empty separators. They may be used to force the substitution of macro  
  arguments in unusual places, e.g. as in  

   str(N) = ['I like N#'];

 ## Multi-threaded and in-place evaluation:  

  * If your image data are large enough and you have several CPUs available, it  
  is likely that the math expression passed to a 'fill', 'eval' or  
  'input' commands is evaluated in parallel, using multiple computation  
  threads.  
  * Starting an expression with ':' or '*' forces the evaluations  
  required for an image to be run in parallel, even if the amount of data to  
  process is small (beware, it may be slower to evaluate in this case!). Specify  
  ':' (rather than '*') to avoid possible image copy done before  
  evaluating the expression (this saves memory, but do this only if you are sure  
  this step is not required!)  
  * If the specified expression starts with '>' or '<', the pixel  
  access operators 'i()', 'i[]', 'j()' and 'j[]' return  
  values of the image being currently modified, in forward ('>') or  
  backward ('<') order. The multi-threading evaluation of the expression is  
  disabled in this case.  
  * Function 'critical(expr)' forces the execution of the given expression  
  in a single thread at a time.  
  * 'begin_t(expr)' and 'end_t(expr)' evaluates the specified  
  expression once for each running thread (so possibly several times) at the  
  beginning and the end of the evaluation procedure.  
  * 'merge(variable,operator)' tells to merge the local variable value  
  computed by threads, with the specified operator, when all threads have  
  finished computing.  
  * Expressions 'i(_#ind,x,_y,_z,_c)=value', 'j(_#ind,x,_y,_z,  
  _c)=value', 'i[_#ind,offset]=value' and 'j[_#ind,offset]=value'  
  set a pixel value at a different location than the running one in the image  
  '[ind]' (or in the associated image if argument '#ind' is omitted),  
  either with global coordinates/offsets (with 'i(...)' and 'i[...]'),  
   or relatively to the current position '(x,y,z,c)' (with 'j(...)'  
  and 'j[...]'). These expressions always return 'value'.  

 10. Image and Data Viewers
     ----------------------

  * G'MIC has some very handy embedded visualization modules, for 1D  
  signals (command plot), 1D/2D/3D images (command display) and 3D  
  vector objects (command display3d). It manages interactive views of the  
  selected image data.  
  * The following actions are available in the interactive viewers:  
    -  '(mousewheel)': Zoom in/out.  
    -  'ESC': Close window.  
    -  'CTRL+D': Increase window size.  
    -  'CTRL+C': Decrease window size.  
    -  'CTRL+R': Reset window size.  
    -  'CTRL+F': Toggle fullscreen mode.  
    -  'CTRL+S': Save current view as a numbered file  
  'gmic_xxxx.ext'.  
    -  'CTRL+O': Save copy of the viewed data, as a numbered file  
  'gmic_xxxx.ext'.  

  * Actions specific to the 1D/2D image viewer (command display) are:  
    -  'Left mouse button': Create an image selection and zoom into it.  
    -  'Middle mouse button', or 'CTRL+left mouse button': Move  
  image.  
    -  'Mouse wheel' or 'PADD+/-': Zoom in/out.  
    -  'Arrow keys': Move image left/right/up/down.  
    -  'CTRL+A': Enable/disable transparency (show alpha channel).  
    -  'CTRL+N': Change normalization mode (can be { none | normal |  
  channel-by-channel }).  
    -  'CTRL+SPACE': Reset view.  
    -  'CTRL+X': Show/hide axes.  
    -  'CTRL+Z': Hold/release aspect ratio.  

  * Actions specific to the 3D volumetric image viewer (command  
  display) are:  
    -  'CTRL+P': Play z-stack of frames as a movie.  
    -  'CTRL+V': Show/hide 3D view on bottom right zone.  
    -  'CTRL+X': Show/hide axes.  
    -  'CTRL+(mousewheel)': Go up/down.  
    -  'SHIFT+(mousewheel)': Go left/right.  
    -  'Numeric PAD': Zoom in/out ('+'/'-') and move through  
  zoomed image (digits).  
    -  'BACKSPACE': Reset zoom scale.  

  * Actions specific to the 3D object viewer (command display3d) are:  
    -  '(mouse)+(left mouse button)': Rotate 3D object.  
    -  '(mouse)+(right mouse button)': Zoom 3D object.  
    -  '(mouse)+(middle mouse button)': Shift 3D object.  
    -  'F1 ... F6': Toggle between different 3D rendering modes.  
    -  'F7/F8': Decrease/increase focale.  
    -  'F9': Select animation mode.  
    -  'F10': Select animation speed.  
    -  'SPACE': Start/stop animation.  
    -  'CTRL+A': Show/hide 3D axes.  
    -  'CTRL+B': Switch between available background.  
    -  'CTRL+G': Save 3D object, as numbered file 'gmic_xxxx.obj'.  
    -  'CTRL+L': Show/hide outline.  
    -  'CTRL+P': Print current 3D pose on stderr.  
    -  'CTRL+T': Switch between single/double-sided 3D modes.  
    -  'CTRL+V': Start animation with video output.  
    -  'CTRL+X': Show/hide 3D bounding box.  
    -  'CTRL+Z': Enable/disable z-buffered rendering.  

 11. Adding Custom Commands
     ----------------------

  * New custom commands can be added by the user, through the use of G'MIC  
  custom commands files.  
  * A command file is a simple text file, where each line starts either by  

   command_name: command_definition  

  or  

   command_definition (continuation)

  * At startup, G'MIC automatically includes user's command file  
  '$HOME/.gmic' (on Unix) or '%APPDATA%/user.gmic' (on  
  Windows). The CLI tool 'gmic' automatically runs the command  
  'cli_start' if defined.  
  * Custom command names must use character set '[a-zA-Z0-9_]' and cannot  
  start with a number.  
  * Any '# comment' expression found in a custom commands file is  
  discarded by the G'MIC parser, wherever it is located in a line.  
  * In a custom command, the following '$-expressions' are recognized and  
  substituted:  
    -  '$*' is substituted by a copy of the specified string of arguments.  
    -  '$"*"' is substituted by a copy of the specified string of  
  arguments, each being double-quoted.  
    -  '$#' is substituted by the maximum index of known arguments (either  
  specified by the user or set to a default value in the custom command).  
    -  '$[]' is substituted by the list of selected image indices that  
  have been specified in the command invocation.  
    -  '$?' is substituted by a printable version of '$[]' to be  
  used in command descriptions.  
    -  '$i' and '${i}' are both substituted by the 'i'-th  
  specified argument. Negative indices such as '${-j}' are allowed and  
  refer to the 'j'-th latest argument. '$0' is substituted by the  
  custom command name.  
    -  '${i=default}' is substituted by the value of '$i' (if  
  defined) or by its new value set to 'default' otherwise ('default'  
  may be a '$-expression' as well).  
    -  '${subset}' is substituted by the argument values (separated by  
  commas ',') of a specified argument subset. For instance expression  
  '${2--2}' is substituted by all specified command arguments except the  
  first and the last one. Expression '${^0}' is then substituted by all  
  arguments of the invoked command (eq. to '$*' if all arguments have been  
  indeed specified).  
    -  '$=var' is substituted by the set of instructions that will assign  
  each argument '$i' to the named variable 'var$i' (for i in  
  '[0...$#]'. This is particularly useful when a custom command want to  
  manage variable numbers of arguments. Variables names must use character set  
  '[a-zA-Z0-9_]' and cannot start with a number.  

  * These particular '$-expressions' for custom commands are always  
  substituted, even in double-quoted items or when the dollar sign '$'  
  is escaped with a backslash '$'. To avoid substitution, place an empty  
  double quoted string just after the '$' (as in '$""1').  
  * Specifying arguments may be skipped when invoking a custom command, by  
  replacing them by commas ',' as in expression  

   flower ,,3  

  Omitted arguments are set to their default values, which must be thus  
  explicitly defined in the code of the corresponding custom command (using  
  default argument expressions as '${1=default}').  
  * If one numbered argument required by a custom command misses a value, an  
  error is thrown by the G'MIC interpreter.  
  * It is possible to specialize the invokation of a '+command' by  
  defining it as  

   +command_name: command_definition  

  * A +-specialization takes priority over the regular command definition when  
  the command is invoked with a prepended '+'.  
  * When only a +-specialization of a command is defined, invoking  
  'command' is actually equivalent to '+command'.  

 12. List of Commands
     ----------------

 All available G'MIC commands are listed below, by categories. An argument  
  specified between '[]' or starting by '_' is optional except when  
  standing for an existing image '[image]', where 'image' can be  
  either an index number or an image name. In this case, the '[]'  
  characters are mandatory when writing the item. Note that all images that serve  
  as illustrations in this reference documentation are normalized in range '[0,  
  255]' before being displayed. You may need to do this explicitly (command  
  'normalize 0,255') if you want to save and view images with the same aspect  
  than those illustrated in the example codes.

 12.1. Global Options
       --------------

 debug (+):

   Activate debug mode.
   When activated, the G'MIC interpreter becomes very verbose and outputs additional log
   messages about its internal state on the standard output (stdout).
   This option is useful for developers or to report possible bugs of the interpreter.

 h:
     Shortcut for command 'help'.

 help:
     command |
     (no arg)

   Display help (optionally for specified command only) and exit.
   (equivalent to shortcut command 'h').

 version:

   Display current version number on stdout.

 12.2. Input / Output
       --------------

 camera (+):
     _camera_index>=0,_nb_frames>0,_skip_frames>=0,_capture_width>=0,_capture_height>=0

   Insert one or several frames from specified camera.
   When 'nb_frames==0', the camera stream is released instead of capturing new images.
   This command requires features from the OpenCV library (not enabled in G'MIC by default).

   Default values: 'camera_index=0' (default camera), 'nb_frames=1', 'skip_frames=0' and 'capture_width=capture_height=0' (default size).

 clut:
     "clut_name",_resolution>0,_cut_and_round={ 0=no | 1=yes }

   Insert one of the 958 pre-defined CLUTs at the end of the image list.

   'clut_name' can be { 2-strip-process | 60s | 60s_faded | 60s_faded_alt | 7drk_21 | action_magenta_01 | action_red_01 | adventure_1453 | agfa_apx_100 | agfa_apx_25 | agfa_precisa_100 |
    agfa_ultra_color_100 | agfa_vista_200 | agressive_highligjtes_recovery_5 | alberto_street | alien_green | amstragram | amstragram+ | analog_film_1 | analogfx_anno_1870_color |
    analogfx_old_style_i | analogfx_old_style_ii | analogfx_old_style_iii | analogfx_sepia_color | analogfx_soft_sepia_i | analogfx_soft_sepia_ii | anime | apocalypse_this_very_moment | aqua |
    aqua_and_orange_dark | arabica_12 | atomic_pink | autumn | ava_614 | avalanche | azrael_93 | bboyz_2 | bc_darkum | beach_aqua_orange | beach_faded_analog | berlin_sky | black_and_white |
    black_star | black_white_01 | black_white_02 | black_white_03 | black_white_04 | black_white_05 | black_white_06 | blade_runner | bleach_bypass | bleachbypass_1 | bleachbypass_2 |
    bleachbypass_3 | bleachbypass_4 | bleech_bypass_green | bleech_bypass_yellow_01 | blue_cold_fade | blue_dark | blue_house | blue_ice | blue_mono | blue_shadows_01 | blues | bob_ford |
    bourbon_64 | bright_green_01 | bright_teal_orange | bright_warm | brightgreen | brown_mobster | brownbm | brownish | bw_1 | bw_10 | bw_2 | bw_3 | bw_4 | bw_5 | bw_6 | bw_7 | bw_8 | bw_9 |
    bw_but_yellow | byers_11 | candlelight | caribe | chemical_168 | chrome_01 | cineblue | cinebm_4k | cinema | cinema_2 | cinema_3 | cinema_4 | cinema_5 | cinema_noir | cinematic-1 |
    cinematic-10 | cinematic-2 | cinematic-3 | cinematic-4 | cinematic-5 | cinematic-6 | cinematic-7 | cinematic-8 | cinematic-9 | cinematic_01 | cinematic_02 | cinematic_03 | cinematic_for_flog
    | cinematic_forest | cinematic_lady_bird | cinematic_mexico | city | city_7 | city_dust | classic_films_01 | classic_films_02 | classic_films_03 | classic_films_04 | classic_films_05 |
    classic_teal_and_orange | clayton_33 | clear_teal_fade | clouseau_54 | cobi_3 | coffee_44 | cold_clear_blue | cold_clear_blue_1 | cold_ice | cold_simplicity_2 | color_rich | colorful_0209 |
    colornegative | conflict_01 | contrail_35 | contrast_with_highlights_protection | contrasty_afternoon | contrasty_green | crispromance | crispwarm | crispwinter | cross_process_cp_130 |
    cross_process_cp_14 | cross_process_cp_15 | cross_process_cp_16 | cross_process_cp_18 | cross_process_cp_3 | cross_process_cp_4 | cross_process_cp_6 | crushin | cubicle_99 | d_o_1 |
    dark_blues_in_sunlight | dark_green_02 | dark_green_1 | dark_man_x | dark_orange_teal | dark_place_01 | darkness | date_39 | day_4nite | day_for_night | day_to_night_kings_blue | deep |
    deep_blue | deep_dark_warm | deep_high_contrast | deep_teal_fade | deep_warm_fade | deepskintones_2 | deepskintones_3 | delicatessen | denoiser_simple_40 | desert_gold_37 | dimension |
    directions_23 | django_25 | domingo_145 | dream_1 | dream_85 | drop_green_tint_14 | dropblues | duotone_blue_red | earth_tone_boost | edgyember | elegance_38 | enchanted | eterna_for_flog |
    expired_69 | expired_fade | expired_polaroid | extreme | fade | fade_to_green | faded | faded_47 | faded_alt | faded_analog | faded_extreme | faded_green | faded_pink-ish | faded_print |
    faded_retro_01 | faded_retro_02 | faded_vivid | fadedlook | fallcolors | faux_infrared | faux_infrared_bw_1 | faux_infrared_color_p_2 | faux_infrared_color_p_3 | faux_infrared_color_r_0a |
    faux_infrared_color_r_0b | faux_infrared_color_yp_1 | fgcinebasic | fgcinebright | fgcinecold | fgcinedrama | fgcinetealorange_1 | fgcinetealorange_2 | fgcinevibrant | fgcinewarm | film_0987
    | film_9879 | film_gb-19 | film_high_contrast | film_print_01 | film_print_02 | filmic | flat_30 | flat_blue_moon | flavin | foggynight | folger_50 | formula_b | french_comedy | frosted |
    frostedbeachpicnic | fuji_160c | fuji_160c_+ | fuji_160c_++ | fuji_160c_- | fuji_3510_constlclip | fuji_3510_constlmap | fuji_3510_cuspclip | fuji_3513_constlclip | fuji_3513_constlmap |
    fuji_3513_cuspclip | fuji_400h | fuji_400h_+ | fuji_400h_++ | fuji_400h_- | fuji_800z | fuji_800z_+ | fuji_800z_++ | fuji_800z_- | fuji_astia_100_generic | fuji_astia_100f | fuji_fp-100c |
    fuji_fp-100c_+ | fuji_fp-100c_++ | fuji_fp-100c_+++ | fuji_fp-100c_++_alt | fuji_fp-100c_- | fuji_fp-100c_-- | fuji_fp-100c_alt | fuji_fp-100c_cool | fuji_fp-100c_cool_+ |
    fuji_fp-100c_cool_++ | fuji_fp-100c_cool_- | fuji_fp-100c_cool_-- | fuji_fp-100c_negative | fuji_fp-100c_negative_+ | fuji_fp-100c_negative_++ | fuji_fp-100c_negative_+++ |
    fuji_fp-100c_negative_++_alt | fuji_fp-100c_negative_- | fuji_fp-100c_negative_-- | fuji_fp-3000b | fuji_fp-3000b_+ | fuji_fp-3000b_++ | fuji_fp-3000b_+++ | fuji_fp-3000b_- | fuji_fp-3000b_--
    | fuji_fp-3000b_hc | fuji_fp-3000b_negative | fuji_fp-3000b_negative_+ | fuji_fp-3000b_negative_++ | fuji_fp-3000b_negative_+++ | fuji_fp-3000b_negative_- | fuji_fp-3000b_negative_-- |
    fuji_fp-3000b_negative_early | fuji_fp_100c | fuji_hdr | fuji_neopan_1600 | fuji_neopan_1600_+ | fuji_neopan_1600_++ | fuji_neopan_1600_- | fuji_neopan_acros_100 | fuji_provia_100_generic |
    fuji_provia_100f | fuji_provia_400f | fuji_provia_400x | fuji_sensia_100 | fuji_superia_100 | fuji_superia_100_+ | fuji_superia_100_++ | fuji_superia_100_- | fuji_superia_1600 |
    fuji_superia_1600_+ | fuji_superia_1600_++ | fuji_superia_1600_- | fuji_superia_200 | fuji_superia_200_xpro | fuji_superia_400 | fuji_superia_400_+ | fuji_superia_400_++ | fuji_superia_400_-
    | fuji_superia_800 | fuji_superia_800_+ | fuji_superia_800_++ | fuji_superia_800_- | fuji_superia_hg_1600 | fuji_superia_reala_100 | fuji_superia_x-tra_800 | fuji_velvia_100_generic |
    fuji_velvia_50 | fuji_xtrans_iii_acros | fuji_xtrans_iii_acros+g | fuji_xtrans_iii_acros+r | fuji_xtrans_iii_acros+ye | fuji_xtrans_iii_astia | fuji_xtrans_iii_classic_chrome |
    fuji_xtrans_iii_mono | fuji_xtrans_iii_mono+g | fuji_xtrans_iii_mono+r | fuji_xtrans_iii_mono+ye | fuji_xtrans_iii_pro_neg_hi | fuji_xtrans_iii_pro_neg_std | fuji_xtrans_iii_provia |
    fuji_xtrans_iii_sepia | fuji_xtrans_iii_velvia | fusion_88 | futuristicbleak_1 | futuristicbleak_2 | futuristicbleak_3 | futuristicbleak_4 | going_for_a_walk | golden | golden_bright |
    golden_fade | golden_mono | golden_night_softner_43 | golden_sony_37 | golden_vibrant | goldengate | goldentime | goldfx_bright_spring_breeze | goldfx_bright_summer_heat |
    goldfx_hot_summer_heat | goldfx_perfect_sunset_01min | goldfx_perfect_sunset_05min | goldfx_perfect_sunset_10min | goldfx_spring_breeze | goldfx_summer_heat | good_morning | green_15 |
    green_2025 | green_action | green_afternoon | green_and_orange | green_blues | green_conflict | green_day_01 | green_day_02 | green_g_09 | green_indoor | green_light | green_mono |
    green_yellow | greenish_contrasty | greenish_fade | greenish_fade_1 | gremerta | hackmanite | hallowen_dark | happyness_133 | hard_teal_orange | harsh_day | harsh_sunset | helios | herderite
    | heulandite | hiddenite | highlights_protection | hilutite | hitman | hlg_1_1 | honey_light | hong_kong | horrorblue | howlite | hydracore | hyla_68 | hypersthene | hypnosis | hypressen |
    ilford_delta_100 | ilford_delta_3200 | ilford_delta_3200_+ | ilford_delta_3200_++ | ilford_delta_3200_- | ilford_delta_400 | ilford_fp_4_plus_125 | ilford_hp_5 | ilford_hp_5_+ |
    ilford_hp_5_++ | ilford_hp_5_- | ilford_hp_5_plus_400 | ilford_hps_800 | ilford_pan_f_plus_50 | ilford_xp_2 | indoor_blue | industrial_33 | infrared_-_dust_pink | instantc | justpeachy |
    jwick_21 | k_tone_vintage_kodachrome | kh_1 | kh_10 | kh_2 | kh_3 | kh_4 | kh_5 | kh_6 | kh_7 | kh_8 | kh_9 | killstreak | kodak_2383_constlclip | kodak_2383_constlmap | kodak_2383_cuspclip |
    kodak_2393_constlclip | kodak_2393_constlmap | kodak_2393_cuspclip | kodak_bw_400_cn | kodak_e-100_gx_ektachrome_100 | kodak_ektachrome_100_vs | kodak_ektachrome_100_vs_generic |
    kodak_ektar_100 | kodak_elite_100_xpro | kodak_elite_chrome_200 | kodak_elite_chrome_400 | kodak_elite_color_200 | kodak_elite_color_400 | kodak_elite_extracolor_100 | kodak_hie_hs_infra |
    kodak_kodachrome_200 | kodak_kodachrome_25 | kodak_kodachrome_64 | kodak_kodachrome_64_generic | kodak_portra_160 | kodak_portra_160_+ | kodak_portra_160_++ | kodak_portra_160_- |
    kodak_portra_160_nc | kodak_portra_160_nc_+ | kodak_portra_160_nc_++ | kodak_portra_160_nc_- | kodak_portra_160_vc | kodak_portra_160_vc_+ | kodak_portra_160_vc_++ | kodak_portra_160_vc_- |
    kodak_portra_400 | kodak_portra_400_+ | kodak_portra_400_++ | kodak_portra_400_- | kodak_portra_400_nc | kodak_portra_400_nc_+ | kodak_portra_400_nc_++ | kodak_portra_400_nc_- |
    kodak_portra_400_uc | kodak_portra_400_uc_+ | kodak_portra_400_uc_++ | kodak_portra_400_uc_- | kodak_portra_400_vc | kodak_portra_400_vc_+ | kodak_portra_400_vc_++ | kodak_portra_400_vc_- |
    kodak_portra_800 | kodak_portra_800_+ | kodak_portra_800_++ | kodak_portra_800_- | kodak_portra_800_hc | kodak_t-max_100 | kodak_t-max_3200 | kodak_t-max_400 | kodak_tmax_3200 |
    kodak_tmax_3200_+ | kodak_tmax_3200_++ | kodak_tmax_3200_- | kodak_tmax_3200_alt | kodak_tri-x_400 | kodak_tri-x_400_+ | kodak_tri-x_400_++ | kodak_tri-x_400_- | kodak_tri-x_400_alt |
    korben_214 | landscape_01 | landscape_02 | landscape_03 | landscape_04 | landscape_05 | landscape_1 | landscape_10 | landscape_2 | landscape_3 | landscape_4 | landscape_5 | landscape_6 |
    landscape_7 | landscape_8 | landscape_9 | lateafternoonwanderlust | latesunset | lc_1 | lc_10 | lc_2 | lc_3 | lc_4 | lc_5 | lc_6 | lc_7 | lc_8 | lc_9 | lenox_340 | life_giving_tree |
    light_blown | lomo | lomography_redscale_100 | lomography_x-pro_slide_200 | london_nights | louetta | low_contrast_blue | low_key_01 | lucky_64 | lushgreensummer | magenta_day |
    magenta_day_01 | magenta_dream | magenta_yellow | magentacoffee | matrix | mckinnon_75 | memories | metropolis | milo_5 | minimalistcaffeination | modern_film | modern_films_01 |
    modern_films_02 | modern_films_03 | modern_films_04 | modern_films_05 | modern_films_06 | modern_films_07 | mono_tinted | monochrome | monochrome_1 | monochrome_2 | moody_1 | moody_10 |
    moody_2 | moody_3 | moody_4 | moody_5 | moody_6 | moody_7 | moody_8 | moody_9 | moonlight | moonlight_01 | moonrise | morning_6 | morroco_16 | mostly_blue | moviz_1 | moviz_10 | moviz_11 |
    moviz_12 | moviz_13 | moviz_14 | moviz_15 | moviz_16 | moviz_17 | moviz_18 | moviz_19 | moviz_2 | moviz_20 | moviz_21 | moviz_22 | moviz_23 | moviz_24 | moviz_25 | moviz_26 | moviz_27 |
    moviz_28 | moviz_29 | moviz_3 | moviz_30 | moviz_31 | moviz_32 | moviz_33 | moviz_34 | moviz_35 | moviz_36 | moviz_37 | moviz_38 | moviz_39 | moviz_4 | moviz_40 | moviz_41 | moviz_42 |
    moviz_43 | moviz_44 | moviz_45 | moviz_46 | moviz_47 | moviz_48 | moviz_5 | moviz_6 | moviz_7 | moviz_8 | moviz_9 | mute_shift | muted_01 | muted_fade | mysticpurplesunset | nah |
    natural_vivid | nemesis | neon_770 | neutral_pump | neutral_teal_orange | neutral_warm_fade | newspaper | night_01 | night_blade_4 | night_king_141 | night_spy | nightfromday | nightlife |
    nostalgiahoney | nostalgic | nw-1 | nw-10 | nw-2 | nw-3 | nw-4 | nw-5 | nw-6 | nw-7 | nw-8 | nw-9 | old_west | once_upon_a_time | only_red | only_red_and_blue | operation_yellow |
    orange_dark_4 | orange_dark_7 | orange_dark_look | orange_tone | orange_underexposed | oranges | paladin | paladin_1875 | pasadena_21 | passing_by | pink_fade | pitaya_15 | pmcinematic_01 |
    pmcinematic_02 | pmcinematic_03 | pmcinematic_04 | pmcinematic_05 | pmcinematic_06 | pmcinematic_07 | pmnight_01 | pmnight_02 | pmnight_03 | pmnight_04 | pmnight_05 | polaroid_664 |
    polaroid_665 | polaroid_665_+ | polaroid_665_++ | polaroid_665_- | polaroid_665_-- | polaroid_665_negative | polaroid_665_negative_+ | polaroid_665_negative_- | polaroid_665_negative_hc |
    polaroid_667 | polaroid_669 | polaroid_669_+ | polaroid_669_++ | polaroid_669_+++ | polaroid_669_- | polaroid_669_-- | polaroid_669_cold | polaroid_669_cold_+ | polaroid_669_cold_- |
    polaroid_669_cold_-- | polaroid_672 | polaroid_690 | polaroid_690_+ | polaroid_690_++ | polaroid_690_- | polaroid_690_-- | polaroid_690_cold | polaroid_690_cold_+ | polaroid_690_cold_++ |
    polaroid_690_cold_- | polaroid_690_cold_-- | polaroid_690_warm | polaroid_690_warm_+ | polaroid_690_warm_++ | polaroid_690_warm_- | polaroid_690_warm_-- | polaroid_polachrome |
    polaroid_px-100uv+_cold | polaroid_px-100uv+_cold_+ | polaroid_px-100uv+_cold_++ | polaroid_px-100uv+_cold_+++ | polaroid_px-100uv+_cold_- | polaroid_px-100uv+_cold_-- |
    polaroid_px-100uv+_warm | polaroid_px-100uv+_warm_+ | polaroid_px-100uv+_warm_++ | polaroid_px-100uv+_warm_+++ | polaroid_px-100uv+_warm_- | polaroid_px-100uv+_warm_-- | polaroid_px-680 |
    polaroid_px-680_+ | polaroid_px-680_++ | polaroid_px-680_- | polaroid_px-680_-- | polaroid_px-680_cold | polaroid_px-680_cold_+ | polaroid_px-680_cold_++ | polaroid_px-680_cold_++_alt |
    polaroid_px-680_cold_- | polaroid_px-680_cold_-- | polaroid_px-680_warm | polaroid_px-680_warm_+ | polaroid_px-680_warm_++ | polaroid_px-680_warm_- | polaroid_px-680_warm_-- | polaroid_px-70
    | polaroid_px-70_+ | polaroid_px-70_++ | polaroid_px-70_+++ | polaroid_px-70_- | polaroid_px-70_-- | polaroid_px-70_cold | polaroid_px-70_cold_+ | polaroid_px-70_cold_++ |
    polaroid_px-70_cold_- | polaroid_px-70_cold_-- | polaroid_px-70_warm | polaroid_px-70_warm_+ | polaroid_px-70_warm_++ | polaroid_px-70_warm_- | polaroid_px-70_warm_-- |
    polaroid_time_zero_expired | polaroid_time_zero_expired_+ | polaroid_time_zero_expired_++ | polaroid_time_zero_expired_- | polaroid_time_zero_expired_-- | polaroid_time_zero_expired_--- |
    polaroid_time_zero_expired_cold | polaroid_time_zero_expired_cold_- | polaroid_time_zero_expired_cold_-- | polaroid_time_zero_expired_cold_--- | portrait_1 | portrait_10 | portrait_2 |
    portrait_3 | portrait_4 | portrait_5 | portrait_6 | portrait_7 | portrait_8 | portrait_9 | progressen | protect_highlights_01 | prussian_blue | pseudogrey | purple | purple_2 |
    red_afternoon_01 | red_day_01 | red_dream_01 | redblueyellow | reds | reds_oranges_yellows | reeve_38 | remy_24 | rest_33 | retro | retro_brown_01 | retro_magenta_01 | retro_summer_3 |
    retro_yellow_01 | rollei_ir_400 | rollei_ortho_25 | rollei_retro_100_tonal | rollei_retro_80s | rotate_muted | rotate_vibrant | rotated | rotated_crush | saturated_blue | saving_private_damon
    | science_fiction | sea | serenity | seringe_4 | serpent | seventies_magazine | sevsuz | shade_kings_ink | shadow_king_39 | shine | skin_tones | smart_contrast | smokey | smooth_clear |
    smooth_cromeish | smooth_fade | smooth_green_orange | smooth_sailing | smooth_teal_orange | soft_fade | softwarming | solarized_color | solarized_color_2 | springmorning | sprocket_231 |
    spy_29 | street | studio_skin_tone_shaper | subtle_blue | subtle_green | subtle_yellow | summer | summer_alt | sunlightlove | sunny | sunny_alt | sunny_rich | sunny_warm | sunset_aqua_orange
    | sunset_intense_violet_blue | sunset_violet_mood | super_warm | super_warm_rich | sutro_fx | sweet_bubblegum | sweet_gelatto | taiga | tarraco | teal_fade | teal_moonlight | tealmagentagold
    | tealorange | tealorange_1 | tealorange_2 | tealorange_3 | technicalfx_backlight_filter | teigen_28 | tensiongreen_1 | tensiongreen_2 | tensiongreen_3 | tensiongreen_4 | terra_4 |
    the_matrices | thriller_2 | toastedgarden | trent_18 | true_colors_8 | turkiest_42 | tweed_71 | ultra_water | undeniable | undeniable_2 | unknown | urban_01 | urban_02 | urban_03 | urban_04 |
    urban_05 | urban_cowboy | uzbek_bukhara | uzbek_marriage | uzbek_samarcande | velvetia | very_warm_greenish | vfb_21 | vibrant | vibrant_alien | vibrant_contrast | vibrant_cromeish | victory
    | vintage | vintage_01 | vintage_02 | vintage_03 | vintage_04 | vintage_05 | vintage_163 | vintage_alt | vintage_brighter | vintage_chrome | vintage_mob | vintage_warmth_1 | violet_taste |
    vireo_37 | warm | warm_dark_contrasty | warm_fade | warm_fade_1 | warm_highlight | warm_neutral | warm_sunset_red | warm_teal | warm_vintage | warm_yellow | well_see | western | westernlut_2
    | whiter_whites | winterlighthouse | wipe | wooden_gold_20 | yellow_55b | yellow_film_01 | yellowstone | you_can_do_it | zed_32 | zeke_39 | zilverfx_bw_solarization | zilverfx_infrared |
    zilverfx_vintage_bw } : Default values: 'resolution=33' and 'cut_and_round=1'.

   Example:
     [#1] clut summer clut alien_green,17 clut orange_dark4,48

 m (+):
     Shortcut for command 'command'.

 command (+):
     _add_debug_info={ 0 | 1 },{ filename | http[s]://URL | "string" }

   Import G'MIC custom commands from specified file, URL or string.
   (equivalent to shortcut command 'm').

   Imported commands are available directly after the 'command' invocation.

   Default value: 'add_debug_info=1'.

   Example:
     [#1] image.jpg command "foo : mirror y deform $""1" +foo[0] 5 +foo[0] 15

 cursor (+):
     _mode = { 0=hide | 1=show }

   Show or hide mouse cursor for selected instant display windows.
   Command selection (if any) stands for instant display window indices instead of image indices.

   Default value: 'mode=1'.

 delete (+):
     filename1[,filename2,...]

   Delete specified filenames on disk. Multiple filenames must be separated by commas.

 d (+):
     Shortcut for command 'display'.

 display (+):
     _X[%]>=0,_Y[%]>=0,_Z[%]>=0,_exit_on_anykey={ 0 | 1 }

   Display selected images in an interactive viewer (use the instant display window [0] if opened).
   (equivalent to shortcut command 'd').

   Arguments 'X','Y','Z' determine the initial selection view, for 3D volumetric images.

   Default value: 'X=Y=Z=0' and 'exit_on_anykey=0'.

   Tutorial: https://gmic.eu/oldtutorial/_display

 d0:
     Shortcut for command 'display0'.

 display0:

   Display selected images without value normalization.
   (equivalent to shortcut command 'd0').

 d2d:
     Shortcut for command 'display2d'.

 display2d:

   Display selected 2d images in an interactive window.
   (equivalent to shortcut command 'd2d').

   This command is used by default by command 'display' when displaying 2d images.
   If selected image is a volumetric image, each slice is displayed on a separate display
   window (up to 10 images can be displayed simultaneously this way), with synchronized moves.
   When interactive window is opened, the following actions are possible:
    * Left mouse button: Create an image selection and zoom into it.
    * Middle mouse button, or CTRL+left mouse button: Move image.
    * Mouse wheel or PADD+/-: Zoom in/out.
    * Arrow keys: Move image left/right/up/down.
    * 'CTRL + A': Enable/disable transparency (show/hide alpha channel).
    * 'CTRL + C': Decrease window size.
    * 'CTRL + D': Increase window size.
    * 'CTRL + F': Toggle fullscreen mode.
    * 'CTRL + N': Change normalization mode (can be { none | normal | channel-by-channel }).
    * 'CTRL + O': Save a copy of the input image, as a numbered file 'gmic_xxxxxx.gmz'.
    * 'CTRL + R': Reset both window size and view.
    * 'CTRL + S': Save a screenshot of the current view, as a numbered file 'gmic_xxxxxx.png'.
    * 'CTRL + SPACE': Reset view.
    * 'CTRL + X': Show/hide axes.
    * 'CTRL + Z': Hold/release aspect ratio.

 d3d:
     Shortcut for command 'display3d'.

 display3d:
     _[background_image],_exit_on_anykey={ 0 | 1 } |
     _exit_on_anykey={ 0 | 1 }

   Display selected 3D objects in an interactive viewer (use the instant display window [0] if opened).
   (equivalent to shortcut command 'd3d').

   Default values: '[background_image]=(default)' and 'exit_on_anykey=0'.

 da:
     Shortcut for command 'display_array'.

 display_array:
     _width>0,_height>0

   Display images in interactive windows where pixel neighborhoods can be explored.

   Default values: 'width=13' and 'height=width'.

 dc:
     Shortcut for command 'display_camera'.

 display_camera:

   Open camera viewer.
   This command requires features from the OpenCV library (not enabled in G'MIC by default).

 dfft:
     Shortcut for command 'display_fft'.

 display_fft:

   Display fourier transform of selected images, with centered log-module and argument.
   (equivalent to shortcut command 'dfft').

   Example:
     [#1] image.jpg +display_fft

 dg:
     Shortcut for command 'display_graph'.

 display_graph:
     _width>=0,_height>=0,_plot_type,_vertex_type,_xmin,_xmax,_ymin,_ymax,_xlabel,_ylabel

   Render graph plot from selected image data.
   'plot_type' can be { 0=none | 1=lines | 2=splines | 3=bar }.
   'vertex_type' can be { 0=none | 1=points | 2,3=crosses | 4,5=circles | 6,7=squares }.
   'xmin','xmax','ymin','ymax' set the coordinates of the displayed xy-axes.
   if specified 'width' or 'height' is '0', then image size is set to half the screen size.

   Default values: 'width=0', 'height=0', 'plot_type=1', 'vertex_type=1', 'xmin=xmax=ymin=ymax=0 (auto)', 'xlabel="x-axis"' and 'ylabel="y-axis"'.

   Example:
     [#1] 128,1,1,1,'cos(x/10+u)' +display_graph 400,300,3

 dh:
     Shortcut for command 'display_histogram'.

 display_histogram:
     _width>=0,_height>=0,_clusters>0,_min_value[%],_max_value[%],_show_axes={ 0 | 1 },_expression.

   Render a channel-by-channel histogram.
   If selected images have several slices, the rendering is performed for all input slices.
   'expression' is a mathematical expression used to transform the histogram data for visualization purpose.
   (equivalent to shortcut command 'dh').

   if specified 'width' or 'height' is '0', then image size is set to half the screen size.

   Default values: 'width=0', 'height=0', 'clusters=256', 'min_value=0%', 'max_value=100%', 'show_axes=1' and 'expression=i'.

   Example:
     [#1] image.jpg +display_histogram 512,300

 display_parametric:
     _width>0,_height>0,_outline_opacity,_vertex_radius>=0,_is_antialiased={ 0 | 1 },_is_decorated={ 0 | 1 },_xlabel,_ylabel

   Render 2D or 3D parametric curve or point clouds from selected image data.
   Curve points are defined as pixels of a 2 or 3-channel image.
   If the point image contains more than 3 channels, additional channels define the (R,G,B) color for each vertex.
   If 'outline_opacity>1', the outline is colored according to the specified vertex colors and
   'outline_opacity-1' is used as the actual drawing opacity.

   Default values: 'width=512', 'height=width', 'outline_opacity=3', 'vertex_radius=0', 'is_antialiased=1','is_decorated=1', 'xlabel="x-axis"' and
    'ylabel="y-axis"'.

   Example:
     [#1] 1024,1,1,2,'t=x/40;if(c==0,sin(t),cos(t))*(exp(cos(t))-2*cos(4*t)-sin(t/12)^5)' display_parametric 512,512
     [#2] 1000,1,1,2,u(-100,100) quantize 4,1 noise 12 channels 0,2 +normalize 0,255 append c display_parametric 512,512,0.1,8

 dp:
     Shortcut for command 'display_parallel'.

 display_parallel:

   Display each selected image in a separate interactive display window.
   (equivalent to shortcut command 'dp').

 dp0:
     Shortcut for command 'display_parallel0'.

 display_parallel0:

   Display each selected image in a separate interactive display window, without value normalization.
   (equivalent to shortcut command 'dp0').

 display_polar:
     _width>32,_height>32,_outline_type,_fill_R,_fill_G,_fill_B,_theta_start,_theta_end,_xlabel,_ylabel

   Render polar curve from selected image data.
   'outline_type' can be { r<0=dots with radius -r | 0=no outline | r>0=lines+dots with radius r }.
   'fill_color' can be { -1=no fill | R,G,B=fill with specified color }.

   Default values: 'width=500', 'height=width', 'outline_type=1', 'fill_R=fill_G=fill_B=200', 'theta_start=0', 'theta_end=360', 'xlabel="x-axis"' and
    'ylabel="y-axis"'.

   Example:
     [#1] 300,1,1,1,'0.3+abs(cos(10*pi*x/w))+u(0.4)' display_polar 512,512,4,200,255,200
     [#2] 3000,1,1,1,'x^3/1e10' display_polar 400,400,1,-1,,,0,{15*360}

 dq:
     Shortcut for command 'display_quiver'.

 display_quiver:
     _size_factor>0,_arrow_size>=0,_color_mode={ 0=monochrome | 1=grayscale | 2=color }

   Render selected images of 2D vectors as a field of 2D arrows.
   (equivalent to shortcut command 'dq').

   Default values: 'size_factor=16', 'arrow_size=1.5' and 'color_mode=1'.

   Example:
     [#1] image.jpg +luminance gradient[-1] xy rv[-2,-1] *[-2] -1 a[-2,-1] c crop 60,10,90,30 +display_quiver[1] ,

 drgba:
     Shortcut for command 'display_rgba'.

 display_rgba:
     _background_RGB_color

   Render selected RGBA images over a checkerboard or colored background.
   (equivalent to shortcut command 'drgba').

   Default values: 'background_RGB_color=undefined' (checkerboard).

   Example:
     [#1] image.jpg +norm threshold[-1] 40% blur[-1] 3 normalize[-1] 0,255 append c display_rgba

 dt:
     Shortcut for command 'display_tensors'.

 display_tensors:
     _size_factor>0,_ellipse_size>=0,_color_mode={ 0=monochrome | 1=grayscale | 2=color },_outline>=0

   Render selected images of tensors as a field of 2D ellipses.
   (equivalent to shortcut command 'dt').

   Default values: 'size_factor=16', 'ellipse_size=1.5', 'color_mode=2' and 'outline=2'.

   Example:
     [#1] image.jpg +diffusiontensors 0.1,0.9 resize2dx. 32 +display_tensors. 64,2

   Tutorial: https://gmic.eu/oldtutorial/_display_tensors

 dw:
     Shortcut for command 'display_warp'.

 display_warp:
     _cell_size>0

   Render selected 2D warping fields.
   (equivalent to shortcut command 'dw').

   Default value: 'cell_size=15'.

   Example:
     [#1] 400,400,1,2,'x=x-w/2;y=y-h/2;r=sqrt(x*x+y*y);a=atan2(y,x);5*sin(r/10)*[cos(a),sin(a)]' +display_warp 10

 e (+):
     Shortcut for command 'echo'.

 echo (+):
     message

   Output specified message on the error output.
   (equivalent to shortcut command 'e').

   Command selection (if any) stands for displayed call stack subset instead of image indices.
   When invoked with a '+' prefix (i.e. '+echo'), the command output its message on stdout rather than stderr.

 echo_file:
     filename,message

   Output specified message, appending it to specified output file.
   (similar to 'echo' for specified output file stream).

 function1d:
     0<=smoothness<=1,x0>=0,y0,x1>=0,y1,...,xn>=0,yn

   Insert continuous 1D function from specified list of keypoints (xk,yk)
   in range [0,max(xk)] (xk are positive integers).

   Default values: 'smoothness=1' and 'x0=y0=0'.

   Example:
     [#1] function1d 1,0,0,10,30,40,20,70,30,80,0 +display_graph 400,300

 i (+):
     Shortcut for command 'input'.

 input (+):
     [type:]filename |
     [type:]http://URL |
     [selection]x_nb_copies>0 |
     { width>0[%] | [image_w] },{ _height>0[%] | [image_h] },{ _depth>0[%] | [image_d] },{ _spectrum>0[%] | [image_s] },_{ value1,_value2,... | 'formula' } |
     (value1{,|;|/|^}value2{,|;|/|^}...[:{x|y|z|c|,|;|/|^}]) |
     0

   Insert a new image taken from a filename or from a copy of an existing image [index],
   or insert new image with specified dimensions and values. Single quotes may be omitted in
   'formula'. Specifying argument '0' inserts an 'empty' image.
   (equivalent to shortcut command 'i').

   Default values: 'nb_copies=1', 'height=depth=spectrum=1' and 'value1=0'.

   Example:
     [#1] input image.jpg
     [#2] input (1,2,3;4,5,6;7,8,9^9,8,7;6,5,4;3,2,1)
     [#3] image.jpg (1,2,3;4,5,6;7,8,9) (255^128^64) 400,400,1,3,'if(x>w/2,x,y)*c'

   Tutorial: https://gmic.eu/tutorial/input

 input_565:
     filename,width>0,height>0,reverse_endianness={ 0 | 1 }

   Insert image data from a raw RGB-565 file, at the end of the list.

   Default value: 'reverse_endianness=0'.

 input_csv:
     "filename",_read_data_as={ 0=numbers | 1=strings | _variable_name }

   Insert number of string array from specified .csv file.
   If 'variable_name' is provided, the string of each cell is stored in a numbered variable '_variable_name_x_y', where 'x' and 'y' are the indices of the cell column and row
   respectively (starting from '0').
   Otherwise, a 'WxH' image is inserted at the end of the list, with each vector-valued pixel 'I(x,y)' encoding the number or the string of each cell.
   This command returns the 'W,H' dimension of the read array, as the status.

   Default value: 'read_data_as=1'.

 input_cube:
     "filename",_convert_1d_cluts_to_3d={ 0 | 1 }.

   Insert CLUT data from a .cube filename (Adobe CLUT file format).

   Default value: 'convert_1d_cluts_to_3d=1'.

 input_flo:
     "filename"

   Insert optical flow data from a .flo filename (vision.middlebury.edu file format).

 ig:
     Shortcut for command 'input_glob'.

 input_glob:
     pattern

   Insert new images from several filenames that match the specified glob pattern.
   (equivalent to shortcut command 'ig').

 input_gpl:
     filename

   Input specified filename as a .gpl palette data file.

 it:
     Shortcut for command 'input_text'.

 input_text:
     filename

   Input specified text-data filename as a new image.
   (equivalent to shortcut command 'it').

 network (+):
     mode={ -1=disabled | 0=enabled w/o timeout | >0=enabled w/ specified timeout in seconds }

   Enable/disable load-from-network and set corresponding timeout.
   (Default mode is 'enabled w/o timeout').

 o (+):
     Shortcut for command 'output'.

 output (+):
     [type:]filename,_format_options

   Output selected images as one or several numbered file(s).
   (equivalent to shortcut command 'o').

   Default value: 'format_options'=(undefined).

 output_565:
     "filename",reverse_endianness={ 0=false | 1=true }

   Output selected images as raw RGB-565 files.

   Default value: 'reverse_endianness=0'.

 output_cube:
     "filename"

   Output selected CLUTs as a .cube file (Adobe CLUT format).

 output_flo:
     "filename"

   Output selected optical flow as a .flo file (vision.middlebury.edu file format).

 output_ggr:
     filename,_gradient_name

   Output selected images as .ggr gradient files (GIMP).
   If no gradient name is specified, it is deduced from the filename.

 output_gmz:
     filename,_datatype

   Output selected images as .gmz files (G'MIC native file format).
   'datatype' can be { bool | uchar | char | ushort | short | uint | int | uint64 | int64 | float | double }.

 output_obj:
     filename,_save_materials={ 0=no | 1=yes }

   Output selected 3D objects as Wavefront 3D object files.
   Set 'save_materials' to '1' to produce a corresponding material file ('.mtl') and eventually texture files.
   Beware, the export to '.obj' files may be quite slow for large 3D objects.

   Default value: 'save_materials=1'.

 ot:
     Shortcut for command 'output_text'.

 output_text:
     filename

   Output selected images as text-data filenames.
   (equivalent to shortcut command 'ot').

 on:
     Shortcut for command 'outputn'.

 outputn:
     filename,_index

   Output selected images as automatically numbered filenames in repeat...done loops.
   (equivalent to shortcut command 'on').

 op:
     Shortcut for command 'outputp'.

 outputp:
     prefix

   Output selected images as prefixed versions of their original filenames.
   (equivalent to shortcut command 'op').

   Default value: 'prefix=_'.

 ow:
     Shortcut for command 'outputw'.

 outputw:

   Output selected images by overwriting their original location.
   (equivalent to shortcut command 'ow').

 ox:
     Shortcut for command 'outputx'.

 outputx:
     extension1,_extension2,_...,_extensionN,_output_at_same_location={ 0 | 1 }

   Output selected images with same base filenames but for N different extensions.
   (equivalent to shortcut command 'ox').

   Default value: 'output_at_same_location=0'.

 parse_cli:
     _output_mode,_{ * | command_name }

   Parse definition of ''-documented commands and output info about them in specified output mode.
   'output_mode' can be { ascii | bashcompletion | html | images | print }.

   Default values: 'output_mode=print' and 'command_name=*'.

 parse_gmd:

   Parse and tokenize selected images, viewed as text strings formatted with the G'MIC markdown syntax.

 gmd2html:
     _include_default_header_footer={ 0=none | 1=Reference | 2=Tutorial | 3=News } |
     (no arg)

   Convert selected gmd-formatted text images to html format.

   Default values: 'include_default_header_footer=1'.

 gmd2ascii:
     _max_line_length>0,_indent_forced_newlines>=0 |
     (no arg)

   Convert selected gmd-formatted text images to ascii format.

   Default values: 'max_line_length=80' and 'indent_forced_newline=0'.

 parse_gui:
     _outputmode,_{ * | filter_name}

   Parse selected filter definitions and generate info about filters in selected output mode.
   'outputmode' can be { json | list | print | strings | update | zart }.
   It is possible to define a custom output mode, by implementing the following commands
   ('outputmode' must be replaced by the name of the custom user output mode):
   . 'parse_gui_outputmode' : A command that outputs the parsing information with a custom format.
   . 'parse_gui_parseparams_outputmode' (optional): A simple command that returns 0 or 1. It tells the parser whether parameters of matching filter must be analyzed (slower) or not.
   . 'parse_gui_trigger_outputmode' (optional): A command that is called by the parser just before parsing the set of each matching filters.
   Here is the list of global variables set by the parser, accessible in command 'parse_gui_outputmode':
   '$_nbfilters': Number of matching filters.
   '$_nongui' (stored as an image): All merged lines in the file that do not correspond to '#@gui' lines.
   For each filter '     * '$_fF_name' : Filter name.
    * '$_fF_path' : Full path.
    * '$_fF_locale' : Filter locale (empty, if not specified).
    * '$_fF_command' : Filter command.
    * '$_fF_commandpreview' : Filter preview command (empty, if not specified).
    * '$_fF_zoomfactor' : Default zoom factor (empty, if not specified).
    * '$_fF_zoomaccurate' : Is preview accurate when zoom changes ? (can be { 0=false | 1=true }).
    * '$_fF_inputmode' : Default preferred input mode (empty, if not specified).
    * '$_fF_hide' : Path of filter hid by current filter (for localized filters, empty if not specified).
    * '$_fF_nbparams' : Number of parameters.
   For each parameter '     * '$_fF_pP_name' : Parameter name.
    * '$_fF_pP_type' : Parameter type.
    * '$_fF_pP_responsivity' : Parameter responsivity (can be { 0 | 1 }).
    * '$_fF_pP_visibility' : Parameter visibility.
    * '$_fF_pP_propagation' : Propagation of the parameter visibility.
    * '$_fF_pP_nbargs' : Number of parameter arguments.
   For each argument '     * '$_fF_pP_aA' : Argument value
   Default parameters: 'filter_name=*' and 'output_format=print'.

 pass (+):
     _shared_state={ -1=status only | 0=non-shared (copy) | 1=shared | 2=adaptive }

   Insert images from parent context of a custom command or a local environment.
   Command selection (if any) stands for a selection of images in the parent context.
   By default (adaptive shared state), selected images are inserted in a shared state if they do not belong
   to the context (selection) of the current custom command or local environment as well.
   Typical use of command 'pass' concerns the design of custom commands that take images as arguments.
   This commands return the list of corresponding indices in the status.

   Default value: 'shared_state=2'.

   Example:
     [#1] command "average : pass$""1 add[^-1] [-1] remove[-1] div 2" sample ? +mirror y +average[0] [1]

 plot (+):
     _plot_type,_vertex_type,_xmin,_xmax,_ymin,_ymax,_exit_on_anykey={ 0 | 1 } |
     'formula',_resolution>=0,_plot_type,_vertex_type,_xmin,xmax,_ymin,_ymax,_exit_on_anykey={ 0 | 1 }

   Display selected images or formula in an interactive viewer (use the instant display window [0] if opened).
   'plot_type' can be { 0=none | 1=lines | 2=splines | 3=bar }.
   'vertex_type' can be { 0=none | 1=points | 2,3=crosses | 4,5=circles | 6,7=squares }.
   'xmin', 'xmax', 'ymin', 'ymax' set the coordinates of the displayed xy-axes.

   Default values: 'plot_type=1', 'vertex_type=1', 'xmin=xmax=ymin=ymax=0 (auto)' and 'exit_on_anykey=0'.

 p (+):
     Shortcut for command 'print'.

 print (+):

   Output information on selected images, on the standard error (stderr).
   (equivalent to shortcut command 'p').

   When invoked with a '+' prefix (i.e. '+print'), the command output its message on stdout rather than stderr.

 random_pattern:
     _width>0,_height>0,_min_detail_level>=0

   Insert a new RGB image of specified size at the end of the image list, rendered with a random pattern.

   Default values: 'width=height=512' and 'min_detail_level=2'.

   Example:
     [#1] repeat 6 random_pattern 256 done

 screen (+):
     _x0[%],_y0[%],_x1[%],_y1[%]

   Take screenshot, optionally grabbed with specified coordinates, and insert it
   at the end of the image list.

 select (+):
     feature_type,_X[%]>=0,_Y[%]>=0,_Z[%]>=0,_exit_on_anykey={ 0 | 1 },_is_deep_selection={ 0 | 1 }

   Interactively select a feature from selected images (use the instant display window [0] if opened).
   'feature_type' can be { 0=point | 1=segment | 2=rectangle | 3=ellipse }.
   Arguments 'X','Y','Z' determine the initial selection view, for 3D volumetric images.
   The retrieved feature is returned as a 3D vector (if 'feature_type==0') or as a 6d vector
   (if 'feature_type!=0') containing the feature coordinates.

   Default values: 'X=Y=Z=(undefined)', 'exit_on_anykey=0' and 'is_deep_selection=0'.

 serialize (+):
     _datatype,_is_compressed={ 0 | 1 },_store_names={ 0 | 1 }

   Serialize selected list of images into a single image, optionnally in a compressed form.
   'datatype' can be { auto | uchar | char | ushort | short | uint | int | uint64 | int64 | float | double }.
   Specify 'datatype' if all selected images have a range of values constrained to a particular datatype,
   in order to minimize the memory footprint.
   The resulting image has only integers values in [0,255] and can then be saved as a raw image of
   unsigned chars (doing so will output a valid .cimg[z] or .gmz file).
   If 'store_names' is set to '1', serialization uses the .gmz format to store data in memory
   (otherwise the .cimg[z] format).

   Default values: 'datatype=auto', 'is_compressed=1' and 'store_names=1'.

   Example:
     [#1] image.jpg +serialize uchar +unserialize[-1]

 shape_circle:
     _size>=0

   Input a 2D circle binary shape with specified size.

   Default value: 'size=512'.

   Example:
     [#1] shape_circle ,

 shape_cupid:
     _size>=0

   Input a 2D cupid binary shape with specified size.

   Default value: 'size=512'.

   Example:
     [#1] shape_cupid ,

 shape_diamond:
     _size>=0

   Input a 2D diamond binary shape with specified size.

   Default value: 'size=512'.

   Example:
     [#1] shape_diamond ,

 shape_dragon:
     _size>=0,_recursion_level>=0,_angle

   Input a 2D Dragon curve with specified size.

   Default value: 'size=512', 'recursion_level=18' and 'angle=0'.

   Example:
     [#1] shape_dragon ,

 shape_fern:
     _size>=0,_density[%]>=0,_angle,0<=_opacity<=1,_type={ 0=Asplenium adiantum-nigrum | 1=Thelypteridaceae }

   Input a 2D Barnsley fern with specified size.

   Default value: 'size=512', 'density=50%', 'angle=30', 'opacity=0.3' and 'type=0'.

   Example:
     [#1] shape_fern ,

 shape_gear:
     _size>=0,_nb_teeth>0,0<=_height_teeth<=100,0<=_offset_teeth<=100,0<=_inner_radius<=100

   Input a 2D gear binary shape with specified size.

   Default value: 'size=512', 'nb_teeth=12', 'height_teeth=20', 'offset_teeth=0' and 'inner_radius=40'.

   Example:
     [#1] shape_gear ,

 shape_heart:
     _size>=0

   Input a 2D heart binary shape with specified size.

   Default value: 'size=512'.

   Example:
     [#1] shape_heart ,

 shape_polygon:
     _size>=0,_nb_vertices>=3,_angle

   Input a 2D polygonal binary shape with specified geometry.

   Default value: 'size=512', 'nb_vertices=5' and 'angle=0'.

   Example:
     [#1] repeat 6 shape_polygon 256,{3+$>} done

 shape_snowflake:
     size>=0,0<=_nb_recursions<=6

   Input a 2D snowflake binary shape with specified size.

   Default values: 'size=512' and 'nb_recursions=5'.

   Example:
     [#1] repeat 6 shape_snowflake 256,$> done

 shape_star:
     _size>=0,_nb_branches>0,0<=_thickness<=1

   Input a 2D star binary shape with specified size.

   Default values: 'size=512', 'nb_branches=5' and 'thickness=0.38'.

   Example:
     [#1] repeat 9 shape_star 256,{$>+2} done

 sh (+):
     Shortcut for command 'shared'.

 shared (+):
     x0[%],x1[%],y[%],z[%],c[%] |
     y0[%],y1[%],z[%],c[%] |
     z0[%],z1[%],c[%] |
     c0[%],c1[%] |
     c0[%] |
     (no arg)

   Insert shared buffers from (opt. points/rows/planes/channels of) selected images.
   Shared buffers cannot be returned by a command, nor a local environment.
   (equivalent to shortcut command 'sh').

   Example:
     [#1] image.jpg shared 1 blur[-1] 3 remove[-1]
     [#2] image.jpg repeat s shared 25%,75%,0,$> mirror[-1] x remove[-1] done

   Tutorial: https://gmic.eu/oldtutorial/_shared

 sp:
     Shortcut for command 'sample'.

 sample:
     _name1={ ? | apples | balloons | barbara | boats | bottles | butterfly | cameraman | car | cat | cliff | chick | colorful | david | dog | duck | eagle | elephant | earth | flower | fruits |
       gmicky | gmicky_mahvin | gmicky_wilber | greece | gummy | house | inside | landscape | leaf | lena | leno | lion | mandrill | monalisa | monkey | parrots | pencils | peppers | portrait0 |
       portrait1 | portrait2 | portrait3 | portrait4 | portrait5 | portrait6 | portrait7 | portrait8 | portrait9 | roddy | rooster | rose | square | swan | teddy | tiger | tulips | wall |
       waterfall | zelda },_name2,...,_nameN,_width={ >=0 | 0 (auto) },_height = { >=0 | 0 (auto) } |
     (no arg)

   Input a new sample RGB image (opt. with specified size).
   (equivalent to shortcut command 'sp').

   Argument 'name' can be replaced by an integer which serves as a sample index.

   Example:
     [#1] repeat 6 sample done

 srand (+):
     value |
     (no arg)

   Set random generator seed.
   If no argument is specified, a random value is used as the random generator seed.

 store (+):
     _is_compressed={ 0 | 1 },variable_name1,_variable_name2,...

   Store selected images into one or several named variables.
   Selected images are transferred to the variables, and are so removed from the image list.
   (except if the prepended variant of the command '+store[selection]' is used).
   If a single variable name is specified, all images of the selection are assigned
   to the named variable. Otherwise, there must be as many variable names as images
   in the selection, and each selected image is assigned to each specified named variable.
   Use command 'input $variable_name' to bring the stored images back in the list.

   Default value: 'is_compressed=0'.

   Example:
     [#1] sample eagle,earth store img1,img2 input $img2 $img1

   Tutorial: https://gmic.eu/tutorial/store

 testimage2d:
     _width>0,_height>0,_spectrum>0

   Input a 2D synthetic image.

   Default values: 'width=512', 'height=width' and 'spectrum=3'.

   Example:
     [#1] testimage2d 512

 um:
     Shortcut for command 'uncommand'.

 uncommand (+):
     command_name[,_command_name2,...] |
     *

   Discard definition of specified custom commands.
   Set argument to '*' for discarding all existing custom commands.
   (equivalent to shortcut command 'um').

 uniform_distribution:
     nb_levels>=1,spectrum>=1

   Input set of uniformly distributed spectrum-d points in [0,1]^spectrum.

   Example:
     [#1] uniform_distribution 64,3 * 255 +distribution3d circles3d[-1] 10

 unserialize (+):

   Recreate lists of images from serialized image buffers, obtained with command 'serialize'.

 up:
     Shortcut for command 'update'.

 update:

   Update commands from the latest definition file on the G'MIC server.
   (equivalent to shortcut command 'up').

 v (+):
     Shortcut for command 'verbose'.

 verbose (+):
     level |
     { + | - }

   Set or increment/decrement the verbosity level. Default level is 0.
   (equivalent to shortcut command 'v').

   When 'level'>0, G'MIC log messages are displayed on the standard error (stderr).

   Default value: 'level=1'.

 wait (+):
     delay |
     (no arg)

   Wait for a given delay (in ms), optionally since the last call to 'wait'.
   or wait for a user event occurring on the selected instant display windows.
   'delay' can be { <0=delay+flush events | 0=event | >0=delay }.
   Command selection (if any) stands for instant display window indices instead of image indices.
   If no window indices are specified and if 'delay' is positive, the command results
   in a 'hard' sleep during specified delay.

   Default value: 'delay=0'.

 warn (+):
     _force_visible={ 0 | 1 },_message

   Print specified warning message, on the standard error (stderr).
   Command selection (if any) stands for displayed call stack subset instead of image indices.

 w (+):
     Shortcut for command 'window'.

 window (+):
     _width[%]>=-1,_height[%]>=-1,_normalization,_fullscreen,_pos_x[%],_pos_y[%],_title

   Display selected images into an instant display window with specified size, normalization type,
   fullscreen mode and title.
   (equivalent to shortcut command 'w').

   If 'width' or 'height' is set to -1, the corresponding dimension is adjusted to the window
   or image size.
   Specify 'pos_x' and 'pos_y' arguments only if the window has to be moved to the specified
   coordinates. Otherwise, they can be avoided.
   'width'=0 or 'height'=0 closes the instant display window.
   'normalization' can be { -1=keep same | 0=none | 1=always | 2=1st-time | 3=auto }.
   'fullscreen' can be { -1=keep same | 0=no | 1=yes }.
   You can manage up to 10 different instant display windows by using the numbered variants
   'w0' (default, eq. to 'w'),'w1',...,'w9' of the command 'w'.
   Invoke 'window' with no selection to make the window visible, if it has been closed by the user.

   Default values: 'width=height=normalization=fullscreen=-1' and 'title=(undefined)'.

 12.3. List Manipulation
       -----------------

 k (+):
     Shortcut for command 'keep'.

 keep (+):

   Keep only selected images.
   (equivalent to shortcut command 'k').

   Example:
     [#1] image.jpg split x keep[0-50%:2] append x
     [#2] image.jpg split x keep[^30%-70%] append x

 mv (+):
     Shortcut for command 'move'.

 move (+):
     position[%]

   Move selected images at specified position.
   (equivalent to shortcut command 'mv').

   Example:
     [#1] image.jpg split x,3 move[1] 0
     [#2] image.jpg split x move[50%--1:2] 0 append x

 nm (+):
     Shortcut for command 'name'.

 name (+):
     "name1","name2",...

   Set names of selected images.
    * If the selection contains a single image, then it is assumed the command has a single name
    argument (possibly containing multiple comas).
    * If the selection contains more than one image, each command argument defines a single image
    name for each image of the selection.
   (equivalent to shortcut command 'nm').

   Example:
     [#1] image.jpg name image blur[image] 2

   Tutorial: https://gmic.eu/tutorial/name

 rm (+):
     Shortcut for command 'remove'.

 remove (+):

   Remove selected images.
   (equivalent to shortcut command 'rm').

   Example:
     [#1] image.jpg split x remove[30%-70%] append x
     [#2] image.jpg split x remove[0-50%:2] append x

 remove_duplicates:

   Remove duplicates images in the selected images list.

   Example:
     [#1] (1,2,3,4,2,4,3,1,3,4,2,1) split x remove_duplicates append x

 remove_empty:

   Remove empty images in the selected image list.

 rmn:
     Shortcut for command 'remove_named'.

 remove_named:
     "name1","name2",...

   Remove all images with specified names from the list of images.
   Does nothing if no images with those names exist.
   (equivalent to shortcut command 'rmn').

 rv (+):
     Shortcut for command 'reverse'.

 reverse (+):

   Reverse positions of selected images.
   (equivalent to shortcut command 'rv').

   Example:
     [#1] image.jpg split x,3 reverse[-2,-1]
     [#2] image.jpg split x,-16 reverse[50%-100%] append x

 sort_list:
     _ordering={ + | - },_criterion

   Sort list of selected images according to the specified image criterion.

   Default values: 'ordering=+', 'criterion=i'.

   Example:
     [#1] (1;4;7;3;9;2;4;7;6;3;9;1;0;3;3;2) split y sort_list +,i append y

 12.4. Mathematical Operators
       ----------------------

 abs (+):

   Compute the pointwise absolute values of selected images.

   Example:
     [#1] image.jpg +sub {ia} abs[-1]
     [#2] 300,1,1,1,'cos(20*x/w)' +abs display_graph 400,300

 acos (+):

   Compute the pointwise arccosine of selected images.

   Example:
     [#1] image.jpg +normalize -1,1 acos[-1]
     [#2] 300,1,1,1,'cut(x/w+0.1*u,0,1)' +acos display_graph 400,300

   Tutorial: https://gmic.eu/oldtutorial/trigometric-and-inverse-trigometric-commands

 acosh (+):

   Compute the pointwise hyperbolic arccosine of selected images.

 + (+):
     Shortcut for command 'add'.

 add (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Add specified value, image or mathematical expression to selected images, or compute the pointwise sum of selected images.
   (equivalent to shortcut command '+').

   Example:
     [#1] image.jpg +add 30% cut 0,255
     [#2] image.jpg +blur 5 normalize 0,255 add[1] [0]
     [#3] image.jpg add '80*cos(80*(x/w-0.5)*(y/w-0.5)+c)' cut 0,255
     [#4] image.jpg repeat 9 +rotate[0] {$>*36},1,0,50%,50% done add div 10

 & (+):
     Shortcut for command 'and'.

 and (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the bitwise AND of selected images with specified value, image or mathematical expression, or compute the pointwise sequential bitwise AND of selected images.
   (equivalent to shortcut command '&').

   Example:
     [#1] image.jpg and {128+64}
     [#2] image.jpg +mirror x and

 argmax:

   Compute the argmax of selected images. Returns a single image
   with each pixel value being the index of the input image with maximal value.

   Example:
     [#1] image.jpg sample lena,lion,square +argmax

 argmaxabs:

   Compute the argmaxabs of selected images. Returns a single image
   with each pixel value being the index of the input image with maxabs value.

 argmin:

   Compute the argmin of selected images. Returns a single image
   with each pixel value being the index of the input image with minimal value.

   Example:
     [#1] image.jpg sample lena,lion,square +argmin

 argminabs:

   Compute the argminabs of selected images. Returns a single image
   with each pixel value being the index of the input image with minabs value.

 asin (+):

   Compute the pointwise arcsine of selected images.

   Example:
     [#1] image.jpg +normalize -1,1 asin[-1]
     [#2] 300,1,1,1,'cut(x/w+0.1*u,0,1)' +asin display_graph 400,300

   Tutorial: https://gmic.eu/oldtutorial/trigometric-and-inverse-trigometric-commands

 asinh (+):

   Compute the pointwise hyperbolic arcsine of selected images.

 atan (+):

   Compute the pointwise arctangent of selected images.

   Example:
     [#1] image.jpg +normalize 0,8 atan[-1]
     [#2] 300,1,1,1,'4*x/w+u' +atan display_graph 400,300

   Tutorial: https://gmic.eu/oldtutorial/trigometric-and-inverse-trigometric-commands

 atan2 (+):
     [x_argument]

   Compute the pointwise oriented arctangent of selected images.
   Each selected image is regarded as the y-argument of the arctangent function, while the
   specified image gives the corresponding x-argument.

   Example:
     [#1] (-1,1) (-1;1) resize 400,400,1,1,3 atan2[1] [0] keep[1] mod {pi/8}

   Tutorial: https://gmic.eu/oldtutorial/trigometric-and-inverse-trigometric-commands

 atanh (+):

   Compute the pointwise hyperbolic arctangent of selected images.

 << (+):
     Shortcut for command 'bsl'.

 bsl (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the bitwise left shift of selected images with specified value, image or mathematical expression, or compute the pointwise sequential bitwise left shift of selected images.
   (equivalent to shortcut command '<<').

   Example:
     [#1] image.jpg bsl 'round(3*x/w,0)' cut 0,255

 >> (+):
     Shortcut for command 'bsr'.

 bsr (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the bitwise right shift of selected images with specified value, image or mathematical expression, or compute the pointwise sequential bitwise right shift of selected images.
   (equivalent to shortcut command '>>').

   Example:
     [#1] image.jpg bsr 'round(3*x/w,0)' cut 0,255

 cos (+):

   Compute the pointwise cosine of selected images.

   Example:
     [#1] image.jpg +normalize 0,{2*pi} cos[-1]
     [#2] 300,1,1,1,'20*x/w+u' +cos display_graph 400,300

   Tutorial: https://gmic.eu/oldtutorial/trigometric-and-inverse-trigometric-commands

 cosh (+):

   Compute the pointwise hyperbolic cosine of selected images.

   Example:
     [#1] image.jpg +normalize -3,3 cosh[-1]
     [#2] 300,1,1,1,'4*x/w+u' +cosh display_graph 400,300

 deg2rad:

   Convert pointwise angle values of selected images, from degrees to radians (apply 'i*pi/180').

 / (+):
     Shortcut for command 'div'.

 div (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Divide selected images by specified value, image or mathematical expression, or compute the pointwise quotient of selected images.
   (equivalent to shortcut command '/').

   Example:
     [#1] image.jpg div '1+abs(cos(x/10)*sin(y/10))'
     [#2] image.jpg +norm add[-1] 1 +div

 div_complex:
     [divider_real,divider_imag],_epsilon>=0

   Perform division of the selected complex pairs (real1,imag1,...,realN,imagN) of images by
   specified complex pair of images (divider_real,divider_imag).
   In complex pairs, the real image must be always located before the imaginary image in the image list.

   Default value: 'epsilon=1e-8'.

 == (+):
     Shortcut for command 'eq'.

 eq (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the boolean equality of selected images with specified value, image or mathematical expression, or compute the boolean equality of selected images.
   (equivalent to shortcut command '==').

   Example:
     [#1] image.jpg round 40 eq {round(ia,40)}
     [#2] image.jpg +mirror x eq

 erf (+):

   Compute the pointwise error function of selected images.

   Example:
     [#1] image.jpg +normalize 0,2 erf[-1]
     [#2] 300,1,1,1,'7*x/w-3.5+u' +erf display_graph 400,300

 exp (+):

   Compute the pointwise exponential of selected images.

   Example:
     [#1] image.jpg +normalize 0,2 exp[-1]
     [#2] 300,1,1,1,'7*x/w+u' +exp display_graph 400,300

 >= (+):
     Shortcut for command 'ge'.

 ge (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the boolean 'greater or equal than' of selected images with specified value, image
   or mathematical expression, or compute the boolean 'greater or equal than' of selected images.
   (equivalent to shortcut command '>=').

   Example:
     [#1] image.jpg ge {ia}
     [#2] image.jpg +mirror x ge

 > (+):
     Shortcut for command 'gt'.

 gt (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the boolean 'greater than' of selected images with specified value, image or mathematical expression, or compute the boolean 'greater than' of selected images.
   (equivalent to shortcut command '>').

   Example:
     [#1] image.jpg gt {ia}
     [#2] image.jpg +mirror x gt

 <= (+):
     Shortcut for command 'le'.

 le (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the boolean 'less or equal than' of selected images with specified value, image or mathematical expression, or compute the boolean 'less or equal than' of selected images.
   (equivalent to shortcut command '<=').

   Example:
     [#1] image.jpg le {ia}
     [#2] image.jpg +mirror x le

 < (+):
     Shortcut for command 'lt'.

 lt (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the boolean 'less than' of selected images with specified value, image or mathematical expression, or compute the boolean 'less than' of selected images.
   (equivalent to shortcut command '<').

   Example:
     [#1] image.jpg lt {ia}
     [#2] image.jpg +mirror x lt

 log (+):

   Compute the pointwise base-e logarithm of selected images.

   Example:
     [#1] image.jpg +add 1 log[-1]
     [#2] 300,1,1,1,'7*x/w+u' +log display_graph 400,300

 log10 (+):

   Compute the pointwise base-10 logarithm of selected images.

   Example:
     [#1] image.jpg +add 1 log10[-1]
     [#2] 300,1,1,1,'7*x/w+u' +log10 display_graph 400,300

 log2 (+):

   Compute the pointwise base-2 logarithm of selected images

   Example:
     [#1] image.jpg +add 1 log2[-1]
     [#2] 300,1,1,1,'7*x/w+u' +log2 display_graph 400,300

 max (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the maximum between selected images and specified value, image or mathematical expression, or compute the pointwise maxima between selected images.

   Example:
     [#1] image.jpg +mirror x max
     [#2] image.jpg max 'R=((x/w-0.5)^2+(y/h-0.5)^2)^0.5;255*R'

 maxabs (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the maxabs between selected images and specified value, image or mathematical expression, or compute the pointwise maxabs between selected images.

 m/ (+):
     Shortcut for command 'mdiv'.

 mdiv (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the matrix division of selected matrices/vectors by specified value, image or mathematical expression, or compute the matrix division of selected images.
   (equivalent to shortcut command 'm/').

 med:

   Compute the median of selected images.

   Example:
     [#1] image.jpg sample lena,lion,square +med

 min (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the minimum between selected images and specified value, image or mathematical expression, or compute the pointwise minima between selected images.

   Example:
     [#1] image.jpg +mirror x min
     [#2] image.jpg min 'R=((x/w-0.5)^2+(y/h-0.5)^2)^0.5;255*R'

 minabs (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the minabs between selected images and specified value, image or mathematical expression, or compute the pointwise minabs between selected images.

 % (+):
     Shortcut for command 'mod'.

 mod (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the modulo of selected images with specified value, image or mathematical expression, or compute the pointwise sequential modulo of selected images.
   (equivalent to shortcut command '%').

   Example:
     [#1] image.jpg +mirror x mod
     [#2] image.jpg mod 'R=((x/w-0.5)^2+(y/h-0.5)^2)^0.5;255*R'

 m* (+):
     Shortcut for command 'mmul'.

 mmul (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the matrix right multiplication of selected matrices/vectors by specified value, image or mathematical expression, or compute the matrix right multiplication of selected images.
   (equivalent to shortcut command 'm*').

   Example:
     [#1] (0,1,0;0,0,1;1,0,0) (1;2;3) +mmul

 * (+):
     Shortcut for command 'mul'.

 mul (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Multiply selected images by specified value, image or mathematical expression, or compute the pointwise product of selected images.
   (equivalent to shortcut command '*').

   See also: add, sub, div.

   Example:
     [#1] image.jpg +mul 2 cut 0,255
     [#2] image.jpg (1,2,3,4,5,6,7,8) ri[-1] [0] mul[0] [-1]
     [#3] image.jpg mul '1-3*abs(x/w-0.5)' cut 0,255
     [#4] image.jpg +luminance negate[-1] +mul

 mul_channels:
     value1,_value2,...,_valueN

   Multiply channels of selected images by specified sequence of values.

   Example:
     [#1] image.jpg +mul_channels 1,0.5,0.8

 mul_complex:
     [multiplier_real,multiplier_imag]

   Perform multiplication of the selected complex pairs (real1,imag1,...,realN,imagN) of images by
   specified complex pair of images (multiplier_real,multiplier_imag).
   In complex pairs, the real image must be always located before the imaginary image in the image list.

 != (+):
     Shortcut for command 'neq'.

 neq (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the boolean inequality of selected images with specified value, image or mathematical expression, or compute the boolean inequality of selected images.
   (equivalent to shortcut command '!=').

   Example:
     [#1] image.jpg round 40 neq {round(ia,40)}

 | (+):
     Shortcut for command 'or'.

 or (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the bitwise OR of selected images with specified value, image or mathematical expression, or compute the pointwise sequential bitwise OR of selected images.
   (equivalent to shortcut command '|').

   Example:
     [#1] image.jpg or 128
     [#2] image.jpg +mirror x or

 ^ (+):
     Shortcut for command 'pow'.

 pow (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Raise selected images to the power of specified value, image or mathematical expression, or compute the pointwise sequential powers of selected images.
   (equivalent to shortcut command '^').

   Example:
     [#1] image.jpg div 255 +pow 0.5 mul 255
     [#2] image.jpg gradient pow 2 add pow 0.2

 rad2deg:

   Convert pointwise angle values of selected images, from radians to degrees (apply 'i*180/pi').

 rol (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the bitwise left rotation of selected images with specified value, image or mathematical expression, or compute the pointwise sequential bitwise left rotation of selected images.

   Example:
     [#1] image.jpg rol 'round(3*x/w,0)' cut 0,255

 ror (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the bitwise right rotation of selected images with specified value, image or mathematical expression, or compute the pointwise sequential bitwise right rotation of selected images.

   Example:
     [#1] image.jpg ror 'round(3*x/w,0)' cut 0,255

 sign (+):

   Compute the pointwise sign of selected images.

   Example:
     [#1] image.jpg +sub {ia} sign[-1]
     [#2] 300,1,1,1,'cos(20*x/w+u)' +sign display_graph 400,300

 sin (+):

   Compute the pointwise sine of selected images.

   Example:
     [#1] image.jpg +normalize 0,{2*pi} sin[-1]
     [#2] 300,1,1,1,'20*x/w+u' +sin display_graph 400,300

   Tutorial: https://gmic.eu/oldtutorial/trigometric-and-inverse-trigometric-commands

 sinc (+):

   Compute the pointwise sinc function of selected images.

   Example:
     [#1] image.jpg +normalize {-2*pi},{2*pi} sinc[-1]
     [#2] 300,1,1,1,'20*x/w+u' +sinc display_graph 400,300

 sinh (+):

   Compute the pointwise hyperbolic sine of selected images.

   Example:
     [#1] image.jpg +normalize -3,3 sinh[-1]
     [#2] 300,1,1,1,'4*x/w+u' +sinh display_graph 400,300

 sqr (+):

   Compute the pointwise square function of selected images.

   Example:
     [#1] image.jpg +sqr
     [#2] 300,1,1,1,'40*x/w+u' +sqr display_graph 400,300

 sqrt (+):

   Compute the pointwise square root of selected images.

   Example:
     [#1] image.jpg +sqrt
     [#2] 300,1,1,1,'40*x/w+u' +sqrt display_graph 400,300

 - (+):
     Shortcut for command 'sub'.

 sub (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Subtract specified value, image or mathematical expression to selected images, or compute the pointwise difference of selected images.
   (equivalent to shortcut command '-').

   Example:
     [#1] image.jpg +sub 30% cut 0,255
     [#2] image.jpg +mirror x sub[-1] [0]
     [#3] image.jpg sub 'i(w/2+0.9*(x-w/2),y)'
     [#4] image.jpg +mirror x sub

 tan (+):

   Compute the pointwise tangent of selected images.

   Example:
     [#1] image.jpg +normalize {-0.47*pi},{0.47*pi} tan[-1]
     [#2] 300,1,1,1,'20*x/w+u' +tan display_graph 400,300

   Tutorial: https://gmic.eu/oldtutorial/trigometric-and-inverse-trigometric-commands

 tanh (+):

   Compute the pointwise hyperbolic tangent of selected images.

   Example:
     [#1] image.jpg +normalize -3,3 tanh[-1]
     [#2] 300,1,1,1,'4*x/w+u' +tanh display_graph 400,300

 xor (+):
     value[%] |
     [image] |
     'formula' |
     (no arg)

   Compute the bitwise XOR of selected images with specified value, image or mathematical expression, or compute the pointwise sequential bitwise XOR of selected images.

   Example:
     [#1] image.jpg xor 128
     [#2] image.jpg +mirror x xor

 12.5. Values Manipulation
       -------------------

 apply_curve:
     0<=smoothness<=1,x0,y0,x1,y1,x2,y2,...,xN,yN

   Apply curve transformation to image values.

   Default values: 'smoothness=1', 'x0=0', 'y0=100'.

   Example:
     [#1] image.jpg +apply_curve 1,0,0,128,255,255,0

 apply_gamma:
     gamma>=0

   Apply gamma correction to selected images.

   Example:
     [#1] image.jpg +apply_gamma 2

 balance_gamma:
     _ref_color1,...

   Compute gamma-corrected color balance of selected image, with respect to specified reference color.

   Default value: 'ref_color1=128'.

   Example:
     [#1] image.jpg +balance_gamma 128,64,64

 cast:
     datatype_source,datatype_target

   Cast datatype of image buffer from specified source type to specified target type.
   'datatype_source' and 'datatype_target' can be { uchar | char | ushort | short | uint | int | uint64 | int64 | float | double }.

 complex2polar:

   Compute complex to polar transforms of selected images.

   Example:
     [#1] image.jpg +fft complex2polar[-2,-1] log[-2] shift[-2] 50%,50%,0,0,2 remove[-1]

 compress_clut:
     _max_error>0,_avg_error>0,_max_nbpoints>=8 | 0 (unlimited),_error_metric={ 0=L2-norm | 1=deltaE_1976 | 2=deltaE_2000 },_reconstruction_colorspace={ 0=srgb | 1=rgb | 2=lab },_try_rbf_first={ 0 | 1
       }

   Compress selected color LUTs as sequences of colored keypoints.

   Default values: 'max_error=1.5', 'avg_error=0.75', 'max_nb_points=2048', 'error_metric=2', 'reconstruction_colorspace=0' and 'try_rbf_first=1'.

 compress_rle:
     _is_binary_data={ 0 | 1 },_maximum_sequence_length>=0

   Compress selected images as 2xN data matrices, using RLE algorithm.
   Set 'maximum_sequence_length=0' to disable maximum length constraint.

   Default values: 'is_binary_data=0' and 'maximum_sequence_length=0'.

   Example:
     [#1] image.jpg resize2dy 100 quantize 4 round +compress_rle , +decompress_rle[-1]

 cumulate (+):
     { x | y | z | c }...{ x | y | z | c } |
     (no arg)

   Compute the cumulative function of specified image data, optionally along the specified axes.

   Example:
     [#1] image.jpg +histogram 256 +cumulate[-1] display_graph[-2,-1] 400,300,3

 c (+):
     Shortcut for command 'cut'.

 cut (+):
     { value0[%] | [image0] },{ value1[%] | [image1] } |
     [image]

   Cut values of selected images in specified range.
   (equivalent to shortcut command 'c').

   Example:
     [#1] image.jpg +add 30% cut[-1] 0,255
     [#2] image.jpg +cut 25%,75%

 decompress_clut:
     _width>0,_height>0,_depth>0,_reconstruction_colorspace={ 0=srgb | 1=rgb | 2=lab }

   Decompress selected colored keypoints into 3D CLUTs, using a mixed RBF/PDE approach.

   Default values: 'width=height=depth=33' and 'reconstruction_colorspace=0'.

 decompress_clut_rbf:
     _width>0,_height>0,_depth>0,_reconstruction_colorspace={ 0=srgb | 1=rgb | 2=lab }

   Decompress selected colored keypoints into 3D CLUTs, using RBF thin plate spline interpolation.

   Default value: 'width=height=depth=33' and 'reconstruction_colorspace=0'.

 decompress_clut_pde:
     _width>0,_height>0,_depth>0,_reconstruction_colorspace={ 0=srgb | 1=rgb | 2=lab }

   Decompress selected colored keypoints into 3D CLUTs, using multiscale diffusion PDE's.

   Default values: 'width=height=depth=33' and 'reconstruction_colorspace=0'.

 decompress_rle:

   Decompress selected data vectors, using RLE algorithm.

 discard (+):
     _value1,_value2,... |
     { x | y | z | c}...{ x | y | z | c},_value1,_value2,... |
     (no arg)

   Discard specified values in selected images or discard neighboring duplicate values,
   optionally only for the values along the first of a specified axis.
   If no arguments are specified, neighboring duplicate values are discarded.
   If all pixels of a selected image are discarded, an empty image is returned.

   Example:
     [#1] (1;2;3;4;3;2;1) +discard 2
     [#2] (1,2,2,3,3,3,4,4,4,4) +discard x

 eigen2tensor:

   Recompose selected pairs of eigenvalues/eigenvectors as 2x2 or 3x3 tensor fields.

   Tutorial: https://gmic.eu/tutorial/eigen2tensor

 endian (+):
     _datatype

   Reverse data endianness of selected images, eventually considering the pixel being of the specified datatype.
   'datatype' can be { bool | uchar | char | ushort | short | uint | int | uint64 | int64 | float | double }.
   This command does nothing for 'bool', 'uchar' and 'char' datatypes.

 equalize (+):
     _nb_levels>0[%],_value_min[%],_value_max[%]

   Equalize histograms of selected images.
   If value range is specified, the equalization is done only for pixels in the specified
   value range.

   Default values: 'nb_levels=256', 'value_min=0%' and 'value_max=100%'.

   Example:
     [#1] image.jpg +equalize
     [#2] image.jpg +equalize 4,0,128

 f (+):
     Shortcut for command 'fill'.

 fill (+):
     value1,_value2,... |
     [image] |
     'formula'

   Fill selected images with values read from the specified value list, existing image
   or mathematical expression. Single quotes may be omitted in 'formula'.
   (equivalent to shortcut command 'f').

   Example:
     [#1] 4,4 fill 1,2,3,4,5,6,7
     [#2] 4,4 (1,2,3,4,5,6,7) fill[-2] [-1]
     [#3] 400,400,1,3 fill "X=x-w/2; Y=y-h/2; R=sqrt(X^2+Y^2); a=atan2(Y,X); if(R<=180,255*abs(cos(c+200*(x/w-0.5)*(y/h-0.5))),850*(a%(0.1*(c+1))))"

   Tutorial: https://gmic.eu/tutorial/fill

 index (+):
     { [palette] | palette_name },0<=_dithering<=1,_map_palette={ 0 | 1 }

   Index selected vector-valued images by specified vector-valued palette.
   'palette_name' can be { default | hsv | lines | hot | cool | jet | flag | cube | rainbow | algae | amp |balance | curl | deep | delta | dense | diff | haline | ice | matter | oxy | phase
   | rain | solar | speed | tarn |tempo | thermal | topo | turbid | aurora | hocuspocus | srb2 | uzebox }

   Default values: 'dithering=0' and 'map_palette=0'.

   Example:
     [#1] image.jpg +index 1,1,1
     [#2] image.jpg (0;255;255^0;128;255^0;0;255) +index[-2] [-1],1,1

   Tutorial: https://gmic.eu/tutorial/gindex

 ir:
     Shortcut for command 'inrange'.

 inrange:
     min[%],max[%],_include_min_boundary={ 0=no | 1=yes },_include_max_boundary={ 0=no | 1=yes }

   Detect pixels whose values are in specified range '[min,max]', in selected images.
   (equivalent to shortcut command 'ir').

   Default value: 'include_min_boundary=include_max_boundary=1'.

   Example:
     [#1] image.jpg +inrange 25%,75%

 map (+):
     [palette],_boundary_conditions |
     palette_name,_boundary_conditions

   Map specified vector-valued palette to selected indexed scalar images.
   'palette_name' can be { default | hsv | lines | hot | cool | jet | flag | cube | rainbow | algae | amp | balance | curl | deep | delta | dense | diff | gray | haline | ice | matter | oxy
   | phase | rain | solar | speed | tarn | tempo | thermal | topo | turbid | aurora | hocuspocus | srb2 | uzebox }
   'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }.

   Default value: 'boundary_conditions=0'.

   Example:
     [#1] image.jpg +luminance map[-1] 3
     [#2] image.jpg +rgb2ycbcr split[-1] c (0,255,0) resize[-1] 256,1,1,1,3 map[-4] [-1] remove[-1] append[-3--1] c ycbcr2rgb[-1]

   Tutorial: https://gmic.eu/tutorial/map

 mix_channels:
     (a00,...,aMN) |
     [matrix]

   Apply specified matrix to channels of selected images.

   Example:
     [#1] image.jpg +mix_channels (0,1,0;1,0,0;0,0,1)

 negate:
     base_value |
     (no arg)

   Negate image values.

   Default value: 'base_value=(undefined)'.

   Example:
     [#1] image.jpg +negate

 noise (+):
     std_deviation>=0[%],_noise_type

   Add random noise to selected images.
   'noise_type' can be { 0=gaussian | 1=uniform | 2=salt&pepper | 3=poisson | 4=rice }.

   Default value: 'noise_type=0'.

   Example:
     [#1] image.jpg +noise[0] 50,0 +noise[0] 50,1 +noise[0] 10,2 cut 0,255
     [#2] 300,300,1,3 [0] noise[0] 20,0 noise[1] 20,1 +histogram 100 display_graph[-2,-1] 400,300,3

 noise_perlin:
     _scale_x[%]>0,_scale_y[%]>0,_scale_z[%]>0,_seed_x,_seed_y,_seed_z

   Render 2D or 3D Perlin noise on selected images, from specified coordinates.
   The Perlin noise is a specific type of smooth noise,
   described here : https://en.wikipedia.org/wiki/Perlin_noise.

   Default values: 'scale_x=scale_y=scale_z=16' and 'seed_x=seed_y=seed_z=0'.

   Example:
     [#1] 500,500,1,3 noise_perlin ,

 noise_poissondisk:
     _radius[%]>0,_max_sample_attempts>0

   Add poisson disk sampling noise to selected images.
   Implements the algorithm from the article "Fast Poisson Disk Sampling in Arbitrary Dimensions",
   by Robert Bridson (SIGGRAPH'2007).

   Default values: 'radius=8' and 'max_sample_attempts=30'.

   Example:
     [#1] 300,300 noise_poissondisk 8

 normp:
     p>=0

   Compute the pointwise Lp-norm norm of vector-valued pixels in selected images.

   Default value: 'p=2'.

   Example:
     [#1] image.jpg +normp[0] 0 +normp[0] 1 +normp[0] 2 +normp[0] inf

 norm:

   Compute the pointwise euclidean norm of vector-valued pixels in selected images.

   Example:
     [#1] image.jpg +norm

   Tutorial: https://gmic.eu/tutorial/norm

 n (+):
     Shortcut for command 'normalize'.

 normalize (+):
     { value0[%] | [image0] },{ value1[%] | [image1] },_constant_case_ratio |
     [image]

   Linearly normalize values of selected images in specified range.
   (equivalent to shortcut command 'n').

   Example:
     [#1] image.jpg split x,2 normalize[-1] 64,196 append x

   Tutorial: https://gmic.eu/tutorial/normalize

 normalize_l2:

   Normalize selected images such that they have a unit L2 norm.

 normalize_sum:

   Normalize selected images such that they have a unit sum.

   Example:
     [#1] image.jpg +histogram 256 normalize_sum[-1] display_graph[-1] 400,300

 not:

   Apply boolean not operation on selected images.

   Example:
     [#1] image.jpg +ge 50% +not[-1]

 orientation:

   Compute the pointwise orientation of vector-valued pixels in selected images.

   Example:
     [#1] image.jpg +orientation +norm[-2] negate[-1] mul[-2] [-1] reverse[-2,-1]

   Tutorial: https://gmic.eu/tutorial/orientation

 oneminus:

   For each selected image, compute one minus image.

   Example:
     [#1] image.jpg normalize 0,1 +oneminus

 otsu:
     _nb_levels>0

   Hard-threshold selected images using Otsu's method.
   The computed thresholds are returned as a list of values in the status.

   Default value: 'nb_levels=256'.

   Example:
     [#1] image.jpg luminance +otsu ,

 polar2complex:

   Compute polar to complex transforms of selected images.

 quantize:
     nb_levels>=1,_keep_values={ 0 | 1 },_quantization_type={ -1=median-cut | 0=k-means | 1=uniform }

   Quantize selected images.

   Default value: 'keep_values=1' and 'quantization_type=0'.

   Example:
     [#1] image.jpg luminance +quantize 3
     [#2] 200,200,1,1,'cos(x/10)*sin(y/10)' +quantize[0] 6 +quantize[0] 4 +quantize[0] 3 +quantize[0] 2

 quantize_area:
     _min_area>0

   Quantize selected images such that each flat region has an area greater or equal to 'min_area'.

   Default value: 'min_area=10'.

   Example:
     [#1] image.jpg quantize 3 +blur 1 round[-1] +quantize_area[-1] 2

 rand (+):
     { value0[%] | [image0] },_{ value1[%] | [image1] } |
     [image]

   Fill selected images with random values uniformly distributed in the specified range.

   Example:
     [#1] 400,400,1,3 rand -10,10 +blur 10 sign[-1]

 replace:
     source,target

   Replace pixel values in selected images.

   Example:
     [#1] (1;2;3;4) +replace 2,3

 replace_inf:
     _expression

   Replace all infinite values in selected images by specified expression.

   Example:
     [#1] (0;1;2) log +replace_inf 2

 replace_nan:
     _expression

   Replace all NaN values in selected images by specified expression.

   Example:
     [#1] (-1;0;2) sqrt +replace_nan 2

 replace_naninf:
     _expression

   Replace all NaN and infinite values in selected images by specified expression.

 replace_seq:
     "search_seq","replace_seq"

   Search and replace a sequence of values in selected images.

   Example:
     [#1] (1;2;3;4;5) +replace_seq "2,3,4","7,8"

 replace_str:
     "search_str","replace_str"

   Search and replace a string in selected images (viewed as strings, i.e. sequences of character codes).

   Example:
     [#1] ('"Hello there, how are you ?"') +replace_str "Hello there","Hi David"

 round (+):
     rounding_value>=0,_rounding_type |
     (no arg)

   Round values of selected images.
   'rounding_type' can be { -1=backward | 0=nearest | 1=forward }.

   Default value: 'rounding_type=0'.

   Example:
     [#1] image.jpg +round 100
     [#2] image.jpg mul {pi/180} sin +round

 roundify:
     gamma>=0

   Apply roundify transformation on float-valued data, with specified gamma.

   Default value: 'gamma=0'.

   Example:
     [#1] 1000 fill '4*x/w' repeat 5 +roundify[0] {$>*0.2} done append c display_graph 400,300

 = (+):
     Shortcut for command 'set'.

 set (+):
     value,_x[%],_y[%],_z[%],_c[%]

   Set pixel value in selected images, at specified coordinates.
   (equivalent to shortcut command '=').

   If specified coordinates are outside the image bounds, no action is performed.

   Default values: 'x=y=z=c=0'.

   Example:
     [#1] 2,2 set 1,0,0 set 2,1,0 set 3,0,1 set 4,1,1
     [#2] image.jpg repeat 10000 set 255,{u(100)}%,{u(100)}%,0,{u(100)}% done

 threshold:
     value[%],_is_soft={ 0 | 1 } :

   Threshold values of selected images.
   'soft' can be { 0=hard-thresholding | 1=soft-thresholding }.

   Default value: 'is_soft=0'.

   Example:
     [#1] image.jpg +threshold[0] 50% +threshold[0] 50%,1

   Tutorial: https://gmic.eu/tutorial/threshold

 vector2tensor:

   Convert selected vector fields to corresponding tensor fields.

 12.6. Colors
       ------

 adjust_colors:
     -100<=_brightness<=100,-100<=_contrast<=100,-100<=_gamma<=100,-100<=_hue_shift<=100,-100<=_saturation<=100,_value_min,_value_max

   Perform a global adjustment of colors on selected images.
   Range of correct image values are considered to be in [value_min,value_max] (e.g. [0,255]).
   If 'value_min==value_max==0', value range is estimated from min/max values of selected images.
   Processed images have pixel values constrained in [value_min,value_max].

   Default values: 'brightness=0', 'contrast=0', 'gamma=0', 'hue_shift=0', 'saturation=0', 'value_min=value_max=0'.

   Example:
     [#1] image.jpg +adjust_colors 0,30,0,0,30

 ac:
     Shortcut for command 'apply_channels'.

 apply_channels:
     "command",color_channels,_value_action={ 0=none | 1=cut | 2=normalize }

   Apply specified command on the chosen color channel(s) of each selected images.
   (equivalent to shortcut command 'ac').

   Argument 'color_channels' refers to a colorspace, and can be basically one of
   { all | rgba | [s]rgb | ryb | lrgb | ycbcr | lab | lch | hsv | hsi | hsl | cmy | cmyk | yiq }.
   You can also make the processing focus on a few particular channels of this colorspace,
   by setting 'color_channels' as 'colorspace_channel' (e.g. 'hsv_h' for the hue).
   All channel values are considered to be provided in the [0,255] range.

   Default value: 'value_action=0'.

   Example:
     [#1] image.jpg +apply_channels "equalize blur 2",ycbcr_cbcr

 autoindex:
     nb_colors>0,0<=_dithering<=1,_method={ 0=median-cut | 1=k-means }

   Index selected vector-valued images by adapted colormaps.

   Default values: 'dithering=0' and 'method=1'.

   Example:
     [#1] image.jpg +autoindex[0] 4 +autoindex[0] 8 +autoindex[0] 16

 bayer2rgb:
     _GM_smoothness,_RB_smoothness1,_RB_smoothness2

   Transform selected RGB-Bayer sampled images to color images.

   Default values: 'GM_smoothness=RB_smoothness=1' and 'RB_smoothness2=0.5'.

   Example:
     [#1] image.jpg rgb2bayer 0 +bayer2rgb 1,1,0.5

 deltaE:
     [image],_metric={ 0=deltaE_1976 | 1=deltaE_2000 },"_to_Lab_command"

   Compute the CIE DeltaE color difference between selected images and specified [image].
   Argument 'to_Lab_command' is a command able to convert colors of [image] into a Lab representation.

   Default values: 'metric=1' and 'to_Lab_command="srgb2lab"'.

   Example:
     [#1] image.jpg +blur 2 +deltaE[0] [1],1,srgb2lab

 cmy2rgb:

   Convert color representation of selected images from CMY to RGB.

 cmyk2rgb:

   Convert color representation of selected images from CMYK to RGB.

 colorblind:
     type={ 0=protanopia | 1=protanomaly | 2=deuteranopia | 3=deuteranomaly | 4=tritanopia | 5=tritanomaly | 6=achromatopsia | 7=achromatomaly }

   Simulate color blindness vision.

   Example:
     [#1] image.jpg +colorblind 0

 colormap:
     nb_levels>=0,_method={ 0=median-cut | 1=k-means },_sort_vectors

   Estimate best-fitting colormap with 'nb_colors' entries, to index selected images.
   Set 'nb_levels==0' to extract all existing colors of an image.
   'sort_vectors' can be { 0=unsorted | 1=by increasing norm | 2=by decreasing occurrence }.

   Default value: 'method=1' and 'sort_vectors=1'.

   Example:
     [#1] image.jpg +colormap[0] 4 +colormap[0] 8 +colormap[0] 16

   Tutorial: https://gmic.eu/oldtutorial/_colormap

 compose_channels:

   Compose all channels of each selected image, using specified arithmetic operator (+,-,or,min,...).

   Default value: '1=+'.

   Example:
     [#1] image.jpg +compose_channels and

   Tutorial: https://gmic.eu/tutorial/compose_channels

 direction2rgb:

   Compute RGB representation of selected 2D direction fields.

   Example:
     [#1] image.jpg luminance gradient append c blur 2 orientation +direction2rgb

 ditheredbw:

   Create dithered B&W version of selected images.

   Example:
     [#1] image.jpg +equalize ditheredbw[-1]

 fc:
     Shortcut for command 'fill_color'.

 fill_color:
     col1,...,colN

   Fill selected images with specified color.
   (equivalent to shortcut command 'fc').

   Example:
     [#1] image.jpg +fill_color 255,0,255

   Tutorial: https://gmic.eu/oldtutorial/_fill_color

 gradient2rgb:
     _is_orientation={ 0 | 1 }

   Compute RGB representation of 2D gradient of selected images.

   Default value: 'is_orientation=0'.

   Example:
     [#1] image.jpg +gradient2rgb 0 equalize[-1]

 hcy2rgb:

   Convert color representation of selected images from HCY to RGB.

 hsi2rgb:

   Convert color representation of selected images from HSI to RGB.

 hsi82rgb:

   Convert color representation of selected images from HSI8 to RGB.

 hsl2rgb:

   Convert color representation of selected images from HSL to RGB.

 hsl82rgb:

   Convert color representation of selected images from HSL8 to RGB.

 hsv2rgb:

   Convert color representation of selected images from HSV to RGB.

   Example:
     [#1] (0,360;0,360^0,0;1,1^1,1;1,1) resize 400,400,1,3,3 hsv2rgb

 hsv82rgb:

   Convert color representation of selected images from HSV8 to RGB.

 int2rgb:

   Convert color representation of selected images from INT24 to RGB.

 jzazbz2rgb:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from RGB to Jzazbz.

   Default value: 'illuminant=2'.

 jzazbz2xyz:

   Convert color representation of selected images from RGB to XYZ.

 lab2lch:

   Convert color representation of selected images from Lab to Lch.

 lab2rgb:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from Lab to RGB.

   Default value: 'illuminant=2'.

   Example:
     [#1] (50,50;50,50^-3,3;-3,3^-3,-3;3,3) resize 400,400,1,3,3 lab2rgb

 lab2srgb:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from Lab to sRGB.

   Default value: 'illuminant=2'.

   Example:
     [#1] (50,50;50,50^-3,3;-3,3^-3,-3;3,3) resize 400,400,1,3,3 lab2rgb

 lab82srgb:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from Lab8 to sRGB.

   Default value: 'illuminant=2'.

   Example:
     [#1] (50,50;50,50^-3,3;-3,3^-3,-3;3,3) resize 400,400,1,3,3 lab2rgb

 lab2xyz:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from Lab to XYZ.

   Default value: 'illuminant=2'.

 lab82rgb:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from Lab8 to RGB.

   Default value: 'illuminant=2'.

 lch2lab:

   Convert color representation of selected images from Lch to Lab.

 lch2rgb:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from Lch to RGB.

   Default value: 'illuminant=2'.

 lch82rgb:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from Lch8 to RGB.

   Default value: 'illuminant=2'.

 luminance:

   Compute luminance of selected sRGB images.

   Example:
     [#1] image.jpg +luminance

   Tutorial: https://gmic.eu/tutorial/luminance

 lightness:

   Compute lightness of selected sRGB images.

   Example:
     [#1] image.jpg +lightness

 lut_contrast:
     _nb_colors>1,_min_rgb_value

   Generate a RGB colormap where consecutive colors have high contrast.
   This function performs a specific score maximization to generate the result, so
   it may take some time when 'nb_colors' is high.

   Default values: 'nb_colors=256' and 'min_rgb_value=64'.

 map_clut:
     [clut] | "clut_name"

   Map specified RGB color LUT to selected images.

   Example:
     [#1] image.jpg uniform_distribution {2^6},3 mirror[-1] x +map_clut[0] [1]

 mix_rgb:
     a11,a12,a13,a21,a22,a23,a31,a32,a33

   Apply 3x3 specified matrix to RGB colors of selected images.

   Default values: 'a11=1', 'a12=a13=a21=0', 'a22=1', 'a23=a31=a32=0' and 'a33=1'.

   Example:
     [#1] image.jpg +mix_rgb 0,1,0,1,0,0,0,0,1

   Tutorial: https://gmic.eu/tutorial/mix_rgb

 oklab2rgb:

   Convert color representation of selected images from OKlab to RGB.
   (see colorspace definition at: https://bottosson.github.io/posts/oklab/ ).
   See also: rgb2oklab.

 palette:
     palette_name | palette_number

   Input specified color palette at the end of the image list.
   'palette_name' can be { default | hsv | lines | hot | cool | jet | flag | cube | rainbow | parula | spring | summer | autumn | winter | bone | copper | pink | vga | algae | amp | balance
   | curl | deep | delta | dense | diff | gray | haline | ice | matter | oxy | phase | rain | solar | speed | tarn | tempo | thermal | topo | turbid | aurora | hocuspocus | srb2 | uzebox | amiga7800
   | amiga7800mess | fornaxvoid1 }

   Example:
     [#1] palette hsv

 pseudogray:
     _max_increment>=0,_JND_threshold>=0,_bits_depth>0

   Generate pseudogray colormap with specified increment and perceptual threshold.
   If 'JND_threshold' is 0, no perceptual constraints are applied.

   Default values: 'max_increment=5', 'JND_threshold=2.3' and 'bits_depth=8'.

   Example:
     [#1] pseudogray 5

 replace_color:
     tolerance[%]>=0,smoothness[%]>=0,src1,src2,...,dest1,dest2,...

   Replace pixels from/to specified colors in selected images.

   Example:
     [#1] image.jpg +replace_color 40,3,204,153,110,255,0,0

 retinex:
     _value_offset>0,_colorspace={ hsi | hsv | lab | lrgb | rgb | ycbcr },0<=_min_cut<=100,0<=_max_cut<=100,_sigma_low>0,_sigma_mid>0,_sigma_high>0

   Apply multi-scale retinex algorithm on selected images to improve color consistency.
   (as described in the page http://www.ipol.im/pub/art/2014/107/).

   Default values: 'offset=1', 'colorspace=hsv', 'min_cut=1', 'max_cut=1', 'sigma_low=15','sigma_mid=80' and 'sigma_high=250'.

 rgb2bayer:
     _start_pattern=0,_color_grid=0

   Transform selected color images to RGB-Bayer sampled images.

   Default values: 'start_pattern=0' and 'color_grid=0'.

   Example:
     [#1] image.jpg +rgb2bayer 0

 rgb2cmy:

   Convert color representation of selected images from RGB to CMY.

   Example:
     [#1] image.jpg rgb2cmy split c

 rgb2cmyk:

   Convert color representation of selected images from RGB to CMYK.

   Example:
     [#1] image.jpg rgb2cmyk split c
     [#2] image.jpg rgb2cmyk split c fill[3] 0 append c cmyk2rgb

 rgb2hcy:

   Convert color representation of selected images from RGB to HCY.

   Example:
     [#1] image.jpg rgb2hcy split c

 rgb2hsi:

   Convert color representation of selected images from RGB to HSI.

   Example:
     [#1] image.jpg rgb2hsi split c

 rgb2hsi8:

   Convert color representation of selected images from RGB to HSI8.

   Example:
     [#1] image.jpg rgb2hsi8 split c

 rgb2hsl:

   Convert color representation of selected images from RGB to HSL.

   Example:
     [#1] image.jpg rgb2hsl split c
     [#2] image.jpg rgb2hsl +split c add[-3] 100 mod[-3] 360 append[-3--1] c hsl2rgb

 rgb2hsl8:

   Convert color representation of selected images from RGB to HSL8.

   Example:
     [#1] image.jpg rgb2hsl8 split c

 rgb2hsv:

   Convert color representation of selected images from RGB to HSV.

   Example:
     [#1] image.jpg rgb2hsv split c
     [#2] image.jpg rgb2hsv +split c add[-2] 0.3 cut[-2] 0,1 append[-3--1] c hsv2rgb

 rgb2hsv8:

   Convert color representation of selected images from RGB to HSV8.

   Example:
     [#1] image.jpg rgb2hsv8 split c

 rgb2int:

   Convert color representation of selected images from RGB to INT24 scalars.

   Example:
     [#1] image.jpg rgb2int

 rgb2jzazbz:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from RGB to Jzazbz.

   Default value: 'illuminant=2'.

 rgb2lab:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from RGB to Lab.

   Default value: 'illuminant=2'.

 rgb2lab8:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from RGB to Lab8.

   Default value: 'illuminant=2'.

   Example:
     [#1] image.jpg rgb2lab8 split c

 rgb2lch:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from RGB to Lch.

   Default value: 'illuminant=2'.

   Example:
     [#1] image.jpg rgb2lch split c

 rgb2lch8:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from RGB to Lch8.

   Default value: 'illuminant=2'.

   Example:
     [#1] image.jpg rgb2lch8 split c

 rgb2luv:

   Convert color representation of selected images from RGB to LUV.

   Example:
     [#1] image.jpg rgb2luv split c

 rgb2oklab:

   Convert color representation of selected images from RGB to Oklab.
   (see colorspace definition at: https://bottosson.github.io/posts/oklab/ ).
   See also: oklab2rgb.

 rgb2ryb:

   Convert color representation of selected images from RGB to RYB.

   Example:
     [#1] image.jpg rgb2ryb split c

 rgb2srgb:

   Convert color representation of selected images from linear RGB to sRGB.

 rgb2xyz:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from RGB to XYZ.

   Default value: 'illuminant=2'.

   Example:
     [#1] image.jpg rgb2xyz split c

 rgb2xyz8:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from RGB to XYZ8.

   Default value: 'illuminant=2'.

   Example:
     [#1] image.jpg rgb2xyz8 split c

 rgb2yiq:

   Convert color representation of selected images from RGB to YIQ.

   Example:
     [#1] image.jpg rgb2yiq split c

 rgb2yiq8:

   Convert color representation of selected images from RGB to YIQ8.

   Example:
     [#1] image.jpg rgb2yiq8 split c

 rgb2ycbcr:

   Convert color representation of selected images from RGB to YCbCr.

   Example:
     [#1] image.jpg rgb2ycbcr split c

 rgb2yuv:

   Convert color representation of selected images from RGB to YUV.

   Example:
     [#1] image.jpg rgb2yuv split c

 rgb2yuv8:

   Convert color representation of selected images from RGB to YUV8.

   Example:
     [#1] image.jpg rgb2yuv8 split c

 remove_opacity:

   Remove opacity channel of selected images.

 ryb2rgb:

   Convert color representation of selected images from RYB to RGB.

 select_color:
     tolerance[%]>=0,col1,...,colN

   Select pixels with specified color in selected images.

   Example:
     [#1] image.jpg +select_color 40,204,153,110

   Tutorial: https://gmic.eu/oldtutorial/_select_color

 sepia:

   Apply sepia tones effect on selected images.

   Example:
     [#1] image.jpg sepia

 solarize:

   Solarize selected images.

   Example:
     [#1] image.jpg solarize

 split_colors:
     _tolerance>=0,_max_nb_outputs>0,_min_area>0

   Split selected images as several image containing a single color.
   One selected image can be split as at most 'max_nb_outputs' images.
   Output images are sorted by decreasing area of extracted color regions and have an additional alpha-channel.

   Default values: 'tolerance=0', 'max_nb_outputs=256' and 'min_area=8'.

   Example:
     [#1] image.jpg quantize 5 +split_colors , display_rgba

 split_opacity:

   Split color and opacity parts of selected images.

 srgb2lab:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from sRGB to Lab.

   Default value: 'illuminant=2'.

   Example:
     [#1] image.jpg srgb2lab split c
     [#2] image.jpg srgb2lab +split c mul[-2,-1] 2.5 append[-3--1] c lab2srgb

 srgb2lab8:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from sRGB to Lab8.

   Default value: 'illuminant=2'.

 srgb2rgb:

   Convert color representation of selected images from sRGB to linear RGB.

 to_a:

   Force selected images to have an alpha channel.

 to_color:

   Force selected images to be in color mode (RGB or RGBA).

 to_colormode:
     mode={ 0=adaptive | 1=G | 2=GA | 3=RGB | 4=RGBA }

   Force selected images to be in a given color mode.

   Default value: 'mode=0'.

 to_gray:

   Force selected images to be in GRAY mode.

   Example:
     [#1] image.jpg +to_gray

 to_graya:

   Force selected images to be in GRAYA mode.

 to_pseudogray:
     _max_step>=0,_is_perceptual_constraint={ 0 | 1 },_bits_depth>0

   Convert selected scalar images ([0-255]-valued) to pseudo-gray color images.

   Default values: 'max_step=5', 'is_perceptual_constraint=1' and 'bits_depth=8'.
   The original pseudo-gray technique has been introduced by Rich Franzen http://r0k.us/graphics/pseudoGrey.html.
   Extension of this technique to arbitrary increments for more tones, has been done by David Tschumperlé.

 to_rgb:

   Force selected images to be in RGB mode.

 to_rgba:

   Force selected images to be in RGBA mode.

 transfer_histogram:
     [reference_image],_nb_levels>0,_color_channels

   Transfer histogram of the specified reference image to selected images.
   Argument 'color channels' is the same as with command 'apply_channels'.

   Default value: 'nb_levels=256' and 'color_channels=all'.

   Example:
     [#1] image.jpg 100,100,1,3,"u([256,200,100])" +transfer_histogram[0] [1]

 transfer_pca:
     [reference_image],_color_channels

   Transfer mean and covariance matrix of specified vector-valued reference image to selected images.
   Argument 'color channels' is the same as with command 'apply_channels'.

   Default value: 'color_channels=all'.

   Example:
     [#1] sample lena,earth +transfer_pca[0] [1]

 transfer_rgb:
     [target],_gamma>=0,_regularization>=0,_luminosity_constraints>=0,_rgb_resolution>=0,_is_constraints={ 0 | 1 }

   Transfer colors from selected source images to selected reference image (given as argument).
   'gamma' determines the importance of color occurrences in the matching process (0=none to 1=huge).
   'regularization' determines the number of guided filter iterations to remove quantization effects.
   'luminosity_constraints' tells if luminosity constraints must be applied on non-confident matched colors.
   'is_constraints' tells if additional hard color constraints must be set (opens an interactive window).

   Default values: 'gamma=0.3','regularization=8', 'luminosity_constraints=0.1', 'rgb_resolution=64' and 'is_constraints=0'.

   Example:
     [#1] sample pencils,wall +transfer_rgb[0] [1],0,0.01

 xyz2jzazbz:

   Convert color representation of selected images from XYZ to RGB.

 xyz2lab:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from XYZ to Lab.

   Default value: 'illuminant=2'.

 xyz2rgb:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from XYZ to RGB.

   Default value: 'illuminant=2'.

 xyz82rgb:
     illuminant={ 0=D50 | 1=D65 | 2=E } |
     (no arg)

   Convert color representation of selected images from XYZ8 to RGB.

   Default value: 'illuminant=2'.

 ycbcr2rgb:

   Convert color representation of selected images from YCbCr to RGB.

 yiq2rgb:

   Convert color representation of selected images from YIQ to RGB.

 yiq82rgb:

   Convert color representation of selected images from YIQ8 to RGB.

 yuv2rgb:

   Convert color representation of selected images from YUV to RGB.

 yuv82rgb:

   Convert selected images from YUV8 to RGB color bases.

 12.7. Geometry Manipulation
       ---------------------

 a (+):
     Shortcut for command 'append'.

 append (+):
     [image],axis,_centering |
     axis,_centering

   Append specified image to selected images, or all selected images together, along specified axis.
   (equivalent to shortcut command 'a').

   'axis' can be { x | y | z | c }.
   Usual 'centering' values are { 0=left-justified | 0.5=centered | 1=right-justified }.

   Default value: 'centering=0'.

   Example:
     [#1] image.jpg split y,10 reverse append y
     [#2] image.jpg repeat 5 +rows[0] 0,{10+18*$>}% done remove[0] append x,0.5
     [#3] image.jpg append[0] [0],y

 append_tiles:
     _M>=0,_N>=0,0<=_centering_x<=1,0<=_centering_y<=1

   Append MxN selected tiles as new images.
   If 'N' is set to 0, number of rows is estimated automatically.
   If 'M' is set to 0, number of columns is estimated automatically.
   If 'M' and 'N' are both set to '0', auto-mode is used.
   If 'M' or 'N' is set to 0, only a single image is produced.
   'centering_x' and 'centering_y' tells about the centering of tiles when they have different sizes.

   Default values: 'M=0', 'N=0', 'centering_x=centering_y=0.5'.

   Example:
     [#1] image.jpg split xy,4 append_tiles ,

 apply_scales:
     "command",number_of_scales>0,_min_scale[%]>=0,_max_scale[%]>=0,_scale_gamma>0,_interpolation

   Apply specified command on different scales of selected images.
   'interpolation' can be { 0=none | 1=nearest | 2=average | 3=linear | 4=grid | 5=bicubic | 6=lanczos }.

   Default value: 'min_scale=25%', 'max_scale=100%' and 'interpolation=3'.

   Example:
     [#1] image.jpg apply_scales "blur 5 sharpen 1000",4

 autocrop (+):
     value1,value2,... |
     (no arg)

   Autocrop selected images by specified vector-valued intensity.
   If no arguments are provided, cropping value is guessed.

   Example:
     [#1] 400,400,1,3 fill_color 64,128,255 ellipse 50%,50%,120,120,0,1,255 +autocrop

 autocrop_components:
     _threshold[%],_min_area[%]>=0,_is_high_connectivity={ 0 | 1 },_output_type={ 0=crop | 1=segmentation | 2=coordinates }

   Autocrop and extract connected components in selected images, according to a mask given as the last channel of
   each of the selected image (e.g. alpha-channel).

   Default values: 'threshold=0%', 'min_area=0.1%', 'is_high_connectivity=0' and 'output_type=1'.

   Example:
     [#1] 256,256 noise 0.1,2 eq 1 dilate_circ 20 label_fg 0,1 normalize 0,255 +neq 0 *[-1] 255 append c +autocrop_components ,

 autocrop_seq:
     value1,value2,... | auto

   Autocrop selected images using the crop geometry of the last one by specified vector-valued intensity,
   or by automatic guessing the cropping value.

   Default value: auto mode.

   Example:
     [#1] image.jpg +fill[-1] 0 ellipse[-1] 50%,50%,30%,20%,0,1,1 autocrop_seq 0

 channels (+):
     { [image0] | c0[%] },_{ [image1] | c1[%] }

   Keep only specified channels of selected images.
   Dirichlet boundary is used when specified channels are out of range.

   Example:
     [#1] image.jpg channels 0,1
     [#2] image.jpg luminance channels 0,2

 columns (+):
     { [image0] | x0[%] },_{ [image1] | x1[%] }

   Keep only specified columns of selected images.
   Dirichlet boundary is used when specified columns are out of range.

   Example:
     [#1] image.jpg columns -25%,50%

 z (+):
     Shortcut for command 'crop'.

 crop (+):
     x0[%],x1[%],_boundary_conditions |
     x0[%],y0[%],x1[%],y1[%],_boundary_conditions |
     x0[%],y0[%],z0[%],x1[%],y1[%],z1[%],_boundary_conditions |
     x0[%],y0[%],z0[%],c0[%],x1[%],y1[%],z1[%],c1[%],_boundary_conditions

   Crop selected images with specified region coordinates.
   (equivalent to shortcut command 'z').

   'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }.

   Default value: 'boundary_conditions=0'.

   Example:
     [#1] image.jpg +crop -230,-230,280,280,1 crop[0] -230,-230,280,280,0
     [#2] image.jpg crop 25%,25%,75%,75%

 diagonal:

   Transform selected vectors as diagonal matrices.

   Example:
     [#1] 1,10,1,1,'y' +diagonal

 elevate:
     _depth,_is_plain={ 0 | 1 },_is_colored={ 0 | 1 }

   Elevate selected 2D images into 3D volumes.

   Default values: 'depth=64', 'is_plain=1' and 'is_colored=1'.

 expand_x:
     size_x>=0,_boundary_conditions={ 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }

   Expand selected images along the x-axis.

   Default value: 'boundary_conditions=1'.

   Example:
     [#1] image.jpg expand_x 30,0

 expand_xy:
     size>=0,_boundary_conditions={ 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }

   Expand selected images along the xy-axes.

   Default value: 'boundary_conditions=1'.

   Example:
     [#1] image.jpg expand_xy 30,0

 expand_xyz:
     size>=0,_boundary_conditions={ 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }

   Expand selected images along the xyz-axes.

   Default value: 'boundary_conditions=1'.

 expand_y:
     size_y>=0,_boundary_conditions={ 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }

   Expand selected images along the y-axis.

   Default value: 'boundary_conditions=1'.

   Example:
     [#1] image.jpg expand_y 30,0

 expand_z:
     size_z>=0,_boundary_conditions={ 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }

   Expand selected images along the z-axis.

   Default value: 'boundary_conditions=1'.

 extract:
     "condition",_output_type={ 0=xyzc-coordinates | 1=xyz-coordinates | 2=scalar-values | 3=vector-values }

   Extract a list of coordinates or values from selected image, where
   specified mathematical condition holds.
   For N coordinates matching, result is a 1xNx1x4 image.

   Default values: 'output_type=0'.

   Example:
     [#1] sp lena +extract "norm(I)>128",3

 extract_region:
     [label_image],_extract_xyz_coordinates={ 0 | 1 },_label_1,...,_label_M

   Extract all pixels of selected images whose corresponding label in '[label_image]' is equal to 'label_m',
   and output them as M column images.

   Default value: 'extract_xyz_coordinates=0'.

   Example:
     [#1] image.jpg +blur 3 quantize. 4,0 +extract_region[0] [1],0,1,3

 montage:
     "_layout_code",_montage_mode={ 0<=centering<=1 | 2<=scale+2<=3 },_output_mode={ 0=single layer | 1=multiple layers },"_processing_command"

   Create a single image montage from selected images, according to specified layout code :
    * 'X' to assemble all images using an automatically estimated layout.
    * 'H' to assemble all images horizontally.
    * 'V' to assemble all images vertically.
    * 'A' to assemble all images as an horizontal array.
    * 'B' to assemble all images as a vertical array.
    * 'Ha:b' to assemble two blocks 'a' and 'b' horizontally.
    * 'Va:b' to assemble two blocks 'a' and 'b' vertically.
    * 'Ra' to rotate a block 'a' by 90 deg. ('RRa' for 180 deg. and 'RRRa' for 270 deg.).
    * 'Ma' to mirror a block 'a' along the X-axis ('MRRa' for the Y-axis).
   A block 'a' can be an image index (treated periodically) or a nested layout expression 'Hb:c','Vb:c','Rb' or
   'Mb' itself.
   For example, layout code 'H0:V1:2' creates an image where image [0] is on the left, and images [1] and [2]
   vertically packed on the right.

   Default values: 'layout_code=X', 'montage_mode=2', output_mode='0' and 'processing_command=""'.

   Example:
     [#1] image.jpg sample ? +plasma[0] shape_cupid 256 normalize 0,255 frame 3,3,0 frame 10,10,255 to_rgb +montage A +montage[^-1] H1:V0:VH2:1H0:3

 mirror (+):
     { x | y | z }...{ x | y | z }

   Mirror selected images along specified axes.

   Example:
     [#1] image.jpg +mirror y +mirror[0] c
     [#2] image.jpg +mirror x +mirror y append_tiles 2,2

 permute (+):
     permutation_string

   Permute selected image axes by specified permutation.
   'permutation' is a combination of the character set {x|y|z|c},
   e.g. 'xycz', 'cxyz', ...

   Example:
     [#1] image.jpg permute yxzc

 r (+):
     Shortcut for command 'resize'.

 resize (+):
     {[image_w] | width>0[%]},_{[image_h] | height>0[%]},_{[image_d] | depth>0[%]},_{[image_s] | spectrum>0[%]},_interpolation,_boundary_conditions,_ax,_ay,_az,_ac

   Resize selected images with specified geometry.
   (equivalent to shortcut command 'r').

   'interpolation' can be { -1=none (memory content) | 0=none | 1=nearest | 2=average | 3=linear | 4=grid | 5=bicubic | 6=lanczos }.
   'boundary_conditions' has different meanings, according to the chosen 'interpolation' mode :
   . When 'interpolation=={ -1 | 1 | 2 | 4 }', 'boundary_conditions' is meaningless.
   . When 'interpolation==0', 'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }.
   . When 'interpolation=={ 3 | 5 | 6 }', 'boundary_conditions' can be { 0=none | 1=neumann }.
   'ax,ay,az,ac' set the centering along each axis when 'interpolation=0 or 4'
   (set to '0' by default, must be defined in range [0,1]).

   Default values: 'interpolation=1', 'boundary_conditions=0' and 'ax=ay=az=ac=0'.

   Example:
     [#1] image.jpg +resize[-1] 256,128,1,3,2 +resize[-1] 120%,120%,1,3,0,1,0.5,0.5 +resize[-1] 120%,120%,1,3,0,0,0.2,0.2 +resize[-1] [0],[0],1,3,4

 ri:
     Shortcut for command 'resize_as_image'.

 resize_as_image:
     [reference],_interpolation,_boundary_conditions,_ax,_ay,_az,_ac

   Resize selected images to the geometry of specified [reference] image.
   (equivalent to shortcut command 'ri').

   Default values: 'interpolation=1', 'boundary_conditions=0' and 'ax=ay=az=ac=0'.

   Example:
     [#1] image.jpg sample duck +resize_as_image[-1] [-2]

 resize_mn:
     width[%]>=0,_height[%]>=0,_depth[%]>=0,_B_value,_C_value

   Resize selected images with Mitchell-Netravali filter (cubic).
   For details about the method, see: https://de.wikipedia.org/wiki/Mitchell-Netravali-Filter.

   Default values: 'height=100%', 'depth=100%', 'B=0.3333' and 'C=0.3333'.

   Example:
     [#1] image.jpg resize2dx 32 resize_mn 800%,800%

 resize_pow2:
     _interpolation,_boundary_conditions,_ax,_ay,_az,_ac

   Resize selected images so that each dimension is a power of 2.
   'interpolation' can be { -1=none (memory content) | 0=none | 1=nearest | 2=average | 3=linear | 4=grid | 5=bicubic | 6=lanczos }.
   'boundary_conditions' has different meanings, according to the chosen 'interpolation' mode :
   . When 'interpolation=={ -1 | 1 | 2 | 4 }', 'boundary_conditions' is meaningless.
   . When 'interpolation==0', 'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }.
   . When 'interpolation=={ 3 | 5 | 6 }', 'boundary_conditions' can be { 0=none | 1=neumann }.
   'ax,ay,az,ac' set the centering along each axis when 'interpolation=0'
   (set to '0' by default, must be defined in range [0,1]).

   Default values: 'interpolation=0', 'boundary_conditions=0' and 'ax=ay=az=ac=0'.

   Example:
     [#1] image.jpg +resize_pow2[-1] 0

 rr2d:
     Shortcut for command 'resize_ratio2d'.

 resize_ratio2d:
     width>0,height>0,_mode={ 0=inside | 1=outside | 2=padded },0=<_interpolation<=6

   Resize selected images while preserving their aspect ratio.
   (equivalent to shortcut command 'rr2d').

   Default values: 'mode=0' and 'interpolation=6'.

 r2dx:
     Shortcut for command 'resize2dx'.

 resize2dx:
     width[%]>0,_interpolation,_boundary_conditions,_ax,_ay,_az,_ac

   Resize selected images along the x-axis, preserving 2D ratio.
   (equivalent to shortcut command 'r2dx').

   'interpolation' can be { -1=none (memory content) | 0=none | 1=nearest | 2=average | 3=linear | 4=grid | 5=bicubic | 6=lanczos }.
   'boundary_conditions' has different meanings, according to the chosen 'interpolation' mode :
   . When 'interpolation=={ -1 | 1 | 2 | 4 }', 'boundary_conditions' is meaningless.
   . When 'interpolation==0', 'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }.
   . When 'interpolation=={ 3 | 5 | 6 }', 'boundary_conditions' can be { 0=none | 1=neumann }.
   'ax,ay,az,ac' set the centering along each axis when 'interpolation=0'
   (set to '0' by default, must be defined in range [0,1]).

   Default values: 'interpolation=3', 'boundary_conditions=0' and 'ax=ay=az=ac=0'.

   Example:
     [#1] image.jpg +resize2dx 100,2 append x

 r2dy:
     Shortcut for command 'resize2dy'.

 resize2dy:
     height[%]>=0,_interpolation,_boundary_conditions,_ax,_ay,_az,_ac

   Resize selected images along the y-axis, preserving 2D ratio.
   (equivalent to shortcut command 'r2dy').

   'interpolation' can be { -1=none (memory content) | 0=none | 1=nearest | 2=average | 3=linear | 4=grid | 5=bicubic | 6=lanczos }.
   'boundary_conditions' has different meanings, according to the chosen 'interpolation' mode :
   . When 'interpolation=={ -1 | 1 | 2 | 4 }', 'boundary_conditions' is meaningless.
   . When 'interpolation==0', 'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }.
   . When 'interpolation=={ 3 | 5 | 6 }', 'boundary_conditions' can be { 0=none | 1=neumann }.
   'ax,ay,az,ac' set the centering along each axis when 'interpolation=0'
   (set to '0' by default, must be defined in range [0,1]).

   Default values: 'interpolation=3', 'boundary_conditions=0' and 'ax=ay=az=ac=0'.

   Example:
     [#1] image.jpg +resize2dy 100,2 append x

 r3dx:
     Shortcut for command 'resize3dx'.

 resize3dx:
     width[%]>0,_interpolation,_boundary_conditions,_ax,_ay,_az,_ac

   Resize selected images along the x-axis, preserving 3D ratio.
   (equivalent to shortcut command 'r3dx').

   'interpolation' can be { -1=none (memory content) | 0=none | 1=nearest | 2=average | 3=linear | 4=grid | 5=bicubic | 6=lanczos }.
   'boundary_conditions' has different meanings, according to the chosen 'interpolation' mode :
   . When 'interpolation=={ -1 | 1 | 2 | 4 }', 'boundary_conditions' is meaningless.
   . When 'interpolation==0', 'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }.
   . When 'interpolation=={ 3 | 5 | 6 }', 'boundary_conditions' can be { 0=none | 1=neumann }.
   'ax,ay,az,ac' set the centering along each axis when 'interpolation=0'
   (set to '0' by default, must be defined in range [0,1]).

   Default values: 'interpolation=3', 'boundary_conditions=0' and 'ax=ay=az=ac=0'.

 r3dy:
     Shortcut for command 'resize3dy'.

 resize3dy:
     height[%]>0,_interpolation,_boundary_conditions,_ax,_ay,_az,_ac

   Resize selected images along the y-axis, preserving 3D ratio.
   (equivalent to shortcut command 'r3dy').

   'interpolation' can be { -1=none (memory content) | 0=none | 1=nearest | 2=average | 3=linear | 4=grid | 5=bicubic | 6=lanczos }.
   'boundary_conditions' has different meanings, according to the chosen 'interpolation' mode :
   . When 'interpolation=={ -1 | 1 | 2 | 4 }', 'boundary_conditions' is meaningless.
   . When 'interpolation==0', 'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }.
   . When 'interpolation=={ 3 | 5 | 6 }', 'boundary_conditions' can be { 0=none | 1=neumann }.
   'ax,ay,az,ac' set the centering along each axis when 'interpolation=0'
   (set to '0' by default, must be defined in range [0,1]).

   Default values: 'interpolation=3', 'boundary_conditions=0' and 'ax=ay=az=ac=0'.

 r3dz:
     Shortcut for command 'resize3dz'.

 resize3dz:
     depth[%]>0,_interpolation,_boundary_conditions,_ax,_ay,_az,_ac

   Resize selected images along the z-axis, preserving 3D ratio.
   (equivalent to shortcut command 'r3dz').

   'interpolation' can be { -1=none (memory content) | 0=none | 1=nearest | 2=average | 3=linear | 4=grid | 5=bicubic | 6=lanczos }.
   'boundary_conditions' has different meanings, according to the chosen 'interpolation' mode :
   . When 'interpolation=={ -1 | 1 | 2 | 4 }', 'boundary_conditions' is meaningless.
   . When 'interpolation==0', 'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }.
   . When 'interpolation=={ 3 | 5 | 6 }', 'boundary_conditions' can be { 0=none | 1=neumann }.
   'ax,ay,az,ac' set the centering along each axis when 'interpolation=0'
   (set to '0' by default, must be defined in range [0,1]).

   Default values: 'interpolation=3', 'boundary_conditions=0' and 'ax=ay=az=ac=0'.

 rotate (+):
     angle,_interpolation,_boundary_conditions,_center_x[%],_center_y[%] |
     u,v,w,angle,interpolation,boundary_conditions,_center_x[%],_center_y[%],_center_z[%]

   Rotate selected images with specified angle (in deg.), and optionally 3D axis (u,v,w).
   'interpolation' can be { 0=none | 1=linear | 2=bicubic }.
   'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }.
   When a rotation center (cx,cy,_cz) is specified, the size of the image is preserved.

   Default values: 'interpolation=1', 'boundary_conditions=0' and 'center_x=center_y=(undefined)'.

   Example:
     [#1] image.jpg +rotate -25,1,2,50%,50% rotate[0] 25

 rotate_tileable:
     angle,_max_size_factor>=0

   Rotate selected images by specified angle and make them tileable.
   If resulting size of an image is too big, the image is replaced by a 1x1 image.

   Default values: 'max_size_factor=8'.

 rows (+):
     { [image0] | y0[%] },_{ [image1] | y1[%] }

   Keep only specified rows of selected images.
   Dirichlet boundary conditions are used when specified rows are out of range.

   Example:
     [#1] image.jpg rows -25%,50%

 scale2x:

   Resize selected images using the Scale2x algorithm.

   Example:
     [#1] image.jpg threshold 50% resize 50%,50% +scale2x

 scale3x:

   Resize selected images using the Scale3x algorithm.

   Example:
     [#1] image.jpg threshold 50% resize 33%,33% +scale3x

 scale_dcci2x:
     _edge_threshold>=0,_exponent>0,_extend_1px={ 0=false | 1=true }

   Double image size using directional cubic convolution interpolation,
   as described in https://en.wikipedia.org/wiki/Directional_Cubic_Convolution_Interpolation.

   Default values: 'edge_threshold=1.15', 'exponent=5' and 'extend_1px=0'.

   Example:
     [#1] image.jpg +scale_dcci2x ,

 seamcarve:
     _width[%]>=0,_height[%]>=0,_is_priority_channel={ 0 | 1 },_is_antialiasing={ 0 | 1 },_maximum_seams[%]>=0

   Resize selected images with specified 2D geometry, using the seam-carving algorithm.

   Default values: 'height=100%', 'is_priority_channel=0', 'is_antialiasing=1' and 'maximum_seams=25%'.

   Example:
     [#1] image.jpg seamcarve 60%

 shift (+):
     vx[%],_vy[%],_vz[%],_vc[%],_boundary_conditions,_interpolation={ 0=nearest_neighbor | 1=linear }

   Shift selected images by specified displacement vector.
   Displacement vector can be non-integer in which case linear interpolation should be chosen.
   'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }.

   Default value: 'boundary_conditions=0' and 'interpolation=0'.

   Example:
     [#1] image.jpg +shift[0] 50%,50%,0,0,0 +shift[0] 50%,50%,0,0,1 +shift[0] 50%,50%,0,0,2

 shrink_x:
     size_x>=0

   Shrink selected images along the x-axis.

   Example:
     [#1] image.jpg shrink_x 30

 shrink_xy:
     size>=0

   Shrink selected images along the xy-axes.

   Example:
     [#1] image.jpg shrink_xy 30

 shrink_xyz:
     size>=0

   Shrink selected images along the xyz-axes.

 shrink_y:
     size_y>=0

   Shrink selected images along the y-axis.

   Example:
     [#1] image.jpg shrink_y 30

 shrink_z:
     size_z>=0

   Shrink selected images along the z-axis.

 slices (+):
     { [image0] | z0[%] },_{ [image1] | z1[%] }

   Keep only specified slices of selected images.
   Dirichlet boundary conditions are used when specified slices are out of range.

 sort (+):
     _ordering={ + | - },_axis={ x | y | z | c }

   Sort pixel values of selected images.
   If 'axis' is specified, the sorting is done according to the data of the first column/row/slice/channel
   of selected images.

   Default values: 'ordering=+' and 'axis=(undefined)'.

   Example:
     [#1] 64 rand 0,100 +sort display_graph 400,300,3

 s (+):
     Shortcut for command 'split'.

 split (+):
     { x | y | z | c }...{ x | y | z | c },_split_mode |
     keep_splitting_values={ + | - },_{ x | y | z | c }...{ x | y | z | c },value1,_value2,... |
     (no arg)

   Split selected images along specified axes, or regarding to a sequence of scalar values
   (optionally along specified axes too).
   (equivalent to shortcut command 's').

   'split_mode' can be { 0=split according to constant values | >0=split in N parts | <0=split in parts of size -N }.

   Default value: 'split_mode=-1'.

   Example:
     [#1] image.jpg split c
     [#2] image.jpg split y,3
     [#3] image.jpg split x,-128
     [#4] 1,20,1,1,"1,2,3,4" +split -,2,3 append[1--1] y
     [#5] (1,2,2,3,3,3,4,4,4,4) +split x,0 append[1--1] y

 split_tiles:
     M!=0,_N!=0,_is_homogeneous={ 0 | 1 }

   Split selected images as a MxN array of tiles.
   If M or N is negative, it stands for the tile size instead.

   Default values: 'N=M' and 'is_homogeneous=0'.

   Example:
     [#1] image.jpg +local split_tiles 5,4 blur 3,0 sharpen 700 append_tiles 4,5 endlocal

 undistort:
     -1<=_amplitude<=1,_aspect_ratio,_zoom,_center_x[%],_center_y[%],_boundary_conditions

   Correct barrel/pincushion distortions occurring with wide-angle lens.
   References:
   [1] Zhang Z. (1999). Flexible camera calibration by viewing a plane from unknown orientation.
   [2] Andrew W. Fitzgibbon (2001). Simultaneous linear estimation of multiple view geometry and lens distortion.
   'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }.

   Default values: 'amplitude=0.25', 'aspect_ratio=0', 'zoom=0', 'center_x=center_y=50%' and 'boundary_conditions=0'.

 y (+):
     Shortcut for command 'unroll'.

 unroll (+):
     _axis={ x | y | z | c }

   Unroll selected images along specified axis.
   (equivalent to shortcut command 'y').

   Default value: 'axis=y'.

   Example:
     [#1] (1,2,3;4,5,6;7,8,9) +unroll y

 upscale_smart:
     width[%],_height[%],_depth,_smoothness>=0,_anisotropy=[0,1],sharpening>=0

   Upscale selected images with an edge-preserving algorithm.

   Default values: 'height=100%', 'depth=100%', 'smoothness=2', 'anisotropy=0.4' and 'sharpening=10'.

   Example:
     [#1] image.jpg resize2dy 100 +upscale_smart 500%,500% append x

 warp (+):
     [warping_field],_mode,_interpolation,_boundary_conditions,_nb_frames>0

   Warp selected images with specified displacement field.
   'mode' can be { 0=backward-absolute | 1=backward-relative | 2=forward-absolute | 3=forward-relative }.
   'interpolation' can be { 0=nearest-neighbor | 1=linear | 2=cubic }.
   'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }.

   Default values: 'mode=0', 'interpolation=1', 'boundary_conditions=1' and 'nb_frames=1'.

   Example:
     [#1] image.jpg 100%,100%,1,2,'X=x/w-0.5;Y=y/h-0.5;R=(X*X+Y*Y)^0.5;A=atan2(Y,X);130*R*if(c==0,cos(4*A),sin(8*A))' warp[-2] [-1],1,1,0 quiver[-1] [-1],10,1,1,1,100

   Tutorial: https://gmic.eu/oldtutorial/_warp

 warp_patch:
     [warping_field],patch_width>=1,_patch_height>=1,_patch_depth>=1,_std_factor>0,_boundary_conditions.

   Patch-warp selected images, with specified 2D or 3D displacement field (in backward-absolute mode).
   Argument 'std_factor' sets the std of the gaussian weights for the patch overlap,
   equal to 'std = std_factor*patch_size'.
   'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }.

   Default values: 'std_factor=0.3' and 'boundary_conditions=3'.

 warp_rbf:
     xs0[%],ys0[%],xt0[%],yt0[%],...,xsN[%],ysN[%],xtN[%],ytN[%]

   Warp selected images using RBF-based interpolation.
   Each argument (xsk,ysk)-(xtk,ytk) corresponds to the coordinates of a keypoint
   respectively on the source and target images. The set of all keypoints define the overall image deformation.

   Example:
     [#1] image.jpg +warp_rbf 0,0,0,0,100%,0,100%,0,100%,100%,100%,100%,0,100%,0,100%,50%,50%,70%,50%,25%,25%,25%,75%

 12.8. Filtering
       ---------

 bandpass:
     _min_freq[%],_max_freq[%]

   Apply bandpass filter to selected images.

   Default values: 'min_freq=0' and 'max_freq=20%'.

   Example:
     [#1] image.jpg bandpass 1%,3%

   Tutorial: https://gmic.eu/oldtutorial/_bandpass

 bilateral (+):
     [guide],std_deviation_s[%]>=0,std_deviation_r[%]>=0,_sampling_s>=0,_sampling_r>=0 |
     std_deviation_s[%]>=0,std_deviation_r[%]>=0,_sampling_s>=0,_sampling_r>=0

   Blur selected images by anisotropic (eventually joint/cross) bilateral filtering.
   If a guide image is provided, it is used for drive the smoothing filter.
   A guide image must be of the same xyz-size as the selected images.
   Set 'sampling' arguments to '0' for automatic adjustment.

   Example:
     [#1] image.jpg repeat 5 bilateral 10,10 done

 b (+):
     Shortcut for command 'blur'.

 blur (+):
     std_deviation>=0[%],_boundary_conditions,_kernel |
     axes,std_deviation>=0[%],_boundary_conditions,_kernel

   Blur selected images by a deriche or gaussian filter (recursive implementation).
   (equivalent to shortcut command 'b').

   'boundary_conditions' can be { 0=dirichlet | 1=neumann }.
   'kernel' can be { 0=deriche | 1=gaussian }.
   When specified, argument 'axes' is a sequence of { x | y | z | c }.
   Specifying one axis multiple times apply also the blur multiple times.

   Default values: 'boundary_conditions=1' and 'kernel=1'.

   Example:
     [#1] image.jpg +blur 5,0 +blur[0] 5,1
     [#2] image.jpg +blur y,10%

   Tutorial: https://gmic.eu/oldtutorial/_blur

 blur_angular:
     amplitude[%],_center_x[%],_center_y[%]

   Apply angular blur on selected images.

   Default values: 'center_x=center_y=50%'.

   Example:
     [#1] image.jpg blur_angular 2%

   Tutorial: https://gmic.eu/oldtutorial/_blur_angular

 blur_bloom:
     _amplitude>=0,_ratio>=0,_nb_iter>=0,_blend_operator={ + | max | min },_kernel={ 0=deriche | 1=gaussian | 2=box | 3=triangle | 4=quadratic },_normalize_scales={ 0 | 1 },_axes

   Apply a bloom filter that blend multiple blur filters of different radii,
   resulting in a larger but sharper glare than a simple blur.
   When specified, argument 'axes' is a sequence of { x | y | z | c }.
   Specifying one axis multiple times apply also the blur multiple times.
   Reference: Masaki Kawase, "Practical Implementation of High Dynamic Range Rendering", GDC 2004.

   Default values: 'amplitude=1', 'ratio=2', 'nb_iter=5', 'blend_operator=+', 'kernel=1', 'normalize_scales=0' and 'axes=(all)'

   Example:
     [#1] image.jpg blur_bloom ,

 blur_linear:
     amplitude1[%],_amplitude2[%],_angle,_boundary_conditions={ 0=dirichlet | 1=neumann }

   Apply linear blur on selected images, with specified angle and amplitudes.

   Default values: 'amplitude2=0', 'angle=0' and 'boundary_conditions=1'.

   Example:
     [#1] image.jpg blur_linear 10,0,45

   Tutorial: https://gmic.eu/oldtutorial/_blur_linear

 blur_radial:
     amplitude[%],_center_x[%],_center_y[%]

   Apply radial blur on selected images.

   Default values: 'center_x=center_y=50%'.

   Example:
     [#1] image.jpg blur_radial 2%

   Tutorial: https://gmic.eu/oldtutorial/_blur_radial

 blur_selective:
     sigma>=0,_edges>0,_nb_scales>0

   Blur selected images using selective gaussian scales.

   Default values: 'sigma=5', 'edges=0.5' and 'nb_scales=5'.

   Example:
     [#1] image.jpg noise 20 cut 0,255 +local[-1] repeat 4 blur_selective , done endlocal

   Tutorial: https://gmic.eu/oldtutorial/_blur_selective

 blur_x:
     amplitude[%]>=0,_boundary_conditions={ 0=dirichlet | 1=neumann }

   Blur selected images along the x-axis.

   Default value: 'boundary_conditions=1'.

   Example:
     [#1] image.jpg +blur_x 6

   Tutorial: https://gmic.eu/oldtutorial/_blur_x

 blur_xy:
     amplitude_x[%],amplitude_y[%],_boundary_conditions={ 0=dirichlet | 1=neumann }

   Blur selected images along the X and Y axes.

   Default value: 'boundary_conditions=1'.

   Example:
     [#1] image.jpg +blur_xy 6

   Tutorial: https://gmic.eu/oldtutorial/_blur_y

 blur_xyz:
     amplitude_x[%],amplitude_y[%],amplitude_z,_boundary_conditions={ 0=dirichlet | 1=neumann }

   Blur selected images along the X, Y and Z axes.

   Default value: 'boundary_conditions=1'.

   Tutorial: https://gmic.eu/oldtutorial/_blur_xyz

 blur_y:
     amplitude[%]>=0,_boundary_conditions={ 0=dirichlet | 1=neumann }

   Blur selected images along the y-axis.

   Default value: 'boundary_conditions=1'.

   Example:
     [#1] image.jpg +blur_y 6

   Tutorial: https://gmic.eu/oldtutorial/_blur_y

 blur_z:
     amplitude[%]>=0,_boundary_conditions={ 0=dirichlet | 1=neumann }

   Blur selected images along the z-axis.

   Default value: 'boundary_conditions=1'.

   Tutorial: https://gmic.eu/oldtutorial/_blur_z

 boxfilter (+):
     size>=0[%],_order,_boundary_conditions,_nb_iter>=0 |
     axes,size>=0[%],_order,_boundary_conditions,_nb_iter>=0

   Blur selected images by a box filter of specified size (fast recursive implementation).
   'order' can be { 0=smooth | 1=1st-derivative | 2=2nd-derivative }.
   'boundary_conditions' can be { 0=dirichlet | 1=neumann }.
   When specified, argument 'axes' is a sequence of { x | y | z | c }.
   Specifying one axis multiple times apply also the blur multiple times.

   Default values: 'order=0', 'boundary_conditions=1' and 'nb_iter=1'.

   Example:
     [#1] image.jpg +boxfilter 5%
     [#2] image.jpg +boxfilter y,3,1

 bump2normal:

   Convert selected bumpmaps to normalmaps.

   Example:
     [#1] 300,300 circle 50%,50%,128,1,1 blur 5% bump2normal

 compose_freq:

   Compose selected low and high frequency parts into new images.

   Example:
     [#1] image.jpg split_freq 2% mirror[-1] x compose_freq

 convolve (+):
     [mask],_boundary_conditions,_is_normalized={ 0 | 1 },_channel_mode,_xcenter,_ycenter,_zcenter,_xstart,_ystart,_zstart,_xend,_yend,_zend,_xstride,_ystride,_zstride,_xdilation,_ydilation,_zdilation,
       interpolation_type

   Convolve selected images by specified mask.
   'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }.
   'channel_mode' can be { 0=sum input channels | 1=one-for-one | 2=expand }.
   'interpolation_type' can be { 0=nearest-neighbor | 1=linear }.

   Default values: 'boundary_conditions=1', 'is_normalized=0', 'channel_mode=1', 'xcenter=ycenter=zcenter=-1' (-1=centered), 'xstart=ystart=zstart=0',
    'xend=yend=zend=-1' (-1=max coordinates), 'xstride=ystride=zstride=1', 'xdilation=ydilation=zdilation=1' and 'interpolation_type=0'.

   Example:
     [#1] image.jpg (0,1,0;1,-4,1;0,1,0) convolve[-2] [-1] keep[-2]
     [#2] image.jpg (0,1,0) resize[-1] 130,1,1,1,3 +convolve[0] [1]

   Tutorial: https://gmic.eu/oldtutorial/_convolve

 convolve_fft:
     [mask],_boundary_conditions

   Convolve selected images with specified mask, in the fourier domain.
   'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }.

   Example:
     [#1] image.jpg 100%,100% gaussian[-1] 20,1,45 +convolve_fft[0] [1]

 correlate (+):
     [mask],_boundary_conditions,_is_normalized={ 0 | 1 },_channel_mode,_xcenter,_ycenter,_zcenter,_xstart,_ystart,_zstart,_xend,_yend,_zend,_xstride,_ystride,_zstride,_xdilation,_ydilation,_zdilation,
       interpolation_type

   Correlate selected images by specified mask.
   'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }.
   'channel_mode' can be { 0=sum input channels | 1=one-for-one | 2=expand }.
   'interpolation_type' can be { 0=nearest-neighbor | 1=linear }.

   Default values: 'boundary_conditions=1', 'is_normalized=0', 'channel_mode=1', 'xcenter=ycenter=zcenter=-1' (-1=centered), 'xstart=ystart=zstart=0',
    'xend=yend=zend=-1' (-1=max coordinates), 'xstride=ystride=zstride=1', 'xdilation=ydilation=zdilation=1' and 'interpolation_type=0'.

   Example:
     [#1] image.jpg (0,1,0;1,-4,1;0,1,0) correlate[-2] [-1] keep[-2]
     [#2] image.jpg +crop 40%,40%,60%,60% +correlate[0] [-1],0,1

 cross_correlation:
     [mask]

   Compute cross-correlation of selected images with specified mask.

   Example:
     [#1] image.jpg +shift -30,-20 +cross_correlation[0] [1]

 curvature:

   Compute isophote curvatures on selected images.

   Example:
     [#1] image.jpg blur 10 curvature

 dct:
     _{ x | y | z }...{ x | y | z } |
     (no arg)

   Compute the discrete cosine transform of selected images, optionally along the specified axes only.
   Output images are always evenly sized, so this command may change the size of the selected images.

   Default values: (no arg)
   See also: idct.

   Example:
     [#1] image.jpg +dct +idct[-1] abs[-2] +[-2] 1 log[-2]

   Tutorial: https://gmic.eu/oldtutorial/_dct-and-idct

 deblur:
     amplitude[%]>=0,_nb_iter>=0,_dt>=0,_regul>=0,_regul_type={ 0=Tikhonov | 1=meancurv. | 2=TV }

   Deblur image using a regularized Jansson-Van Cittert algorithm.

   Default values: 'nb_iter=10', 'dt=20', 'regul=0.7' and 'regul_type=1'.

   Example:
     [#1] image.jpg blur 3 +deblur 3,40,20,0.01

 deblur_goldmeinel:
     sigma>=0,_nb_iter>=0,_acceleration>=0,_kernel_type={ 0=deriche | 1=gaussian }.

   Deblur selected images using Gold-Meinel algorithm

   Default values: 'nb_iter=8', 'acceleration=1' and 'kernel_type=1'.

   Example:
     [#1] image.jpg +blur 1 +deblur_goldmeinel[-1] 1

 deblur_richardsonlucy:
     sigma>=0, nb_iter>=0, _kernel_type={ 0=deriche | 1=gaussian }.

   Deblur selected images using Richardson-Lucy algorithm.

   Default values: 'nb_iter=50' and 'kernel_type=1'.

   Example:
     [#1] image.jpg +blur 1 +deblur_richardsonlucy[-1] 1

 deconvolve_fft:
     [kernel],_regularization>=0

   Deconvolve selected images by specified mask in the fourier space.

   Default value: 'regularization>=0'.

   Example:
     [#1] image.jpg +gaussian 5 +convolve_fft[0] [1] +deconvolve_fft[-1] [1]

 deinterlace:
     _method={ 0 | 1 }

   Deinterlace selected images ('method' can be { 0=standard or 1=motion-compensated }).

   Default value: 'method=0'.

   Example:
     [#1] image.jpg +rotate 3,1,1,50%,50% resize 100%,50% resize 100%,200%,1,3,4 shift[-1] 0,1 add +deinterlace 1

 denoise (+):
     [guide],std_deviation_s[%]>=0,_std_deviation_r[%]>=0,_patch_size>0,_lookup_size>0,_smoothness,_fast_approx={ 0 | 1 } |
     std_deviation_s[%]>=0,_std_deviation_r[%]>=0,_patch_size>0,_lookup_size>0,_smoothness,_fast_approx={ 0 | 1 }

   Denoise selected images by non-local patch averaging.

   Default values: 'std_deviation_p=10', 'patch_size=5', 'lookup_size=6' and 'smoothness=1'.

   Example:
     [#1] image.jpg +denoise 5,5,8

 denoise_haar:
     _threshold>=0,_nb_scales>=0,_cycle_spinning>0

   Denoise selected images using haar-wavelet thresholding with cycle spinning.
   Set 'nb_scales==0' to automatically determine the optimal number of scales.

   Default values: 'threshold=1.4', 'nb_scale=0' and 'cycle_spinning=10'.

   Example:
     [#1] image.jpg noise 20 cut 0,255 +denoise_haar[-1] 0.8

 denoise_patchpca:
     _strength>=0,_patch_size>0,_lookup_size>0,_spatial_sampling>0

   Denoise selected images using the patch-pca algorithm.

   Default values: 'patch_size=7', 'lookup_size=11', 'details=1.8' and 'spatial_sampling=5'.

   Example:
     [#1] image.jpg +noise 20 cut[-1] 0,255 +denoise_patchpca[-1] ,

 deriche (+):
     std_deviation>=0[%],order={ 0 | 1 | 2 },axis={ x | y | z | c },_boundary_conditions

   Apply Deriche recursive filter on selected images, along specified axis and with
   specified standard deviation, order and boundary conditions.
   'boundary_conditions' can be { 0=dirichlet | 1=neumann }.

   Default value: 'boundary_conditions=1'.

   Example:
     [#1] image.jpg deriche 3,1,x
     [#2] image.jpg +deriche 30,0,x deriche[-2] 30,0,y add

   Tutorial: https://gmic.eu/oldtutorial/_deriche

 dilate (+):
     size>=0 |
     size_x>=0,size_y>=0,size_z>=0 |
     [kernel],_boundary_conditions,_is_real={ 0=binary-mode | 1=real-mode }

   Dilate selected images by a rectangular or the specified structuring element.
   'boundary_conditions' can be { 0=dirichlet | 1=neumann }.

   Default values: 'size_z=1', 'boundary_conditions=1' and 'is_real=0'.

   Example:
     [#1] image.jpg +dilate 10

 dilate_circ:
     _size>=0,_boundary_conditions,_is_normalized={ 0 | 1 }

   Apply circular dilation of selected images by specified size.

   Default values: 'boundary_conditions=1' and 'is_normalized=0'.

   Example:
     [#1] image.jpg +dilate_circ 7

 dilate_oct:
     _size>=0,_boundary_conditions,_is_normalized={ 0 | 1 }

   Apply octagonal dilation of selected images by specified size.

   Default values: 'boundary_conditions=1' and 'is_normalized=0'.

   Example:
     [#1] image.jpg +dilate_oct 7

 dilate_threshold:
     size_x>=1,size_y>=1,size_z>=1,_threshold>=0,_boundary_conditions

   Dilate selected images in the (X,Y,Z,I) space.
   'boundary_conditions' can be { 0=dirichlet | 1=neumann }.

   Default values: 'size_y=size_x', 'size_z=1', 'threshold=255' and 'boundary_conditions=1'.

 divergence:

   Compute divergence of selected vector fields.

   Example:
     [#1] image.jpg luminance +gradient append[-2,-1] c divergence[-1]

 dog:
     _sigma1>=0[%],_sigma2>=0[%]

   Compute difference of gaussian on selected images.

   Default values: 'sigma1=2%' and 'sigma2=3%'.

   Example:
     [#1] image.jpg dog 2,3

 diffusiontensors:
     _sharpness>=0,0<=_anisotropy<=1,_alpha[%],_sigma[%],is_sqrt={ 0 | 1 }

   Compute the diffusion tensors of selected images for edge-preserving smoothing algorithms.

   Default values: 'sharpness=0.7', 'anisotropy=0.3', 'alpha=0.6', 'sigma=1.1' and 'is_sqrt=0'.

   Example:
     [#1] image.jpg diffusiontensors 0.8 abs pow 0.2

   Tutorial: https://gmic.eu/oldtutorial/_diffusiontensors

 edges:
     _threshold[%]>=0

   Estimate contours of selected images.

   Default value: 'edges=15%'

   Example:
     [#1] image.jpg +edges 15%

 erode (+):
     size>=0 |
     size_x>=0,size_y>=0,_size_z>=0 |
     [kernel],_boundary_conditions,_is_real={ 0=binary-mode | 1=real-mode }

   Erode selected images by a rectangular or the specified structuring element.
   'boundary_conditions' can be { 0=dirichlet | 1=neumann }.

   Default values: 'size_z=1', 'boundary_conditions=1' and 'is_real=0'.

   Example:
     [#1] image.jpg +erode 10

 erode_circ:
     _size>=0,_boundary_conditions,_is_normalized={ 0 | 1 }

   Apply circular erosion of selected images by specified size.

   Default values: 'boundary_conditions=1' and 'is_normalized=0'.

   Example:
     [#1] image.jpg +erode_circ 7

 erode_oct:
     _size>=0,_boundary_conditions,_is_normalized={ 0 | 1 }

   Apply octagonal erosion of selected images by specified size.

   Default values: 'boundary_conditions=1' and 'is_normalized=0'.

   Example:
     [#1] image.jpg +erode_oct 7

 erode_threshold:
     size_x>=1,size_y>=1,size_z>=1,_threshold>=0,_boundary_conditions

   Erode selected images in the (X,Y,Z,I) space.
   'boundary_conditions' can be { 0=dirichlet | 1=neumann }.

   Default values: 'size_y=size_x', 'size_z=1', 'threshold=255' and 'boundary_conditions=1'.

 fft (+):
     _{ x | y | z }...{ x | y | z }

   Compute the direct fourier transform (real and imaginary parts) of selected images,
   optionally along the specified axes only.
   See also: ifft.

   Example:
     [#1] image.jpg luminance +fft append[-2,-1] c norm[-1] log[-1] shift[-1] 50%,50%,0,0,2
     [#2] image.jpg w2={int(w/2)} h2={int(h/2)} fft shift $w2,$h2,0,0,2 ellipse $w2,$h2,30,30,0,1,0 shift -$w2,-$h2,0,0,2 ifft remove[-1]

   Tutorial: https://gmic.eu/oldtutorial/_fft

 g (+):
     Shortcut for command 'gradient'.

 gradient (+):
     { x | y | z }...{ x | y | z },_scheme |
     (no arg)

   Compute the gradient components (first derivatives) of selected images.
   (equivalent to shortcut command 'g').

   'scheme' can be { -1=backward | 0=centered | 1=forward | 2=sobel | 3=rotation-invariant (default) | 4=deriche | 5=vanvliet }.
   (no arg) compute all significant components.

   Default value: 'scheme=0'.

   Example:
     [#1] image.jpg gradient

   Tutorial: https://gmic.eu/oldtutorial/_gradient

 gradient_norm:

   Compute gradient norm of selected images.

   Example:
     [#1] image.jpg gradient_norm equalize

   Tutorial: https://gmic.eu/oldtutorial/_gradient_norm

 gradient_orientation:
     _dimension={1,2,3}

   Compute N-d gradient orientation of selected images.

   Default value: 'dimension=3'.

   Example:
     [#1] image.jpg +gradient_orientation 2

 guided (+):
     [guide],radius[%]>=0,regularization[%]>=0 |
     radius[%]>=0,regularization[%]>=0

   Blur selected images by guided image filtering.
   If a guide image is provided, it is used to drive the smoothing process.
   A guide image must be of the same xyz-size as the selected images.
   This command implements the filtering algorithm described in:
   He, Kaiming; Sun, Jian; Tang, Xiaoou, "Guided Image Filtering",
   IEEE Transactions on Pattern Analysis and Machine Intelligence, vol.35, no.6, pp.1397,1409, June 2013

   Example:
     [#1] image.jpg +guided 5,400

 haar:
     scale>0

   Compute the direct haar multiscale wavelet transform of selected images.
   See also: ihaar.

   Tutorial: https://gmic.eu/oldtutorial/_haar

 heat_flow:
     _nb_iter>=0,_dt,_keep_sequence={ 0 | 1 }

   Apply iterations of the heat flow on selected images.

   Default values: 'nb_iter=10', 'dt=30' and 'keep_sequence=0'.

   Example:
     [#1] image.jpg +heat_flow 20

 hessian (+):
     { xx | xy | xz | yy | yz | zz }...{ xx | xy | xz | yy | yz | zz } |
     (no arg)

   Compute the hessian components (second derivatives) of selected images.
   (no arg) compute all significant components.

   Example:
     [#1] image.jpg hessian

 idct:
     _{ x | y | z }...{ x | y | z } |
     (no arg)

   Compute the inverse discrete cosine transform of selected images, optionally along the specified axes only.
   Output images are always evenly sized, so this command may change the size of the selected images.
   (dct images obtained with the 'dct' command are evenly sized anyway).

   Default values: (no arg)
   See also: dct.

   Tutorial: https://gmic.eu/oldtutorial/_dct-and-idct

 iee:

   Compute gradient-orthogonal-directed 2nd derivative of image(s).

   Example:
     [#1] image.jpg iee

 ifft (+):
     _{ x | y | z }...{ x | y | z }

   Compute the inverse fourier transform (real and imaginary parts) of selected images.
   optionally along the specified axes only.
   See also: fft.

   Tutorial: https://gmic.eu/oldtutorial/_fft

 ihaar:
     scale>0

   Compute the inverse haar multiscale wavelet transform of selected images.
   See also: haar.

   Tutorial: https://gmic.eu/oldtutorial/_haar

 ilaplacian:
     { nb_iterations>0 | 0 },_[initial_estimate]

   Invert selected Laplacian images.
   If given 'nb_iterations' is '0', inversion is done in Fourier space (single iteration),
   otherwise, by applying 'nb_iterations' of a Laplacian-inversion PDE flow.
   Note that the resulting inversions are just estimation of possible/approximated solutions.

   Default values: 'nb_iterations=0' and '[initial_estimated]=(undefined)'.

   Example:
     [#1] image.jpg +laplacian +ilaplacian[-1] 0

 inn:

   Compute gradient-directed 2nd derivative of image(s).

   Example:
     [#1] image.jpg inn

 inpaint (+):
     [mask] |
     [mask],0,_fast_method |
     [mask],_patch_size>=1,_lookup_size>=1,_lookup_factor>=0,_lookup_increment!=0,_blend_size>=0,0<=_blend_threshold<=1,_blend_decay>=0,_blend_scales>=1,_is_blend_outer={ 0 | 1 }

   Inpaint selected images by specified mask.
   If no patch size (or 0) is specified, inpainting is done using a fast average or median algorithm.
   Otherwise, it used a patch-based reconstruction method, that can be very time consuming.
   'fast_method' can be { 0=low-connectivity average | 1=high-connectivity average | 2=low-connectivity median | 3=high-connectivity median }.

   Default values: 'patch_size=0', 'fast_method=1', 'lookup_size=22', 'lookup_factor=0.5', 'lookup_increment=1', 'blend_size=0', 'blend_threshold=0',
    'blend_decay=0.05', 'blend_scales=10' and 'is_blend_outer=1'.

   Example:
     [#1] image.jpg 100%,100% ellipse 50%,50%,30,30,0,1,255 ellipse 20%,20%,30,10,0,1,255 +inpaint[-2] [-1] remove[-2]
     [#2] image.jpg 100%,100% circle 30%,30%,30,1,255,0,255 circle 70%,70%,50,1,255,0,255 +inpaint[0] [1],5,15,0.5,1,9,0 remove[1]

 inpaint_pde:
     [mask],_nb_scales[%]>=0,_diffusion_type={ 0=isotropic | 1=delaunay-guided | 2=edge-guided | 3=mask-guided },_diffusion_iter>=0

   Inpaint selected images by specified mask using a multiscale transport-diffusion algorithm.
   If 'diffusion type==3', non-zero values of the mask (e.g. a distance function) are used
   to guide the diffusion process.

   Default values: 'nb_scales=75%', 'diffusion_type=1' and 'diffusion_iter=20'.

   Example:
     [#1] image.jpg 100%,100% ellipse[-1] 30%,30%,40,30,0,1,255 +inpaint_pde[0] [1]

 inpaint_flow:
     [mask],_nb_global_iter>=0,_nb_local_iter>=0,_dt>0,_alpha>=0,_sigma>=0

   Apply iteration of the inpainting flow on selected images.

   Default values: 'nb_global_iter=10', 'nb_local_iter=100', 'dt=5', 'alpha=1' and 'sigma=3'.

   Example:
     [#1] image.jpg 100%,100% ellipse[-1] 30%,30%,40,30,0,1,255 inpaint_flow[0] [1]

 inpaint_holes:
     maximal_area[%]>=0,_tolerance>=0,_is_high_connectivity={ 0 | 1 }

   Inpaint all connected regions having an area less than specified value.

   Default values: 'maximal_area=4', 'tolerance=0' and 'is_high_connectivity=0'.

   Example:
     [#1] image.jpg noise 5%,2 +inpaint_holes 8,40

 inpaint_morpho:
     [mask]

   Inpaint selected images by specified mask using morphological operators.

   Example:
     [#1] image.jpg 100%,100% ellipse[-1] 30%,30%,40,30,0,1,255 +inpaint_morpho[0] [1]

 inpaint_matchpatch:
     [mask],_nb_scales={ 0=auto | >0 },_patch_size>0,_nb_iterations_per_scale>0,_blend_size>=0,_allow_outer_blending={ 0 | 1 },_is_already_initialized={ 0 | 1 }

   Inpaint selected images by specified binary mask, using a multi-scale matchpatch algorithm.

   Default values: 'nb_scales=0', 'patch_size=9', 'nb_iterations_per_scale=10', 'blend_size=5','allow_outer_blending=1' and 'is_already_initialized=0'.

   Example:
     [#1] image.jpg 100%,100% ellipse[-1] 30%,30%,40,30,0,1,255 +inpaint_matchpatch[0] [1]

 kuwahara:
     size>0

   Apply Kuwahara filter of specified size on selected images.

   Example:
     [#1] image.jpg kuwahara 9

 laplacian:

   Compute Laplacian of selected images.

   Example:
     [#1] image.jpg laplacian

 lic:
     _amplitude>0,_channels>0

   Render LIC representation of selected vector fields.

   Default values: 'amplitude=30' and 'channels=1'.

   Example:
     [#1] 400,400,1,2,'if(c==0,x-w/2,y-h/2)' +lic 200,3 quiver[-2] [-2],10,1,1,1,255

 map_tones:
     _threshold>=0,_gamma>=0,_smoothness>=0,nb_iter>=0

   Apply tone mapping operator on selected images, based on Poisson equation.

   Default values: 'threshold=0.1', 'gamma=0.8', 'smoothness=0.5' and 'nb_iter=30'.

   Example:
     [#1] image.jpg +map_tones ,

 map_tones_fast:
     _radius[%]>=0,_power>=0

   Apply fast tone mapping operator on selected images.

   Default values: 'radius=3%' and 'power=0.3'.

   Example:
     [#1] image.jpg +map_tones_fast ,

 meancurvature_flow:
     _nb_iter>=0,_dt,_keep_sequence={ 0 | 1 }

   Apply iterations of the mean curvature flow on selected images.

   Default values: 'nb_iter=10', 'dt=30' and 'keep_sequence=0'.

   Example:
     [#1] image.jpg +meancurvature_flow 20

 median (+):
     size>=0,_threshold>0

   Apply (opt. thresholded) median filter on selected images with structuring element size x size.

   Example:
     [#1] image.jpg +median 5

 nlmeans:
     [guide],_patch_radius>0,_spatial_bandwidth>0,_tonal_bandwidth>0,_patch_measure_command |
     _patch_radius>0,_spatial_bandwidth>0,_tonal_bandwidth>0,_patch_measure_command

   Apply non local means denoising of Buades et al, 2005. on selected images.
   The patch is a gaussian function of 'std_patch_radius'.
   The spatial kernel is a rectangle of radius 'spatial_bandwidth'.
   The tonal kernel is exponential ('exp(-d^2/_tonal_bandwidth^2)')
   with 'd' the euclidean distance between image patches.

   Default values: 'patch_radius=4', 'spatial_bandwidth=4', 'tonal_bandwidth=10' and 'patch_measure_command=-norm'.

   Example:
     [#1] image.jpg +noise 10 nlmeans[-1] 4,4,{0.6*${-std_noise}}

 nlmeans_core:
     _reference_image,_scaling_map,_patch_radius>0,_spatial_bandwidth>0

   Apply non local means denoising using a image for weight and a map for scaling

 normalize_local:
     _amplitude>=0,_radius>0,_n_smooth>=0[%],_a_smooth>=0[%],_is_cut={ 0 | 1 },_min=0,_max=255

   Normalize selected images locally.

   Default values: 'amplitude=3', 'radius=16', 'n_smooth=4%', 'a_smooth=2%', 'is_cut=1', 'min=0' and 'max=255'.

   Example:
     [#1] image.jpg normalize_local 8,10

 normalized_cross_correlation:
     [mask]

   Compute normalized cross-correlation of selected images with specified mask.

   Example:
     [#1] image.jpg +shift -30,-20 +normalized_cross_correlation[0] [1]

 percentile:
     [mask],0<=_min_percentile[%]<=100,0<=_max_percentile[%]<=100.

   Apply percentile averaging filter to selected images.

   Default values: 'min_percentile=0' and 'max_percentile=100'.

   Example:
     [#1] image.jpg shape_circle 11,11 +percentile[0] [1],25,75

 peronamalik_flow:
     K_factor>0,_nb_iter>=0,_dt,_keep_sequence={ 0 | 1 }

   Apply iterations of the Perona-Malik flow on selected images.

   Default values: 'K_factor=20', 'nb_iter=5', 'dt=5' and 'keep_sequence=0'.

   Example:
     [#1] image.jpg +heat_flow 20

 phase_correlation:
     [destination]

   Estimate translation vector between selected source images and specified destination.

   Example:
     [#1] image.jpg +shift -30,-20 +phase_correlation[0] [1] unroll[-1] y

 pde_flow:
     _nb_iter>=0,_dt,_velocity_command,_keep_sequence={ 0 | 1 }

   Apply iterations of a generic PDE flow on selected images.

   Default values: 'nb_iter=10', 'dt=30', 'velocity_command=laplacian' and 'keep_sequence=0'.

   Example:
     [#1] image.jpg +pde_flow 20

 periodize_poisson:

   Periodize selected images using a Poisson solver in Fourier space.

   Example:
     [#1] image.jpg +periodize_poisson array 2,2,2

 rbf:
     dx,_x0,_x1,_phi(r) |
     dx,dy,_x0,_y0,_x1,_y1,_phi(r) |
     dx,dy,dz,x0,y0,z0,x1,y1,z1,phi(r)

   Reconstruct 1D/2D or 3D image from selected sets of keypoints, by RBF-interpolation.
   A set of keypoints is represented by a vector-valued image, where each pixel represents a single keypoint.
   Vector components of a keypoint have the following meaning:
      -  For 1D reconstruction: [ x_k, f1(k),...fN(k) ].
      -  For 2D reconstruction: [ x_k,y_k, f1(k),...,fN(k) ].
      -  For 3D reconstruction: [ x_k,y_k,z_k, f1(k),...,fN(k) ].
   Values 'x_k','y_k' and 'z_k' are the spatial coordinates of keypoint 'k'.
   Values 'f1(k),..,fN(k)' are the 'N' components of the vector value of keypoint 'k'.
   The command reconstructs an image with specified size 'dx'x'dy'x'dz', with 'N' channels.

   Default values: 'x0=y0=z0=0', 'x1=dx-1', 'y1=dy-1', 'z1=dz-1', 'phi(r)=r^2*log(1e-5+r)'.

   Example:
     [#1] sp colorful r2dx 400 100%,100% noise_poissondisk. 10 1,{is},1,5 eval[-2] "begin(p=0);i?(I[#-1,p++]=[x,y,I(#0)])" to_rgb[1] mul[0,1] dilate_circ[0] 5 +rbf[-1] {0,[w,h]} c[-1] 0,255
     [#2] 32,1,1,5,u([400,400,255,255,255]) rbf 400,400 c 0,255

 red_eye:
     0<=_threshold<=100,_smoothness>=0,0<=attenuation<=1

   Attenuate red-eye effect in selected images.

   Default values: 'threshold=75', 'smoothness=3.5' and 'attenuation=0.1'.

   Example:
     [#1] image.jpg +red_eye ,

 remove_hotpixels:
     _mask_size>0, _threshold[%]>0

   Remove hot pixels in selected images.

   Default values: 'mask_size=3' and 'threshold=10%'.

   Example:
     [#1] image.jpg noise 10,2 +remove_hotpixels ,

 remove_pixels:
     number_of_pixels[%]>=0

   Remove specified number of pixels (i.e. set them to 0) from the set of non-zero pixels in selected images.

   Example:
     [#1] image.jpg +remove_pixels 50%

 rolling_guidance:
     std_deviation_s[%]>=0,std_deviation_r[%]>=0,_precision>=0

   Apply the rolling guidance filter on selected image.
   Rolling guidance filter is a fast image abstraction filter, described in:
   "Rolling Guidance Filter", Qi Zhang Xiaoyong, Shen Li, Xu Jiaya Jia, ECCV'2014.

   Default values: 'std_deviation_s=4', 'std_deviation_r=10' and 'precision=0.5'.

   Example:
     [#1] image.jpg +rolling_guidance , +-

 sharpen (+):
     amplitude>=0 |
     amplitude>=0,edge>=0,_alpha,_sigma

   Sharpen selected images by inverse diffusion or shock filters methods.
   'edge' must be specified to enable shock-filter method.

   Default values: 'alpha=0' and 'sigma=0'.

   Example:
     [#1] image.jpg sharpen 300
     [#2] image.jpg blur 5 sharpen 300,1

 smooth (+):
     amplitude[%]>=0,_sharpness>=0,0<=_anisotropy<=1,_alpha[%],_sigma[%],_dl>0,_da>0,_precision>0,_interpolation,_fast_approx={ 0 | 1 } |
     nb_iterations>=0,_sharpness>=0,_anisotropy,_alpha,_sigma,_dt>0,0 |
     [tensor_field],_amplitude>=0,_dl>0,_da>0,_precision>0,_interpolation,_fast_approx={ 0 | 1 } |
     [tensor_field],_nb_iters>=0,_dt>0,0

   Smooth selected images anisotropically using diffusion PDE's, with specified field of
   diffusion tensors.
   'interpolation' can be { 0=nearest | 1=linear | 2=runge-kutta }.

   Default values: 'sharpness=0.7', 'anisotropy=0.3', 'alpha=0.6', 'sigma=1.1', 'dl=0.8', 'da=30', 'precision=2', 'interpolation=0' and
    'fast_approx=1'.

   Example:
     [#1] image.jpg repeat 3 smooth 40,0,1,1,2 done
     [#2] image.jpg 100%,100%,1,2 rand[-1] -100,100 repeat 2 smooth[-1] 100,0.2,1,4,4 done warp[0] [-1],1,1

   Tutorial: https://gmic.eu/oldtutorial/_smooth

 split_freq:
     smoothness>0[%]

   Split selected images into low and high frequency parts.

   Example:
     [#1] image.jpg split_freq 2%

 solve_poisson:
     "laplacian_command",_nb_iterations>=0,_time_step>0,_nb_scales>=0

   Solve Poisson equation so that applying 'laplacian[n]' is close to the result of 'laplacian_command[n]'.
   Solving is performed using a multi-scale gradient descent algorithm.
   If 'nb_scales=0', the number of scales is automatically determined.

   Default values: 'nb_iterations=60', 'dt=5' and 'nb_scales=0'.

   Example:
     [#1] image.jpg command "foo : gradient x" +solve_poisson foo +foo[0] +laplacian[1]

 split_details:
     _nb_scales>0,_base_scale[%]>=0,_detail_scale[%]>=0

   Split selected images into 'nb_scales' detail scales.
   If 'base_scale''detail_scale'0, the image decomposition is done with 'a trous' wavelets.
   Otherwise, it uses laplacian pyramids with linear standard deviations.

   Default values: 'nb_scales=4', 'base_scale=0' and 'detail_scale=0'.

   Example:
     [#1] image.jpg split_details ,

 structuretensors (+):
     _scheme={ 0=centered | 1=forward/backward }

   Compute the structure tensor field of selected images.

   Default value: 'scheme=1'.

   Example:
     [#1] image.jpg structuretensors abs pow 0.2

   Tutorial: https://gmic.eu/oldtutorial/_structuretensors

 solidify:
     _smoothness[%]>=0,_diffusion_type={ 0=isotropic | 1=delaunay-oriented | 2=edge-oriented },_diffusion_iter>=0

   Solidify selected transparent images.

   Default values: 'smoothness=75%', 'diffusion_type=1' and 'diffusion_iter=20'.

   Example:
     [#1] image.jpg 100%,100% circle[-1] 50%,50%,25%,1,255 append c +solidify , display_rgba

 syntexturize:
     _width[%]>0,_height[%]>0

   Resynthetize 'width'x'height' versions of selected micro-textures by phase randomization.
   The texture synthesis algorithm is a straightforward implementation of the method described in :
   http://www.ipol.im/pub/art/2011/ggm_rpn/.

   Default values: 'width=height=100%'.

   Example:
     [#1] image.jpg crop 2,282,50,328 +syntexturize 320,320

 syntexturize_matchpatch:
     _width[%]>0,_height[%]>0,_nb_scales>=0,_patch_size>0,_blending_size>=0,_precision>=0

   Resynthetize 'width'x'height' versions of selected micro-textures using a patch-matching algorithm.
   If 'nbscales==0', the number of scales used is estimated from the image size.

   Default values: 'width=height=100%', 'nb_scales=0', 'patch_size=7', 'blending_size=5' and 'precision=1'.

   Example:
     [#1] image.jpg crop 25%,25%,75%,75% syntexturize_matchpatch 512,512

 tv_flow:
     _nb_iter>=0,_dt,_keep_sequence={ 0 | 1 }

   Apply iterations of the total variation flow on selected images.

   Default values: 'nb_iter=10', 'dt=30' and 'keep_sequence=0'.

   Example:
     [#1] image.jpg +tv_flow 40

 unsharp:
     radius[%]>=0,_amount>=0,_threshold[%]>=0

   Apply unsharp mask on selected images.

   Default values: 'amount=2' and 'threshold=0'.

   Example:
     [#1] image.jpg blur 3 +unsharp 1.5,15 cut 0,255

 unsharp_octave:
     _nb_scales>0,_radius[%]>=0,_amount>=0,threshold[%]>=0

   Apply octave sharpening on selected images.

   Default values: 'nb_scales=4', 'radius=1', 'amount=2' and 'threshold=0'.

   Example:
     [#1] image.jpg blur 3 +unsharp_octave 4,5,15 cut 0,255

 vanvliet (+):
     std_deviation>=0[%],order={ 0 | 1 | 2 | 3 },axis={ x | y | z | c },_boundary_conditions

   Apply Vanvliet recursive filter on selected images, along specified axis and with
   specified standard deviation, order and boundary conditions.
   'boundary_conditions' can be { 0=dirichlet | 1=neumann }.

   Default value: 'boundary_conditions=1'.

   Example:
     [#1] image.jpg +vanvliet 3,1,x
     [#2] image.jpg +vanvliet 30,0,x vanvliet[-2] 30,0,y add

 voronoi:

   Compute the discrete Voronoi diagram of non-zero pixels in selected images.

   Example:
     [#1] 400,400 noise 0.2,2 eq 1 +label_fg 0 voronoi[-1] +gradient[-1] xy,1 append[-2,-1] c norm[-1] ==[-1] 0 map[-2] 2,2 mul[-2,-1] normalize[-2] 0,255 dilate_circ[-2] 4 reverse max

 watermark_fourier:
     text,_size>0

   Add a textual watermark in the frequency domain of selected images.

   Default value: 'size=33'.

   Example:
     [#1] image.jpg +watermark_fourier "Watermarked!" +display_fft remove[-3,-1] normalize 0,255 append[-4,-2] y append[-2,-1] y

 watershed (+):
     [priority_image],_is_high_connectivity={ 0 | 1 }

   Compute the watershed transform of selected images.

   Default value: 'is_high_connectivity=1'.

   Example:
     [#1] 400,400 noise 0.2,2 eq 1 +distance 1 mul[-1] -1 label[-2] watershed[-2] [-1] mod[-2] 256 map[-2] 0 reverse

 12.9. Features Extraction
       -------------------

 area:
     tolerance>=0,is_high_connectivity={ 0 | 1 }

   Compute area of connected components in selected images.

   Default values: 'is_high_connectivity=0'.

   Example:
     [#1] image.jpg luminance stencil[-1] 1 +area 0

   Tutorial: https://gmic.eu/oldtutorial/_area

 area_fg:
     tolerance>=0,is_high_connectivity={ 0 | 1 }

   Compute area of connected components for non-zero values in selected images.
   Similar to 'area' except that 0-valued pixels are not considered.

   Default values: 'is_high_connectivity=0'.

   Example:
     [#1] image.jpg luminance stencil[-1] 1 +area_fg 0

 at_line:
     x0[%],y0[%],z0[%],x1[%],y1[%],z1[%]

   Retrieve pixels of the selected images belonging to the specified line (x0,y0,z0)-(x1,y1,z1).

   Example:
     [#1] image.jpg +at_line 0,0,0,100%,100%,0 line[0] 0,0,100%,100%,1,0xFF00FF00,255,0,0

 at_quadrangle:
     x0[%],y0[%],x1[%],y1[%],x2[%],y2[%],x3[%],y3[%],_interpolation,_boundary_conditions |
     x0[%],y0[%],z0[%],x1[%],y1[%],z1[%],x2[%],y2[%],z2[%],x3[%],y3[%],z3[%],_interpolation,_boundary_conditions

   Retrieve pixels of the selected images belonging to the specified 2D or 3D quadrangle.
   'interpolation' can be { 0=nearest-neighbor | 1=linear | 2=cubic }.
   'boundary_conditions' can be { 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }.

   Example:
     [#1] image.jpg params=5%,5%,95%,5%,60%,95%,40%,95% +at_quadrangle $params polygon.. 4,$params,0.5,255

 barycenter:

   Compute the barycenter vector of pixel values.

   Example:
     [#1] 256,256 ellipse 50%,50%,20%,20%,0,1,1 deform 20 +barycenter +ellipse[-2] {@0,1},5,5,0,10

 delaunay:

   Generate discrete 2D Delaunay triangulation of non-zero pixels in selected images.
   Input images must be scalar.
   Each pixel of the output image is a triplet (a,b,c) meaning the pixel belongs to
   the Delaunay triangle 'ABC' where 'a','b','c' are the labels of the pixels 'A','B','C'.

   Example:
     [#1] 400,400 rand 32,255 100%,100% noise. 0.4,2 eq. 1 mul +delaunay
     [#2] image.jpg b 1% 100%,100% noise. 0.8,2 eq. 1 mul +delaunay channels 0,2

 detect_skin:
     0<=tolerance<=1,_skin_x,_skin_y,_skin_radius>=0

   Detect skin in selected color images and output an appartenance probability map.
   Detection is performed using CbCr chromaticity data of skin pixels.
   If arguments 'skin_x', 'skin_y' and 'skin_radius' are provided, skin pixels are learnt
   from the sample pixels inside the circle located at ('skin_x','skin_y') with radius 'skin_radius'.

   Default value: 'tolerance=0.5' and 'skin_x=skiny=radius=-1'.

 displacement (+):
     [source_image],_smoothness,_precision>=0,_nb_scales>=0,_iteration_max>=0,is_backward={ 0 | 1 },_[guide]

   Estimate displacement field between specified source and selected target images.
   If 'smoothness>=0', regularization type is set to isotropic, else to anisotropic.
   If 'nbscales==0', the number of scales used is estimated from the image size.

   Default values: 'smoothness=0.1', 'precision=5', 'nb_scales=0', 'iteration_max=10000', 'is_backward=1' and '[guide]=(unused)'.

   Example:
     [#1] image.jpg +rotate 3,1,0,50%,50% +displacement[-1] [-2] quiver[-1] [-1],15,1,1,1,{1.5*iM}

 distance (+):
     isovalue[%],_metric |
     isovalue[%],[metric],_method

   Compute the unsigned distance function to specified isovalue, opt. according to a custom metric.
   'metric' can be { 0=chebyshev | 1=manhattan | 2=euclidean | 3=squared-euclidean }.
   'method' can be { 0=fast-marching | 1=low-connectivity dijkstra | 2=high-connectivity dijkstra | 3=1+return path | 4=2+return path }.

   Default value: 'metric=2' and 'method=0'.

   Example:
     [#1] image.jpg threshold 20% distance 0 pow 0.3
     [#2] 400,400 set 1,50%,50% +distance[0] 1,2 +distance[0] 1,1 distance[0] 1,0 mod 32 threshold 16 append c

   Tutorial: https://gmic.eu/oldtutorial/_distance

 fftpolar:

   Compute fourier transform of selected images, as centered magnitude/phase images.

   Example:
     [#1] image.jpg fftpolar ellipse 50%,50%,10,10,0,1,0 ifftpolar

 histogram (+):
     nb_levels>0[%],_min_value[%],_max_value[%]

   Compute the histogram of selected images.
   If value range is set, the histogram is estimated only for pixels in the specified
   value range. Argument 'max_value' must be specified if 'min_value' is set.

   Default values: 'min_value=0%' and 'max_value=100%'.

   Example:
     [#1] image.jpg +histogram 64 display_graph[-1] 400,300,3

 histogram_nd:
     nb_levels>0[%],_value0[%],_value1[%]

   Compute the 1D,2D or 3D histogram of selected multi-channels images (having 1,2 or 3 channels).
   If value range is set, the histogram is estimated only for pixels in the specified
   value range.

   Default values: 'value0=0%' and 'value1=100%'.

   Example:
     [#1] image.jpg channels 0,1 +histogram_nd 256

 histogram_cumul:
     _nb_levels>0,_is_normalized={ 0 | 1 },_val0[%],_val1[%]

   Compute cumulative histogram of selected images.

   Default values: 'nb_levels=256', 'is_normalized=0', 'val0=0%' and 'val1=100%'.

   Example:
     [#1] image.jpg +histogram_cumul 256 histogram[0] 256 display_graph 400,300,3

 histogram_pointwise:
     nb_levels>0[%],_value0[%],_value1[%]

   Compute the histogram of each vector-valued point of selected images.
   If value range is set, the histogram is estimated only for values in the specified
   value range.

   Default values: 'value0=0%' and 'value1=100%'.

 hough:
     _width>0,_height>0,gradient_norm_voting={ 0 | 1 }

   Compute hough transform (theta,rho) of selected images.

   Default values: 'width=512', 'height=width' and 'gradient_norm_voting=1'.

   Example:
     [#1] image.jpg +blur 1.5 hough[-1] 400,400 blur[-1] 0.5 add[-1] 1 log[-1]

 ifftpolar:

   Compute inverse fourier transform of selected images, from centered magnitude/phase images.

 isophotes:
     _nb_levels>0

   Render isophotes of selected images on a transparent background.

   Default value: 'nb_levels=64'

   Example:
     [#1] image.jpg blur 2 isophotes 6 dilate_circ 5 display_rgba

 label (+):
     _tolerance>=0,is_high_connectivity={ 0 | 1 },_is_L2_norm={ 0 | 1 }

   Label connected components in selected images.

   Default values: 'tolerance=0', 'is_high_connectivity=0' and 'is_L2_norm=1'.

   Example:
     [#1] image.jpg luminance threshold 60% label normalize 0,255 map 0
     [#2] 400,400 set 1,50%,50% distance 1 mod 16 threshold 8 label mod 255 map 2

   Tutorial: https://gmic.eu/oldtutorial/_label

 label_fg:
     tolerance>=0,is_high_connectivity={ 0 | 1 }

   Label connected components for non-zero values (foreground) in selected images.
   Similar to 'label' except that 0-valued pixels are not labeled.

   Default value: 'is_high_connectivity=0'.

 laar:

   Extract the largest axis-aligned rectangle in non-zero areas of selected images.
   Rectangle coordinates are returned in status, as a sequence of numbers x0,y0,x1,y1.

   Example:
     [#1] shape_cupid 256 coords=${-laar} normalize 0,255 to_rgb rectangle $coords,0.5,0,128,0

 max_patch:
     _patch_size>=1

   Return locations of maximal values in local patch-based neighborhood of given size for selected images.

   Default value: 'patch_size=16'.

   Example:
     [#1] image.jpg norm +max_patch 16

 min_patch:
     _patch_size>=1

   Return locations of minimal values in local patch-based neighborhood of given size for selected images.

   Default value: 'patch_size=16'.

   Example:
     [#1] image.jpg norm +min_patch 16

 minimal_path:
     x0[%]>=0,y0[%]>=0,z0[%]>=0,x1[%]>=0,y1[%]>=0,z1[%]>=0,_is_high_connectivity={ 0 | 1 }

   Compute minimal path between two points on selected potential maps.

   Default value: 'is_high_connectivity=0'.

   Example:
     [#1] image.jpg +gradient_norm fill[-1] 1/(1+i) minimal_path[-1] 0,0,0,100%,100%,0 pointcloud[-1] 0 *[-1] 280 to_rgb[-1] ri[-1] [-2],0 or

 mse:

   Compute MSE (Mean-Squared Error) matrix between selected images.

   Example:
     [#1] image.jpg +noise 30 +noise[0] 35 +noise[0] 38 cut. 0,255 mse

 patches:
     patch_width>0,patch_height>0,patch_depth>0,x0,y0,z0,_x1,_y1,_z1,...,_xN,_yN,_zN

   Extract N+1 patches from selected images, centered at specified locations.

   Example:
     [#1] image.jpg +patches 64,64,1,153,124,0,184,240,0,217,126,0,275,38,0

 matchpatch (+):
     [patch_image],patch_width>=1,_patch_height>=1,_patch_depth>=1,_nb_iterations>=0,_nb_randoms>=0,_patch_penalization,_output_score={ 0 | 1 },_[guide]

   Estimate correspondence map between selected images and specified patch image, using
   a patch-matching algorithm.
   Each pixel of the returned correspondence map gives the location (p,q) of the closest patch in
   the specified patch image. If 'output_score=1', the third channel also gives the corresponding
   matching score for each patch as well.
   If 'patch_penalization' is >=0, SSD is penalized with patch occurrences.
   If 'patch_penalization' is <0, SSD is inf-penalized when distance between patches are less than '-patch_penalization'.

   Default values: 'patch_height=patch_width', 'patch_depth=1', 'nb_iterations=5', 'nb_randoms=5', 'patch_penalization=0', 'output_score=0' and
    'guide=(undefined)'.

   Example:
     [#1] image.jpg sample colorful +matchpatch[0] [1],3 +warp[-2] [-1],0

 plot2value:

   Retrieve values from selected 2D graph plots.

   Example:
     [#1] 400,300,1,1,'if(y>300*abs(cos(x/10+2*u)),1,0)' +plot2value +display_graph[-1] 400,300

 pointcloud:
     _type = { -X=-X-opacity | 0=binary | 1=cumulative | 2=label | 3=retrieve coordinates },_width,_height>0,_depth>0

   Render a set of point coordinates, as a point cloud in a 1D/2D or 3D binary image
   (or do the reverse, i.e. retrieve coordinates of non-zero points from a rendered point cloud).
   Input point coordinates can be a NxMx1x1, Nx1x1xM or 1xNx1xM image, where 'N' is the number of points,
   and M the point coordinates.
   If 'M'>3, the 3-to-M components sets the (M-3)-dimensional color at each point.
   Parameters 'width','height' and 'depth' are related to the size of the final image :
      -  If set to 0, the size is automatically set along the specified axis.
      -  If set to N>0, the size along the specified axis is N.
      -  If set to N<0, the size along the specified axis is at most N.
   Points with coordinates that are negative or higher than specified ('width','height','depth')
   are not plotted.

   Default values: 'type=0' and 'max_width=max_height=max_depth=0'.

   Example:
     [#1] 3000,2 rand 0,400 +pointcloud 0 dilate[-1] 3
     [#2] 3000,2 rand 0,400 {w} {w},3 rand[-1] 0,255 append y +pointcloud 0 dilate[-1] 3

 psnr:
     _max_value

   Compute PSNR (Peak Signal-to-Noise Ratio) matrix between selected images.

   Default value: 'max_value=255'.

   Example:
     [#1] image.jpg +noise 30 +noise[0] 35 +noise[0] 38 cut[-1] 0,255 psnr 255 replace_inf 0

 segment_watershed:
     _threshold>=0

   Apply watershed segmentation on selected images.

   Default values: 'threshold=2'.

   Example:
     [#1] image.jpg segment_watershed 2

 shape2bump:
     _resolution>=0,0<=_weight_avg_max_avg<=1,_dilation,_smoothness>=0

   Estimate bumpmap from binary shape in selected images.

   Default value: 'resolution=256', 'weight_avg_max=0.75', 'dilation=0' and 'smoothness=100'.

 skeleton:
     _boundary_conditions={ 0=dirichlet | 1=neumann }

   Compute skeleton of binary shapes using distance transform and constrained thinning.

   Default value: 'boundary_conditions=1'.

   Example:
     [#1] shape_cupid 320 +skeleton 0

 slic:
     size>0,_regularity>=0,_nb_iterations>0

   Segment selected 2D images with superpixels, using the SLIC algorithm (Simple Linear Iterative Clustering).
   Scalar images of increasingly labeled pixels are returned.
   Reference paper: Achanta, R., Shaji, A., Smith, K., Lucchi, A., Fua, P., & Susstrunk, S. (2010). SLIC Superpixels (No. EPFL-REPORT-149300).

   Default values: 'size=16', 'regularity=10' and 'nb_iterations=10'.

   Example:
     [#1] image.jpg +srgb2lab slic[-1] 16 +blend shapeaverage f[-2] "j(1,0)==i && j(0,1)==i" *[-1] [-2]

 ssd_patch:
     [patch],_use_fourier={ 0 | 1 },_boundary_conditions={ 0=dirichlet | 1=neumann }

   Compute fields of SSD between selected images and specified patch.
   Argument 'boundary_conditions' is valid only when 'use_fourier=0'.

   Default value: 'use_fourier=0' and 'boundary_conditions=0'.

   Example:
     [#1] image.jpg +crop 20%,20%,35%,35% +ssd_patch[0] [1],0,0

 thinning:
     _boundary_conditions={ 0=dirichlet | 1=neumann }

   Compute skeleton of binary shapes using morphological thinning
   (beware, this is a quite slow iterative process)

   Default value: 'boundary_conditions=1'.

   Example:
     [#1] shape_cupid 320 +thinning

 tones:
     N>0

   Get N tones masks from selected images.

   Example:
     [#1] image.jpg +tones 3

 topographic_map:
     _nb_levels>0,_smoothness

   Render selected images as topographic maps.

   Default values: 'nb_levels=16' and 'smoothness=2'.

   Example:
     [#1] image.jpg topographic_map 10

 tsp:
     _precision>=0

   Try to solve the 'travelling salesman' problem, using a combination of greedy search and 2-opt algorithms.
   Selected images must have dimensions Nx1x1xC to represent N cities each with C-dimensional coordinates.
   This command re-order the selected data along the x-axis so that the point sequence becomes a shortest path.

   Default values: 'precision=256'.

   Example:
     [#1] 256,1,1,2 rand 0,512 tsp , 512,512,1,3 repeat w#0 circle[-1] {0,I[$>]},2,1,255,255,255 line[-1] {0,boundary=2;[I[$>],I[$>+1]]},1,255,128,0 done keep[-1]

 variance_patch:
     _patch_size>=1

   Compute variance of each images patch centered at (x,y), in selected images.

   Default value: 'patch_size=16'

   Example:
     [#1] image.jpg +variance_patch

 12.10. Image Drawing
        -------------

 arrow:
     x0[%],y0[%],x1[%],y1[%],_thickness[%]>=0,_head_length[%]>=0,_head_thickness[%]>=0,_opacity,_pattern,_color1,...

   Draw specified arrow on selected images.
   'pattern' is an hexadecimal number starting with '0x' which can be omitted
   even if a color is specified. If a pattern is specified, the arrow is
   drawn outlined instead of filled.

   Default values: 'thickness=1%', 'head_length=10%', 'head_thickness=3%', 'opacity=1', 'pattern=(undefined)' and 'color1=0'.

   Example:
     [#1] 400,400,1,3 repeat 100 arrow 50%,50%,{u(100)}%,{u(100)}%,3,20,10,0.3,${-rgb} done

 axes:
     x0,x1,y0,y1,_font_height>=0,_opacity,_pattern,_color1,...

   Draw xy-axes on selected images.
   'pattern' is an hexadecimal number starting with '0x' which can be omitted
   even if a color is specified.
   To draw only one x-axis at row Y, set both 'y0' and 'y1' to Y.
   To draw only one y-axis at column X, set both 'x0' and 'x1' to X.

   Default values: 'font_height=14', 'opacity=1', 'pattern=(undefined)' and 'color1=0'.

   Example:
     [#1] 400,400,1,3,255 axes -1,1,1,-1

 ball:
     _size>0, _R,_G,_B,0<=_specular_light<=8,0<=_specular_size<=8,_shadow>=0

   Input a 2D RGBA colored ball sprite.

   Default values: 'size=64', 'R=255', 'G=R', 'B=R', 'specular_light=0.8', 'specular_size=1' and 'shading=1.5'.

   Example:
     [#1] repeat 9 ball {1.5^($>+2)},${-rgb} done append x

 chessboard:
     size1>0,_size2>0,_offset1,_offset2,_angle,_opacity,_color1,...,_color2,...

   Draw chessboard on selected images.

   Default values: 'size2=size1', 'offset1=offset2=0', 'angle=0', 'opacity=1', 'color1=0' and 'color2=255'.

   Example:
     [#1] image.jpg chessboard 32,32,0,0,25,0.3,255,128,0,0,128,255

 cie1931:

   Draw CIE-1931 chromaticity diagram on selected images.

   Example:
     [#1] 500,400,1,3 cie1931

 circle:
     x[%],y[%],R[%],_opacity,_pattern,_color1,...

   Draw specified colored circle on selected images.
   A radius of '100%' stands for 'sqrt(width^2+height^2)'.
   'pattern' is an hexadecimal number starting with '0x' which can be omitted
   even if a color is specified. If a pattern is specified, the circle is
   drawn outlined instead of filled.

   Default values: 'opacity=1', 'pattern=(undefined)' and 'color1=0'.

   Example:
     [#1] image.jpg repeat 300 circle {u(100)}%,{u(100)}%,{u(30)},0.3,${-rgb} done circle 50%,50%,100,0.7,255

 close_binary:
     0<=_endpoint_rate<=100,_endpoint_connectivity>=0,_spline_distmax>=0,_segment_distmax>=0,0<=_spline_anglemax<=180,_spline_roundness>=0,_area_min>=0,_allow_self_intersection={ 0 | 1 }

   Automatically close open shapes in binary images (defining white strokes on black background).

   Default values: 'endpoint_rate=75', 'endpoint_connectivity=2', 'spline_distmax=80', 'segment_distmax=20', 'spline_anglemax=90', 'spline_roundness=1',
    'area_min=100', 'allow_self_intersection=1'.

 ellipse (+):
     x[%],y[%],R[%],r[%],_angle,_opacity,_pattern,_color1,...

   Draw specified colored ellipse on selected images.
   A radius of '100%' stands for 'sqrt(width^2+height^2)'.
   'pattern' is an hexadecimal number starting with '0x' which can be omitted
   even if a color is specified. If a pattern is specified, the ellipse is
   drawn outlined instead of filled.

   Default values: 'opacity=1', 'pattern=(undefined)' and 'color1=0'.

   Example:
     [#1] image.jpg repeat 300 ellipse {u(100)}%,{u(100)}%,{u(30)},{u(30)},{u(180)},0.3,${-rgb} done ellipse 50%,50%,100,100,0,0.7,255

 flood (+):
     x[%],_y[%],_z[%],_tolerance>=0,_is_high_connectivity={ 0 | 1 },_opacity,_color1,...

   Flood-fill selected images using specified value and tolerance.

   Default values: 'y=z=0', 'tolerance=0', 'is_high_connectivity=0', 'opacity=1' and 'color1=0'.

   Example:
     [#1] image.jpg repeat 1000 flood {u(100)}%,{u(100)}%,0,20,0,1,${-rgb} done

 gaussian:
     _sigma1[%],_sigma2[%],_angle

   Draw a centered gaussian on selected images, with specified standard deviations and orientation.

   Default values: 'sigma1=3', 'sigma2=sigma1' and 'angle=0'.

   Example:
     [#1] 400,400 gaussian 100,30,45

   Tutorial: https://gmic.eu/oldtutorial/_gaussian

 graph (+):
     [function_image],_plot_type,_vertex_type,_ymin,_ymax,_opacity,_pattern,_color1,... |
     'formula',_resolution>=0,_plot_type,_vertex_type,_xmin,xmax,_ymin,_ymax,_opacity,_pattern,_color1,...

   Draw specified function graph on selected images.
   'plot_type' can be { 0=none | 1=lines | 2=splines | 3=bar }.
   'vertex_type' can be { 0=none | 1=points | 2,3=crosses | 4,5=circles | 6,7=squares }.
   'pattern' is an hexadecimal number starting with '0x' which can be omitted
   even if a color is specified.

   Default values: 'plot_type=1', 'vertex_type=1', 'ymin=ymax=0 (auto)', 'opacity=1', 'pattern=(undefined)'
   and 'color1=0'.

   Example:
     [#1] image.jpg +rows 50% blur[-1] 3 split[-1] c div[0] 1.5 graph[0] [1],2,0,0,0,1,255,0,0 graph[0] [2],2,0,0,0,1,0,255,0 graph[0] [3],2,0,0,0,1,0,0,255 keep[0]

 grid:
     size_x[%]>=0,size_y[%]>=0,_offset_x[%],_offset_y[%],_opacity,_pattern,_color1,...

   Draw xy-grid on selected images.
   'pattern' is an hexadecimal number starting with '0x' which can be omitted
   even if a color is specified.

   Default values: 'offset_x=offset_y=0', 'opacity=1', 'pattern=(undefined)' and 'color1=0'.

   Example:
     [#1] image.jpg grid 10%,10%,0,0,0.5,255
     [#2] 400,400,1,3,255 grid 10%,10%,0,0,0.3,0xCCCCCCCC,128,32,16

 j (+):
     Shortcut for command 'image'.

 image (+):
     [sprite],_x[%|~],_y[%|~],_z[%|~],_c[%|~],_opacity,_[opacity_mask],_max_opacity_mask

   Draw specified sprite image on selected images.
   (equivalent to shortcut command 'j').

   If one of the x,y,z or c argument ends with a '~', its value is expected to be
   a centering ratio (in [0,1]) rather than a position.
   Usual centering ratio are { 0=left-justified | 0.5=centered | 1=right-justified }.

   Default values: 'x=y=z=c=0', 'opacity=1', 'opacity_mask=(undefined)' and 'max_opacity_mask=1'.

   Example:
     [#1] image.jpg +crop 40%,40%,60%,60% resize[-1] 200%,200%,1,3,5 frame[-1] 2,2,0 image[0] [-1],30%,30% keep[0]

 line (+):
     x0[%],y0[%],x1[%],y1[%],_opacity,_pattern,_color1,...

   Draw specified colored line on selected images.
   'pattern' is an hexadecimal number starting with '0x' which can be omitted
   even if a color is specified.

   Default values: 'opacity=1', 'pattern=(undefined)' and 'color1=0'.

   Example:
     [#1] image.jpg repeat 500 line 50%,50%,{u(w)},{u(h)},0.5,${-rgb} done line 0,0,100%,100%,1,0xCCCCCCCC,255 line 100%,0,0,100%,1,0xCCCCCCCC,255

 linethick:
     x0[%],y0[%],x1[%],y1[%],_thickness,_opacity,_color1

   Draw specified colored thick line on selected images.

   Default values: 'thickness=2', 'opacity=1' and 'color1=0'.

   Example:
     [#1] 400,400,1,3 repeat 100 linethick {u([w,h,w,h,5])},0.5,${-rgb} done

 mandelbrot (+):
     z0r,z0i,z1r,z1i,_iteration_max>=0,_is_julia={ 0 | 1 },_c0r,_c0i,_opacity

   Draw mandelbrot/julia fractal on selected images.

   Default values: 'iteration_max=100', 'is_julia=0', 'c0r=c0i=0' and 'opacity=1'.

   Example:
     [#1] 400,400 mandelbrot -2.5,-2,2,2,1024 map 0 +blur 2 elevation3d[-1] -0.2

 marble:
     _image_weight,_pattern_weight,_angle,_amplitude,_sharpness>=0,_anisotropy>=0,_alpha,_sigma,_cut_low>=0,_cut_high>=0

   Render marble like pattern on selected images.

   Default values: 'image_weight=0.2', 'pattern_weight=0.1', 'angle=45', 'amplitude=0', 'sharpness=0.4' and 'anisotropy=0.8',
   'alpha=0.6', 'sigma=1.1' and 'cut_low=cut_high=0'.

   Example:
     [#1] image.jpg +marble ,

 maze:
     _width>0,_height>0,_cell_size>0

   Input maze with specified size.

   Example:
     [#1] maze 30,20 negate normalize 0,255

 maze_mask:
     _cellsize>0

   Input maze according to size and shape of selected mask images.
   Mask may contain disconnected shapes.

   Example:
     [#1] 0 text "G'MIC",0,0,53,1,1 dilate 3 autocrop 0 frame 1,1,0 maze_mask 8 dilate 3 negate mul 255

 newton_fractal:
     z0r,z0i,z1r,z1i,_angle,0<=_descent_method<=2,_iteration_max>=0,_convergence_precision>0,_expr_p(z),_expr_dp(z),_expr_d2p(z)

   Draw newton fractal on selected images, for complex numbers in range (z0r,z0i) - (z1r,z1i).
   Resulting images have 3 channels whose meaning is [ last_zr, last_zi, nb_iter_used_for_convergence ].
   'descent_method' can be { 0=secant | 1=newton | 2=householder }.

   Default values: 'angle=0', 'descent_method=1', 'iteration_max=200', 'convergence_precision=0.01', 'expr_p(z)=z^^3-1', 'expr_dp(z)=3*z^^2' and
    'expr_d2z(z)=6*z'.

   Example:
     [#1] 400,400 newton_fractal -1.5,-1.5,1.5,1.5,0,2,200,0.01,"z^^6 + z^^3 - 1","6*z^^5 + 3*z^^2","30*z^^4 + 6*z" f "[ atan2(i1,i0)*90+20,1,cut(i2/30,0.2,0.7) ]" hsl2rgb

 j3d (+):
     Shortcut for command 'object3d'.

 object3d (+):
     [object3d],_x[%],_y[%],_z,_opacity,_rendering_mode,_is_double_sided={ 0 | 1 },_is_zbuffer={ 0 | 1 },_focale,_light_x,_light_y,_light_z,_specular_lightness,_specular_shininess

   Draw specified 3D object on selected images.
   (equivalent to shortcut command 'j3d').

   'rendering_mode' can be { 0=dots | 1=wireframe | 2=flat | 3=flat-shaded | 4=gouraud-shaded | 5=phong-shaded }.

   Default values: 'x=y=z=0', 'opacity=1' and 'is_zbuffer=1'. All other arguments take their default values
   from the 3D environment variables.

   Example:
     [#1] image.jpg torus3d 100,10 cone3d 30,-120 add3d[-2,-1] rotate3d. 1,1,0,60 object3d[0] [-1],50%,50% keep[0]

 pack_sprites:
     _nb_scales>=0,0<=_min_scale<=100,_allow_rotation={ 0=0 deg. | 1=180 deg. | 2=90 deg. | 3=any },_spacing,_precision>=0,max_iterations>=0

   Try to randomly pack as many sprites as possible onto the 'empty' areas of an image.
   Sprites can be eventually rotated and scaled during the packing process.
   First selected image is the canvas that will be filled with the sprites.
   Its last channel must be a binary mask whose zero values represent potential locations for drawing the sprites.
   All other selected images represent the sprites considered for packing.
   Their last channel must be a binary mask that represents the sprite shape (i.e. a 8-connected component).
   The order of sprite packing follows the order of specified sprites in the image list.
   Sprite packing is done on random locations and iteratively with decreasing scales.
   'nb_scales' sets the number of decreasing scales considered for all specified sprites to be packed.
   'min_scale' (in %) sets the minimal size considered for packing (specified as a percentage of the
   original sprite size).
   'spacing' can be positive or negative.
   'precision' tells about the desired number of failed trials before ending the filling process.

   Default values: 'nb_scales=5', 'min_scale=25', 'allow_rotation=3', 'spacing=1', 'precision=7' and 'max_iterations=256'.

   Example:
     [#1] 512,512,1,3,"min(255,y*c/2)" 100%,100% circle 50%,50%,100,1,255 append c image.jpg resize2dy[-1] 24 to_rgba pack_sprites 3,25

 piechart:
     label_height>=0,label_R,label_G,label_B,"label1",value1,R1,G1,B1,...,"labelN",valueN,RN,GN,BN

   Draw pie chart on selected (RGB) images.

   Example:
     [#1] image.jpg piechart 25,0,0,0,"Red",55,255,0,0,"Green",40,0,255,0,"Blue",30,128,128,255,"Other",5,128,128,128

 plasma (+):
     _alpha,_beta,_scale>=0

   Draw a random colored plasma fractal on selected images.
   This command implements the so-called 'Diamond-Square' algorithm.

   Default values: 'alpha=1', 'beta=1' and 'scale=8'.

   Example:
     [#1] 400,400,1,3 plasma

   Tutorial: https://gmic.eu/oldtutorial/_plasma

 point (+):
     x[%],y[%],_z[%],_opacity,_color1,...

   Set specified colored pixel on selected images.

   Default values: 'z=0', 'opacity=1' and 'color1=0'.

   Example:
     [#1] image.jpg repeat 10000 point {u(100)}%,{u(100)}%,0,1,${-rgb} done

 polka_dots:
     diameter>=0,_density,_offset1,_offset2,_angle,_aliasing,_shading,_opacity,_color,...

   Draw dots pattern on selected images.

   Default values: 'density=20', 'offset1=offset2=50', 'angle=0', 'aliasing=10', 'shading=1', 'opacity=1' and 'color=255'.

   Example:
     [#1] image.jpg polka_dots 10,15,0,0,20,10,1,0.5,0,128,255

 polygon (+):
     N>=1,x1[%],y1[%],...,xN[%],yN[%],_opacity,_pattern,_color1,...

   Draw specified colored N-vertices polygon on selected images.
   'pattern' is an hexadecimal number starting with '0x' which can be omitted
   even if a color is specified. If a pattern is specified, the polygon is
   drawn outlined instead of filled.

   Default values: 'opacity=1', 'pattern=(undefined)' and 'color1=0'.

   Example:
     [#1] image.jpg polygon 4,20%,20%,80%,30%,80%,70%,20%,80%,0.3,0,255,0 polygon 4,20%,20%,80%,30%,80%,70%,20%,80%,1,0xCCCCCCCC,255
     [#2] image.jpg 2,16,1,1,'u(if(x,{h},{w}))' polygon[-2] {h},{^},0.6,255,0,255 remove[-1]

 quiver:
     [function_image],_sampling[%]>0,_factor>=0,_is_arrow={ 0 | 1 },_opacity,_color1,...

   Draw specified 2D vector/orientation field on selected images.

   Default values: 'sampling=5%', 'factor=1', 'is_arrow=1', 'opacity=1', 'pattern=(undefined)'
   and 'color1=0'.

   Example:
     [#1] 100,100,1,2,'if(c==0,x-w/2,y-h/2)' 500,500,1,3,255 quiver[-1] [-2],10
     [#2] image.jpg +resize2dy 600 luminance[0] gradient[0] mul[1] -1 reverse[0,1] append[0,1] c blur[0] 8 orientation[0] quiver[1] [0],20,1,1,0.8,255

 rectangle:
     x0[%],y0[%],x1[%],y1[%],_opacity,_pattern,_color1,...

   Draw specified colored rectangle on selected images.
   'pattern' is an hexadecimal number starting with '0x' which can be omitted
   even if a color is specified. If a pattern is specified, the rectangle is
   drawn outlined instead of filled.

   Default values: 'opacity=1', 'pattern=(undefined)' and 'color1=0'.

   Example:
     [#1] image.jpg repeat 30 rectangle {u(100)}%,{u(100)}%,{u(100)}%,{u(100)}%,0.3,${-rgb} done

 rorschach:
     'smoothness[%]>=0','mirroring={ 0=none | 1=x | 2=y | 3=xy }

   Render rorschach-like inkblots on selected images.

   Default values: 'smoothness=5%' and 'mirroring=1'.

   Example:
     [#1] 400,400 rorschach 3%

 sierpinski:
     recursion_level>=0

   Draw Sierpinski triangle on selected images.

   Default value: 'recursion_level=7'.

   Example:
     [#1] image.jpg sierpinski 7

 spiralbw:
     width>0,_height>0,_is_2dcoords={ 0 | 1 }

   Input a 2D rectangular spiral image with specified size.

   Default values: 'height=width' and 'is_2dcoords=0'.

   Example:
     [#1] spiralbw 16
     [#2] image.jpg spiralbw {[w,h]},1 +warp[0] [1],0 +warp[2] [1],2

 spline:
     x0[%],y0[%],u0[%],v0[%],x1[%],y1[%],u1[%],v1[%],_opacity,_color1,...

   Draw specified colored spline curve on selected images (cubic hermite spline).

   Default values: 'opacity=1' and 'color1=0'.

   Example:
     [#1] image.jpg repeat 30 spline {u(100)}%,{u(100)}%,{u(-600,600)},{u(-600,600)},{u(100)}%,{u(100)}%,{u(-600,600)},{u(-600,600)},0.6,255 done

 tetraedron_shade:
     x0,y0,z0,x1,y1,z1,x2,y2,z2,x3,y3,z3,R0,G0,B0,...,R1,G1,B1,...,R2,G2,B2,...,R3,G3,B3,...

   Draw tetraedron with interpolated colors on selected (volumetric) images.

 t (+):
     Shortcut for command 'text'.

 text (+):
     text,_x[%|~],_y[%|~],_font_height[%]>=0,_opacity,_color1,...

   Draw specified colored text string on selected images.
   (equivalent to shortcut command 't').

   If one of the x or y argument ends with a '~', its value is expected to be
   a centering ratio (in [0,1]) rather than a position.
   Usual centering ratio are { 0=left-justified | 0.5=centered | 1=right-justified }.
   Sizes '13' and '128' are special and correspond to binary fonts (no-antialiasing).
   Any other font size is rendered with anti-aliasing.
   Specifying an empty target image resizes it to new dimensions such that the image contains
   the entire text string.

   Default values: 'x=y=0.01~', 'font_height=16', 'opacity=1' and 'color1=0'.

   Example:
     [#1] image.jpg resize2dy 600 y=0 repeat 30 text {2*$>}" : This is a nice text, isn't it ?",10,$y,{2*$>},0.9,255 y+={2*$>} done
     [#2] 0 text "G'MIC",0,0,23,1,255

 to:
     Shortcut for command 'text_outline'.

 text_outline:
     text,_x[%|~],_y[%|~],_font_height[%]>0,_outline>=0,_opacity,_color1,...

   Draw specified colored and outlined text string on selected images.
   If one of the x or y argument ends with a '~', its value is expected to be
   a centering ratio (in [0,1]) rather than a position.
   Usual centering ratio are { 0=left-justified | 0.5=centered | 1=right-justified }.

   Default values: 'x=y=0.01~', 'font_height=7.5%', 'outline=2', 'opacity=1', 'color1=color2=color3=255' and 'color4=255'.

   Example:
     [#1] image.jpg text_outline "Hi there!",10,10,63,3

 triangle_shade:
     x0,y0,x1,y1,x2,y2,R0,G0,B0,...,R1,G1,B1,...,R2,G2,B2,...

   Draw triangle with interpolated colors on selected images.

   Example:
     [#1] image.jpg triangle_shade 20,20,400,100,120,200,255,0,0,0,255,0,0,0,255

 truchet:
     _scale>0,_radius>=0,_pattern_type={ 0=straight | 1=curved }

   Fill selected images with random truchet patterns.

   Default values: 'scale=32', 'radius=5' and 'pattern_type=1'.

   Example:
     [#1] 400,300 truchet ,

 turbulence:
     _radius>0,_octaves={1,2,3...,12},_alpha>0,_difference={-10,10},_mode={0,1,2,3}

   Render fractal noise or turbulence on selected images.

   Default values: 'radius=32', 'octaves=6', 'alpha=3', 'difference=0' and 'mode=0'.

   Example:
     [#1] 400,400,1,3 turbulence 16

   Tutorial: https://gmic.eu/oldtutorial/_turbulence

 yinyang:

   Draw a yin-yang symbol on selected images.

   Example:
     [#1] 400,400 yinyang

 12.11. Matrix Computation
        ------------------

 dijkstra (+):
     starting_node>=0,ending_node>=0

   Compute minimal distances and paths from specified adjacency matrices by the Dijkstra algorithm.

 eigen (+):

   Compute the eigenvalues and eigenvectors of selected symmetric matrices or matrix fields.
   If one selected image has 3 or 6 channels, it is regarded as a field of 2x2 or 3x3 symmetric matrices,
   whose eigen elements are computed at each point of the field.

   Example:
     [#1] (1,0,0;0,2,0;0,0,3) +eigen
     [#2] image.jpg structuretensors blur 2 eigen split[0] c

   Tutorial: https://gmic.eu/oldtutorial/_eigen

 invert (+):
     solver={ 0=SVD | 1=LU }

   Compute the inverse of the selected matrices.
   SVD solver is slower but less numerically instable than LU.

   Default value: 'solver=1'.

   Example:
     [#1] (0,1,0;0,0,1;1,0,0) +invert

 orthogonalize:
     _mode = { 0=orthogonalize | 1=orthonormalize }

   Orthogonalize or orthonormalize selected matrices, using Modified Gram-Schmidt process.

   Default value: 'mode=0'.

 meigen:
     m>=1

   Compute an approximation of the 'm' largest eigenvalues and eigenvectors of selected symmetric matrices,
   using the Arnoldi iteration method (https://en.wikipedia.org/wiki/Arnoldi_iteration).
   A larger 'm' goes with better numerical precision.

   Example:
     [#1] (1,0,0;0,2,0;0,0,3) +meigen 3

 mproj (+):
     [dictionary],_method,_max_iter={ 0=auto | >0 },_max_residual>=0

   Find best matching projection of selected matrices onto the span of an over-complete
   dictionary D, using the orthogonal projection or Matching Pursuit algorithm.
   Selected images are 2D-matrices in which each column represent a signal to project.
   '[dictionary]' is a matrix in which each column is an element of the dictionary D.
   'method' tells what projection algorithm must be applied. It can be:
      - 0 = orthogonal projection (least-squares solution using LU-based solver).
      - 1 = matching pursuit.
      - 2 = matching pursuit, with a single orthogonal projection step at the end.
      - >=3 = orthogonal matching pursuit where an orthogonal projection step is performed
              every 'method-2' iterations.
   'max_iter' sets the max number of iterations processed for each signal.
   If set to '0' (default), 'max_iter' is equal to the number of columns in D.
   (only meaningful for matching pursuit and its variants).
   'max_residual' gives a stopping criterion on signal reconstruction accuracy.
   (only meaningful for matching pursuit and its variants).
   For each selected image, the result is returned as a matrix W
   whose columns correspond to the weights associated to each column of D,
   such that the matrix product D*W is an approximation of the input matrix.

   Default values: 'method=0', 'max_iter=0' and 'max_residual=1e-6'.

 solve (+):
     [image]

   Solve linear system AX = B for selected B-matrices and specified A-matrix.
   If the system is under- or over-determined, the least squares solution is returned
   (using SVD-based solver).

   Example:
     [#1] (0,1,0;1,0,0;0,0,1) (1;2;3) +solve[-1] [-2]

 svd (+):

   Compute SVD decomposition of selected matrices.

   Example:
     [#1] 10,10,1,1,'if(x==y,x+u(-0.2,0.2),0)' +svd

 transpose:

   Transpose selected matrices.

   Example:
     [#1] image.jpg +transpose

 trisolve (+):
     [image]

   Solve tridiagonal system AX = B for selected B-vectors and specified tridiagonal A-matrix.
   Tridiagonal matrix must be stored as a 3 column vector, where 2nd column contains the
   diagonal coefficients, while 1st and 3rd columns contain the left and right coefficients.

   Example:
     [#1] (0,0,1;1,0,0;0,1,0) (1;2;3) +trisolve[-1] [-2]

 12.12. 3D Meshes
        ---------

 +3d (+):
     Shortcut for command 'add3d'.

 add3d (+):
     tx,_ty,_tz |
     [object3d] |
     (no arg)

   Shift selected 3D objects with specified displacement vector, or merge them with specified
   3D object, or merge all selected 3D objects together.
   (equivalent to shortcut command '+3d').

   Default values: 'ty=tz=0'.

   Example:
     [#1] sphere3d 10 repeat 5 +add3d[-1] 10,{u(-10,10)},0 color3d[-1] ${-rgb} done add3d
     [#2] repeat 20 torus3d 15,2 color3d[-1] ${-rgb} mul3d[-1] 0.5,1 if $>%2 rotate3d[-1] 0,1,0,90 fi add3d[-1] 70 add3d rotate3d[-1] 0,0,1,18 done double3d 0

 animate3d:
     _width>0,_height>0,_angle_dx,_angle_dy,_angle_dz,_zoom_factor>=0,_filename

   Animate selected 3D objects in a window.
   If argument 'filename' is provided, each frame of the animation is saved as a numbered filename.

   Default values: 'width=640', 'height=480', 'angle_dx=0', 'angle_dy=1', 'angle_dz=0', 'zoom_factor=1' and 'filename=(undefined)'.

 apply_camera3d:
     pos_x,pos_y,pos_z,target_x,target_y,target_z,up_x,up_y,up_z

   Apply 3D camera matrix to selected 3D objects.

   Default values: 'target_x=0', 'target_y=0', 'target_z=0', 'up_x=0', 'up_y=-1' and 'up_z=0'.

 apply_matrix3d:
     a11,a12,a13,...,a31,a32,a33

   Apply specified 3D rotation matrix to selected 3D objects.

   Example:
     [#1] torus3d 10,1 +apply_matrix3d {mul(rot(1,0,1,-15°),[1,0,0,0,2,0,0,0,8],3)} double3d 0

 array3d:
     size_x>=1,_size_y>=1,_size_z>=1,_offset_x[%],_offset_y[%],_offset_y[%]

   Duplicate a 3D object along the X,Y and Z axes.

   Default values: 'size_y=1', 'size_z=1' and 'offset_x=offset_y=offset_z=100%'.

   Example:
     [#1] torus3d 10,1 +array3d 5,5,5,110%,110%,300%

 arrow3d:
     x0,y0,z0,x1,y1,z1,_radius[%]>=0,_head_length[%]>=0,_head_radius[%]>=0

   Input 3D arrow with specified starting and ending 3D points.

   Default values: 'radius=5%', 'head_length=25%' and 'head_radius=15%'.

   Example:
     [#1] repeat 10 a={$>*2*pi/10} arrow3d 0,0,0,{cos($a)},{sin($a)},-0.5 done +3d

 axes3d:
     _size_x,_size_y,_size_z,_font_size>0,_label_x,_label_y,_label_z,_is_origin={ 0=no | 1=yes }

   Input 3D axes with specified sizes along the x,y and z orientations.

   Default values: 'size_x=size_y=size_z=1', 'font_size=23', 'label_x=X', 'label_y=Y', 'label_z=Z' and 'is_origin=1'

   Example:
     [#1] axes3d ,

 boundingbox3d:

   Replace selected 3D objects by their 3D bounding boxes.

   Example:
     [#1] torus3d 100,30 +boundingbox3d +3d[-1] [-2]

 box3d:
     _size_x,_size_y,_size_z

   Input 3D box at (0,0,0), with specified geometry.

   Default values: 'size_x=1' and 'size_z=size_y=size_x'.

   Example:
     [#1] box3d 100,40,30 +primitives3d 1 color3d[-2] ${-rgb}

 c3d:
     Shortcut for command 'center3d'.

 center3d:

   Center selected 3D objects at (0,0,0).
   (equivalent to shortcut command 'c3d').

   Example:
     [#1] repeat 100 circle3d {u(100)},{u(100)},{u(100)},2 done add3d color3d[-1] 255,0,0 +center3d color3d[-1] 0,255,0 add3d

 circle3d:
     _x0,_y0,_z0,_radius>=0

   Input 3D circle at specified coordinates.

   Default values: 'x0=y0=z0=0' and 'radius=1'.

   Example:
     [#1] repeat 500 a={$>*pi/250} circle3d {cos(3*$a)},{sin(2*$a)},0,{$a/50} color3d[-1] ${-rgb},0.4 done add3d

 circles3d:
     _radius>=0,_is_wireframe={ 0 | 1 }

   Convert specified 3D objects to sets of 3D circles with specified radius.

   Default values: 'radius=1' and 'is_wireframe=1'.

   Example:
     [#1] image.jpg luminance resize2dy 40 threshold 50% * 255 pointcloud3d color3d[-1] 255,255,255 circles3d 0.7

 col3d (+):
     Shortcut for command 'color3d'.

 color3d (+):
     R,_G,_B,_opacity |
     (no arg)

   Set color and opacity of selected 3D objects.
   (equivalent to shortcut command 'col3d').

   Default value: 'B=G=R' and 'opacity=(undefined)'.

   Example:
     [#1] torus3d 100,10 double3d 0 repeat 7 +rotate3d[-1] 1,0,0,20 color3d[-1] ${-rgb} done add3d

 colorcube3d:

   Input 3D color cube.

   Example:
     [#1] colorcube3d mode3d 2 +primitives3d 1

 cone3d:
     _radius,_height,_nb_subdivisions>0

   Input 3D cone at (0,0,0), with specified geometry.

   Default value: 'radius=1','height=1' and 'nb_subdivisions=24'.

   Example:
     [#1] cone3d 10,40 +primitives3d 1 color3d[-2] ${-rgb}

 cubes3d:
     _size>=0

   Convert specified 3D objects to sets of 3D cubes with specified size.

   Default value: 'size=1'.

   Example:
     [#1] image.jpg luminance resize2dy 40 threshold 50% * 255 pointcloud3d color3d[-1] 255,255,255 cubes3d 1

 cup3d:
     _resolution>0

   Input 3D cup object.

   Default value: 'resolution=128'.

   Example:
     [#1] cup3d ,

 cylinder3d:
     _radius,_height,_nb_subdivisions>0

   Input 3D cylinder at (0,0,0), with specified geometry.

   Default value: 'radius=1','height=1' and 'nb_subdivisions=24'.

   Example:
     [#1] cylinder3d 10,40 +primitives3d 1 color3d[-2] ${-rgb}

 delaunay3d:

   Generate 3D delaunay triangulations from selected images.
   One assumes that the selected input images are binary images containing the set of points to mesh.
   The output 3D object is a mesh composed of non-oriented triangles.

   Example:
     [#1] 500,500 noise 0.05,2 eq 1 * 255 +delaunay3d color3d[1] 255,128,0 dilate_circ[0] 5 to_rgb[0] +object3d[0] [1],0,0,0,1,1 max[-1] [0]

 distribution3d:

   Get 3D color distribution of selected images.

   Example:
     [#1] image.jpg distribution3d colorcube3d primitives3d[-1] 1 add3d

 /3d (+):
     Shortcut for command 'div3d'.

 div3d (+):
     factor |
     factor_x,factor_y,_factor_z

   Scale selected 3D objects isotropically or anisotropically, with the inverse of specified
   factors.
   (equivalent to shortcut command '/3d').

   Default value: 'factor_z=0'.

   Example:
     [#1] torus3d 5,2 repeat 5 +add3d[-1] 12,0,0 div3d[-1] 1.2 color3d[-1] ${-rgb} done add3d

 db3d (+):
     Shortcut for command 'double3d'.

 double3d (+):
     _is_double_sided={ 0 | 1 }

   Enable/disable double-sided mode for 3D rendering.
   (equivalent to shortcut command 'db3d').

   Default value: 'is_double_sided=1'.

   Example:
     [#1] mode3d 1 repeat 2 torus3d 100,30 rotate3d[-1] 1,1,0,60 double3d $> snapshot3d[-1] 400 done

 elevation3d:
     { z-factor | [elevation_map] | 'formula' },base_height={ -1 | >=0 } |
     (no arg)

   Generate 3D elevation of selected images, opt. with a specified elevation map.
   When invoked with (no arg) or 'z-factor', the elevation map is computed as the pointwise L2 norm of the
   pixel values. Otherwise, the elevation map is taken from the specified image or formula.

   Example:
     [#1] image.jpg +blur 5 elevation3d. 0.75
     [#2] 128,128,1,3,u(255) plasma 10,3 blur 4 sharpen 10000 n 0,255 elevation3d[-1] 'X=(x-64)/6;Y=(y-64)/6;-100*exp(-(X^2+Y^2)/30)*abs(cos(X)*sin(Y))'

 empty3d:

   Input empty 3D object.

   Example:
     [#1] empty3d

 extrude3d:
     _depth>0,_resolution>0,_smoothness[%]>=0

   Generate extruded 3D object from selected binary XY-profiles.

   Default values: 'depth=16', 'resolution=1024' and 'smoothness=0.5%'.

   Example:
     [#1] image.jpg threshold 50% extrude3d 16

 f3d (+):
     Shortcut for command 'focale3d'.

 focale3d (+):
     focale

   Set 3D focale.
   (equivalent to shortcut command 'f3d').

   Set 'focale' to 0 to enable parallel projection (instead of perspective).
   Set negative 'focale' will disable 3D sprite zooming.

   Default value: 'focale=700'.

   Example:
     [#1] repeat 5 torus3d 100,30 rotate3d[-1] 1,1,0,60 focale3d {$<*90} snapshot3d[-1] 400 done remove[0]

 gaussians3d:
     _size>0,_opacity

   Convert selected 3D objects into set of 3D gaussian-shaped sprites.

   Example:
     [#1] image.jpg r2dy 32 distribution3d gaussians3d 20 colorcube3d primitives3d[-1] 1 +3d

 gmic3d:

   Input a 3D G'MIC logo.

   Example:
     [#1] gmic3d +primitives3d 1

 gyroid3d:
     _resolution>0,_zoom

   Input 3D gyroid at (0,0,0), with specified resolution.

   Default values: 'resolution=32' and 'zoom=5'.

   Example:
     [#1] gyroid3d 48 +primitives3d 1

 histogram3d:

   Get 3D color histogram of selected images.

   Example:
     [#1] image.jpg resize2dx 64 histogram3d circles3d 3 opacity3d. 0.75 colorcube3d primitives3d[-1] 1 add3d

 image6cube3d:

   Generate 3D mapped cubes from 6-sets of selected images.

   Example:
     [#1] image.jpg animate flower,"30,0","30,5",6 image6cube3d

 imageblocks3d:
     _maximum_elevation,_smoothness[%]>=0

   Generate 3D blocks from selected images.
   Transparency of selected images is taken into account.

   Default values: 'maximum_elevation=10' and 'smoothness=0'.

   Example:
     [#1] image.jpg resize2dy 32 imageblocks3d -20 mode3d 3

 imagecube3d:

   Generate 3D mapped cubes from selected images.

   Example:
     [#1] image.jpg imagecube3d

 imageplane3d:

   Generate 3D mapped planes from selected images.

   Example:
     [#1] image.jpg imageplane3d

 imagepyramid3d:

   Generate 3D mapped pyramids from selected images.

   Example:
     [#1] image.jpg imagepyramid3d

 imagerubik3d:
     _xy_tiles>=1,0<=xy_shift<=100,0<=z_shift<=100

   Generate 3D mapped rubik's cubes from selected images.

   Default values: 'xy_tiles=3', 'xy_shift=5' and 'z_shift=5'.

   Example:
     [#1] image.jpg imagerubik3d ,

 imagesphere3d:
     _resolution1>=3,_resolution2>=3

   Generate 3D mapped sphere from selected images.

   Default values: 'resolution1=32' and 'resolutions2=16'.

   Example:
     [#1] image.jpg imagesphere3d 32,16

 isoline3d (+):
     isovalue[%] |
     'formula',value,_x0,_y0,_x1,_y1,_size_x>0[%],_size_y>0[%]

   Extract 3D isolines with specified value from selected images or from specified formula.

   Default values: 'x0=y0=-3', 'x1=y1=3' and 'size_x=size_y=256'.

   Example:
     [#1] image.jpg blur 1 isoline3d 50%
     [#2] isoline3d 'X=x-w/2;Y=y-h/2;(X^2+Y^2)%20',10,-10,-10,10,10

 isosurface3d (+):
     isovalue[%] |
     'formula',value,_x0,_y0,_z0,_x1,_y1,_z1,_size_x>0[%],_size_y>0[%],_size_z>0[%]

   Extract 3D isosurfaces with specified value from selected images or from specified formula.

   Default values: 'x0=y0=z0=-3', 'x1=y1=z1=3' and 'size_x=size_y=size_z=32'.

   Example:
     [#1] image.jpg resize2dy 128 luminance threshold 50% expand_z 2,0 blur 1 isosurface3d 50% mul3d 1,1,30
     [#2] isosurface3d 'x^2+y^2+abs(z)^abs(4*cos(x*y*z*3))',3

 label3d:
     "text",font_height>=0,_opacity,_color1,...

   Generate 3D text label.

   Default values: 'font_height=13', 'opacity=1' and 'color=255,255,255'.

 label_points3d:
     _label_size>0,_opacity

   Add a numbered label to all vertices of selected 3D objects.

   Default values: 'label_size=13' and 'opacity=0.8'.

   Example:
     [#1] torus3d 100,40,6,6 label_points3d 23,1 mode3d 1

 lathe3d:
     _resolution>0,_smoothness[%]>=0,_max_angle>=0

   Generate 3D object from selected binary XY-profiles.

   Default values: 'resolution=128', 'smoothness=0.5%' and 'max_angle=361'.

   Example:
     [#1] 300,300 rand -1,1 blur 40 sign normalize 0,255 lathe3d ,

 l3d (+):
     Shortcut for command 'light3d'.

 light3d (+):
     position_x,position_y,position_z |
     [texture] |
     (no arg)

   Set the light coordinates or the light texture for 3D rendering.
   (equivalent to shortcut command 'l3d').

   (no arg) resets the 3D light to default.

   Example:
     [#1] torus3d 100,30 double3d 0 specs3d 1.2 repeat 5 light3d {$>*100},0,-300 +snapshot3d[0] 400 done remove[0]

 line3d:
     x0,y0,z0,x1,y1,z1

   Input 3D line at specified coordinates.

   Example:
     [#1] repeat 100 a={$>*pi/50} line3d 0,0,0,{cos(3*$a)},{sin(2*$a)},0 color3d. ${-rgb} done add3d

 lissajous3d:
     resolution>1,a,A,b,B,c,C

   Input 3D lissajous curves (x(t)=sin(at+A*2pi), y(t)=sin(bt+B*2pi), z(t)=sin(ct+C*2pi)).

   Default values: 'resolution=1024', 'a=2', 'A=0', 'b=1', 'B=0', 'c=0' and 'C=0'.

   Example:
     [#1] lissajous3d ,

 m3d (+):
     Shortcut for command 'mode3d'.

 mode3d (+):
     _mode

   Set static 3D rendering mode.
   (equivalent to shortcut command 'm3d').

   'mode' can be { -1=bounding-box | 0=dots | 1=wireframe | 2=flat | 3=flat-shaded | 4=gouraud-shaded | 5=phong-shaded }.");
   Bounding-box mode ('mode==-1') is active only for the interactive 3D viewer.

   Default value: 'mode=4'.

   Example:
     [#1] (0,1,2,3,4,5) double3d 0 repeat w torus3d 100,30 rotate3d[-1] 1,1,0,60 mode3d {0,@$>} snapshot3d[-1] 300 done remove[0]

 md3d (+):
     Shortcut for command 'moded3d'.

 moded3d (+):
     _mode

   Set dynamic 3D rendering mode for interactive 3D viewer.
   (equivalent to shortcut command 'md3d').

   'mode' can be { -1=bounding-box | 0=dots | 1=wireframe | 2=flat | 3=flat-shaded | 4=gouraud-shaded | 5=phong-shaded }.

   Default value: 'mode=-1'.

 *3d (+):
     Shortcut for command 'mul3d'.

 mul3d (+):
     factor |
     factor_x,factor_y,_factor_z

   Scale selected 3D objects isotropically or anisotropically, with specified factors.
   (equivalent to shortcut command '*3d').

   Default value: 'factor_z=0'.

   Example:
     [#1] torus3d 5,2 repeat 5 +add3d[-1] 10,0,0 mul3d[-1] 1.2 color3d[-1] ${-rgb} done add3d

 n3d:
     Shortcut for command 'normalize3d'.

 normalize3d:

   Normalize selected 3D objects to unit size.
   (equivalent to shortcut command 'n3d').

   Example:
     [#1] repeat 100 circle3d {u(3)},{u(3)},{u(3)},0.1 done add3d color3d[-1] 255,0,0 +normalize3d[-1] color3d[-1] 0,255,0 add3d

 o3d (+):
     Shortcut for command 'opacity3d'.

 opacity3d (+):
     _opacity

   Set opacity of selected 3D objects.
   (equivalent to shortcut command 'o3d').

   Default value: 'opacity=1'.

   Example:
     [#1] torus3d 100,10 double3d 0 repeat 7 +rotate3d[-1] 1,0,0,20 opacity3d[-1] {u} done add3d

 parametric3d:
     _x(a,b),_y(a,b),_z(a,b),_amin,_amax,_bmin,_bmax,_res_a>0,_res_b>0,_res_x>0,_res_y>0,_res_z>0,_smoothness>=0,_isovalue>=0

   Input 3D object from specified parametric surface '(a,b) ⟶ (x(a,b),y(a,b),z(a,b))'.

   Default values: 'x=(2+cos(b))*sin(a)', 'y=(2+cos(b))*cos(a)', 'c=sin(b)', 'amin=-pi', 'amax=pi', 'bmin=-pi', 'bmax=pi', 'res_a=512',
    'res_b=res_a', 'res_x=64', 'res_y=res_x', 'res_z=res_y', 'smoothness=2%' and 'isovalue=10%'.

   Example:
     [#1] parametric3d ,

 pca_patch3d:
     _patch_size>0,_M>0,_N>0,_normalize_input={ 0 | 1 },_normalize_output={ 0 | 1 },_lambda_xy

   Get 3D patch-pca representation of selected images.
   The 3D patch-pca is estimated from M patches on the input image, and displayed as a cloud of N 3D points.

   Default values: 'patch_size=7', 'M=1000', 'N=3000', 'normalize_input=1', 'normalize_output=0', and 'lambda_xy=0'.

   Example:
     [#1] image.jpg pca_patch3d 7

 plane3d:
     _size_x,_size_y,_nb_subdivisions_x>0,_nb_subdisivions_y>0

   Input 3D plane at (0,0,0), with specified geometry.

   Default values: 'size_x=1', 'size_y=size_x' and 'nb_subdivisions_x=nb_subdivisions_y=24'.

   Example:
     [#1] plane3d 50,30 +primitives3d 1 color3d[-2] ${-rgb}

 point3d:
     x0,y0,z0

   Input 3D point at specified coordinates.

   Example:
     [#1] repeat 1000 a={$>*pi/500} point3d {cos(3*$a)},{sin(2*$a)},0 color3d[-1] ${-rgb} done add3d

 pointcloud3d:

   Convert selected planar or volumetric images to 3D point clouds.

   Example:
     [#1] image.jpg luminance resize2dy 100 threshold 50% mul 255 pointcloud3d color3d[-1] 255,255,255

 pose3d:
     p1,...,p12

   Apply 3D pose matrix to selected 3D objects.

   Example:
     [#1] torus3d 100,20 pose3d 0.152437,1.20666,-0.546366,0,-0.535962,0.559129,1.08531,0,1.21132,0.0955431,0.548966,0,0,0,-206,1 snapshot3d 400

 p3d:
     Shortcut for command 'primitives3d'.

 primitives3d:
     mode

   Convert primitives of selected 3D objects.
   (equivalent to shortcut command 'p3d').

   'mode' can be { 0=points | 1=outlines | 2=non-textured }.

   Example:
     [#1] sphere3d 30 primitives3d 1 torus3d 50,10 color3d[-1] ${-rgb} add3d

 projections3d:
     _x[%],_y[%],_z[%],_is_bounding_box={ 0 | 1 }

   Generate 3D xy,xz,yz projection planes from specified volumetric images.

 pyramid3d:
     width,height

   Input 3D pyramid at (0,0,0), with specified geometry.

   Example:
     [#1] pyramid3d 100,-100 +primitives3d 1 color3d[-2] ${-rgb}

 quadrangle3d:
     x0,y0,z0,x1,y1,z1,x2,y2,z2,x3,y3,z3

   Input 3D quadrangle at specified coordinates.

   Example:
     [#1] quadrangle3d -10,-10,10,10,-10,10,10,10,10,-10,10,10 repeat 10 +rotate3d[-1] 0,1,0,30 color3d[-1] ${-rgb},0.6 done add3d mode3d 2

 random3d:
     nb_points>=0

   Input random 3D point cloud in [0,1]^3.

   Example:
     [#1] random3d 100 circles3d 0.1 opacity3d 0.5

 rv3d (+):
     Shortcut for command 'reverse3d'.

 reverse3d (+):

   Reverse primitive orientations of selected 3D objects.
   (equivalent to shortcut command 'rv3d').

   Example:
     [#1] torus3d 100,40 double3d 0 +reverse3d

 r3d (+):
     Shortcut for command 'rotate3d'.

 rotate3d (+):
     u,v,w,angle

   Rotate selected 3D objects around specified axis with specified angle (in deg.).
   (equivalent to shortcut command 'r3d').

   Example:
     [#1] torus3d 100,10 double3d 0 repeat 7 +rotate3d[-1] 1,0,0,20 done add3d

 rotation3d:
     u,v,w,angle

   Input 3x3 rotation matrix with specified axis and angle (in deg).

   Example:
     [#1] rotation3d 1,0,0,0 rotation3d 1,0,0,90 rotation3d 1,0,0,180

 sierpinski3d:
     _recursion_level>=0,_width,_height

   Input 3d Sierpinski pyramid.

   Example:
     [#1] sierpinski3d 3,100,-100 +primitives3d 1 color3d[-2] ${-rgb}

 size3d:

   Return bounding box size of the last selected 3D object.

 skeleton3d:
     _metric,_frame_type={ 0=squares | 1=diamonds | 2=circles | 3=auto },_skeleton_opacity,_frame_opacity,_is_frame_wireframe={ 0 | 1 }

   Build 3D skeletal structure object from 2d binary shapes located in selected images.
   'metric' can be { 0=chebyshev | 1=manhattan | 2=euclidean }.

   Default values: 'metric=2', 'bones_type=3', 'skeleton_opacity=1' and 'frame_opacity=0.1'.

   Example:
     [#1] shape_cupid 480 +skeleton3d ,

 snapshot3d:
     _size>0,_zoom>=0,_backgroundR,_backgroundG,_backgroundB,_backgroundA |
     [background_image],zoom>=0

   Take 2d snapshots of selected 3D objects.
   Set 'zoom' to 0 to disable object auto-scaling.

   Default values: 'size=512', 'zoom=1' and '[background_image]=(default)'.

   Example:
     [#1] torus3d 100,20 rotate3d 1,1,0,60 snapshot3d 400,1.2,128,64,32
     [#2] torus3d 100,20 rotate3d 1,1,0,60 sample ? +snapshot3d[0] [1],1.2

 sl3d (+):
     Shortcut for command 'specl3d'.

 specl3d (+):
     value>=0

   Set lightness of 3D specular light.
   (equivalent to shortcut command 'sl3d').

   Default value: 'value=0.15'.

   Example:
     [#1] (0,0.3,0.6,0.9,1.2) repeat w torus3d 100,30 rotate3d[-1] 1,1,0,60 color3d[-1] 255,0,0 specl3d {0,@$>} snapshot3d[-1] 400 done remove[0]

 ss3d (+):
     Shortcut for command 'specs3d'.

 specs3d (+):
     value>=0

   Set shininess of 3D specular light.
   (equivalent to shortcut command 'ss3d').

   Default value: 'value=0.8'.

   Example:
     [#1] (0,0.3,0.6,0.9,1.2) repeat w torus3d 100,30 rotate3d[-1] 1,1,0,60 color3d[-1] 255,0,0 specs3d {0,@$>} snapshot3d[-1] 400 done remove[0]

 sphere3d (+):
     radius,_nb_recursions>=0

   Input 3D sphere at (0,0,0), with specified geometry.

   Default value: 'nb_recursions=3'.

   Example:
     [#1] sphere3d 100 +primitives3d 1 color3d[-2] ${-rgb}

 spherical3d:
     _nb_azimuth>=3,_nb_zenith>=3,_radius_function(phi,theta)

   Input 3D spherical object at (0,0,0), with specified geometry.

   Default values: 'nb_zenith=nb_azimut=64' and 'radius_function="abs(1+0.5*cos(3*phi)*sin(4*theta))"'.

   Example:
     [#1] spherical3d 64 +primitives3d 1

 spline3d:
     x0[%],y0[%],z0[%],u0[%],v0[%],w0[%],x1[%],y1[%],z1[%],u1[%],v1[%],w1[%],_nb_vertices>=2

   Input 3D spline with specified geometry.

   Default values: 'nb_vertices=128'.

   Example:
     [#1] repeat 100 spline3d {u},{u},{u},{u},{u},{u},{u},{u},{u},{u},{u},{u},128 color3d[-1] ${-rgb} done box3d 1 primitives3d[-1] 1 add3d

 s3d (+):
     Shortcut for command 'split3d'.

 split3d (+):
     _full_split={ 0 | 1 }

   Split selected 3D objects into feature vectors :
    * If 'full_split==0', { header, sizes, vertices, primitives, colors, opacities }.
    * If 'full_split==1', { header, sizes, vertices, p0,...,pP, c0,...,cP, o0,...,oP }.
   (equivalent to shortcut command 's3d').

   To recreate the 3D object, append all produced images along the y-axis.

   Default value: 'full_split=0'.

   Example:
     [#1] box3d 100 +split3d

 sprite3d:

   Convert selected images as 3D sprites.
   Selected images with alpha channels are managed.

   Example:
     [#1] image.jpg sprite3d

 sprites3d:
     [sprite],_sprite_has_alpha_channel={ 0 | 1 }

   Convert selected 3D objects as a sprite cloud.
   Set 'sprite_has_alpha_channel' to 1 to make the last channel of the selected sprite be a transparency mask.

   Default value: 'mask_has_alpha_channel=0'.

   Example:
     [#1] torus3d 100,20 image.jpg resize2dy[-1] 64 100%,100% gaussian[-1] 30%,30% *[-1] 255 append[-2,-1] c +sprites3d[0] [1],1 display_rgba[-2]

 star3d:
     _nb_branches>0,0<=_thickness<=1

   Input 3D star at position '(0,0,0)', with specified geometry.

   Default values: 'nb_branches=5' and 'thickness=0.38'.

   Example:
     [#1] star3d , +primitives3d 1 color3d[-2] ${-rgb}

 streamline3d (+):
     x[%],y[%],z[%],_L>=0,_dl>0,_interpolation,_is_backward={ 0 | 1 },_is_oriented={ 0 | 1 } |
     'formula',x,y,z,_L>=0,_dl>0,_interpolation,_is_backward={ 0 | 1 },_is_oriented={ 0 | 1 }

   Extract 3D streamlines from selected vector fields or from specified formula.
   'interpolation' can be { 0=nearest integer | 1=1st-order | 2=2nd-order | 3=4th-order }.

   Default values: 'dl=0.1', 'interpolation=2', 'is_backward=0' and 'is_oriented=0'.

   Example:
     [#1] 100,100,100,3 rand -10,10 blur 3 repeat 300 +streamline3d[0] {u(100)},{u(100)},{u(100)},1000,1,1 color3d[-1] ${-rgb} done remove[0] box3d 100 primitives3d[-1] 1 add3d

 -3d (+):
     Shortcut for command 'sub3d'.

 sub3d (+):
     tx,_ty,_tz

   Shift selected 3D objects with the opposite of specified displacement vector.
   (equivalent to shortcut command '3d').

   Default values: 'ty=tz=0'.

   Example:
     [#1] sphere3d 10 repeat 5 +sub3d[-1] 10,{u(-10,10)},0 color3d[-1] ${-rgb} done add3d

 superformula3d:
     resolution>1,m>=1,n1,n2,n3

   Input 2D superformula curve as a 3D object.

   Default values: 'resolution=1024', 'm=8', 'n1=1', 'n2=5' and 'n3=8'.

   Example:
     [#1] superformula3d ,

 tensors3d:
     _radius_factor>=0,_shape={ 0=box | >=N=ellipsoid },_radius_min>=0

   Generate 3D tensor fields from selected images.
   when 'shape'>0, it gives the ellipsoid shape precision.

   Default values: 'radius_factor=1', 'shape=2' and 'radius_min=0.05'.

   Example:
     [#1] 6,6,6,9,"U = [x,y,z] - [w,h,d]/2; U/=norm(U); mul(U,U,3) + 0.3*eye(3)" tensors3d 0.8

 text_pointcloud3d:
     _"text1",_"text2",_smoothness

   Input 3D text pointcloud from the two specified strings.

   Default values: 'text1="text1"', 'text2="text2"' and 'smoothness=1'.

   Example:
     [#1] text_pointcloud3d "G'MIC","Rocks!"

 text3d:
     text,_font_height>0,_depth>0,_smoothness

   Input a 3D text object from specified text.

   Default values: 'font_height=53', 'depth=10' and 'smoothness=1.5'.

   Example:
     [#1] text3d "G'MIC as a0D logo!"

 t3d:
     Shortcut for command 'texturize3d'.

 texturize3d:
     [ind_texture],_[ind_coords]

   Texturize selected 3D objects with specified texture and coordinates.
   (equivalent to shortcut command 't3d').

   When '[ind_coords]' is omitted, default XY texture projection is performed.

   Default value: 'ind_coords=(undefined)'.

   Example:
     [#1] image.jpg torus3d 100,30 texturize3d[-1] [-2] keep[-1]

 torus3d:
     _radius1,_radius2,_nb_subdivisions1>2,_nb_subdivisions2>2

   Input 3D torus at (0,0,0), with specified geometry.

   Default values: 'radius1=1', 'radius2=0.3', 'nb_subdivisions1=24' and 'nb_subdivisions2=12'.

   Example:
     [#1] torus3d 10,3 +primitives3d 1 color3d[-2] ${-rgb}

 triangle3d:
     x0,y0,z0,x1,y1,z1,x2,y2,z2

   Input 3D triangle at specified coordinates.

   Example:
     [#1] repeat 100 a={$>*pi/50} triangle3d 0,0,0,0,0,3,{cos(3*$a)},{sin(2*$a)},0 color3d[-1] ${-rgb} done add3d

 volume3d:

   Transform selected 3D volumetric images as 3D parallelepipedic objects.

   Example:
     [#1] image.jpg animate blur,0,5,30 append z volume3d

 weird3d:
     _resolution>0

   Input 3D weird object at (0,0,0), with specified resolution.

   Default value: 'resolution=32'.

   Example:
     [#1] weird3d 48 +primitives3d 1 color3d[-2] ${-rgb}

 12.13. Control Flow
        ------------

 ap:
     Shortcut for command 'apply_parallel'.

 apply_parallel:
     "command"

   Apply specified command on each of the selected images, by parallelizing it for all image of the list.
   (equivalent to shortcut command 'ap').

   Example:
     [#1] image.jpg +mirror x +mirror y apply_parallel "blur 3"

 apc:
     Shortcut for command 'apply_parallel_channels'.

 apply_parallel_channels:
     "command"

   Apply specified command on each of the selected images, by parallelizing it for all channel
   of the images independently.
   (equivalent to shortcut command 'apc').

   Example:
     [#1] image.jpg apply_parallel_channels "blur 3"

 apo:
     Shortcut for command 'apply_parallel_overlap'.

 apply_parallel_overlap:
     "command",overlap[%],nb_threads={ 0=auto | 1 | 2 | 4 | 8 | 16 }

   Apply specified command on each of the selected images, by parallelizing it on 'nb_threads'
   overlapped sub-images.
   (equivalent to shortcut command 'apo').

   'nb_threads' must be a power of 2.

   Default values: 'overlap=0','nb_threads=0'.

   Example:
     [#1] image.jpg +apply_parallel_overlap "smooth 500,0,1",1

 at:
     Shortcut for command 'apply_tiles'.

 apply_tiles:
     "command",_tile_width[%]>0,_tile_height[%]>0,_tile_depth[%]>0,_overlap_width[%]>=0,_overlap_height[%]>=0,_overlap_depth[%]>=0,_boundary_conditions={ 0=dirichlet | 1=neumann | 2=periodic |
       3=mirror }

   Apply specified command on each tile (neighborhood) of the selected images, eventually with overlapping tiles.
   (equivalent to shortcut command 'at').

   Default values: 'tile_width=tile_height=tile_depth=10%','overlap_width=overlap_height=overlap_depth=0' and 'boundary_conditions=1'.

   Example:
     [#1] image.jpg +equalize[0] 256 +apply_tiles[0] "equalize 256",16,16,1,50%,50%

 apply_timeout:
     "command",_timeout={ 0=no timeout | >0=with specified timeout (in seconds) }

   Apply a command with a timeout.
   Set variable '$_is_timeout' to '1' if timeout occurred, '0' otherwise.

   Default value: 'timeout=20'.

 check (+):
     condition

   Evaluate specified condition and display an error message if evaluated to false.
   If 'expression' is not a math expression, it is regarded as a filename and checked if it exists.

 check3d (+):
     _is_full_check={ 0 | 1 }

   Check validity of selected 3D vector objects, and display an error message
   if one of the selected images is not a valid 3D vector object.
   Full 3D object check is slower but more precise.

   Default value: 'is_full_check=1'.

 continue (+):

   Go to end of current 'repeat...done', 'do...while' or 'local...endlocal' block.

   Example:
     [#1] image.jpg repeat 10 blur 1 if 1==1 continue fi deform 10 done

 break (+):

   Break current 'repeat...done', 'do...while' or 'local...endlocal' block.

   Example:
     [#1] image.jpg repeat 10 blur 1 if 1==1 break fi deform 10 done

 do (+):

   Start a 'do...while' block.

   Example:
     [#1] image.jpg luminance i={ia+2} do set 255,{u(100)}%,{u(100)}% while ia<$i

 done (+):

   End a 'repeat/for...done' block, and go to associated 'repeat/for' position, if iterations remain.

 elif (+):
     condition

   Start a 'elif...[else]...fi' block if previous 'if' was not verified
   and test if specified condition holds
   'condition' is a mathematical expression, whose evaluation is interpreted as { 0=false | other=true }..

   Tutorial: https://gmic.eu/tutorial/iffi

 else (+):

   Execute following commands if previous 'if' or 'elif' conditions failed.

   Tutorial: https://gmic.eu/tutorial/iffi

 fi (+):

   End a 'if...[elif]...[else]...fi' block.
   (equivalent to shortcut command 'fi').

   Tutorial: https://gmic.eu/tutorial/iffi

 endl (+):
     Shortcut for command 'endlocal'.

 endlocal (+):

   End a 'local...endlocal' block.
   (equivalent to shortcut command 'endl').

 error (+):
     message

   Print specified error message on the standard error (stderr) and exit interpreter, except
   if error is caught by a 'onfail' command.
   Command selection (if any) stands for displayed call stack subset instead of image indices.

 eval (+):
     expression

   Evaluate specified math expression.
    * If no command selection is specified, the expression is evaluated once and its result is set to status.
    * If command selection is specified, the evaluation is looped over selected images. Status is not modified.
     (in this latter case, 'eval' is similar to 'fill' without assigning the image values).

 x (+):
     Shortcut for command 'exec'.

 exec (+):
     _is_verbose={ 0 | 1 },"command"

   Execute external command using a system call.
   The status value is then set to the error code returned by the system call.
   If 'is_verbose=1', the executed command is allowed to output on stdout/stderr.
   (equivalent to shortcut command 'x').

   Default value: 'is_verbose=1'.

 xo:
     Shortcut for command 'exec_out'.

 exec_out:
     _mode,"command"

   Execute external command using a system call, and return resulting 'stdout' and/or 'stderr'.
   'mode' can be { 0=stdout | 1=stderr | 2=stdout+stderr }.

 for (+):
     condition

   Start a 'for...done' block.

   Example:
     [#1] image.jpg resize2dy 32 400,400,1,3 x=0 for $x<400 image[1] [0],$x,$x x+=40 done

 if (+):
     condition

   Start a 'if...[elif]...[else]...fi' block and test if specified condition holds.
   'condition' is a mathematical expression, whose evaluation is interpreted as { 0=false | other=true }.

   Example:
     [#1] image.jpg if ia<64 add 50% elif ia<128 add 25% elif ia<192 sub 25% else sub 50% fi cut 0,255

   Tutorial: https://gmic.eu/tutorial/iffi

 l (+):
     Shortcut for command 'local'.

 local (+):

   Start a 'local...[onfail]...endlocal' block, with selected images.
   (equivalent to shortcut command 'l').

   Example:
     [#1] image.jpg local[] 300,300,1,3 rand[0] 0,255 blur 4 sharpen 1000 endlocal
     [#2] image.jpg +local repeat 3 deform 20 done endlocal

   Tutorial: https://gmic.eu/oldtutorial/_local

 mutex (+):
     index,_action={ 0=unlock | 1=lock }

   Lock or unlock specified mutex for multi-threaded programming.
   A locked mutex can be unlocked only by the same thread. All mutexes are unlocked by default.
   'index' designates the mutex index, in [0,255].

   Default value: 'action=1'.

 noarg (+):

   Used in a custom command, 'noarg' tells the command that its argument list have not been used
   finally, and so they must be evaluated next in the G'MIC pipeline, just as if the custom
   command takes no arguments at all.
   Use this command to write a custom command which can decide if it takes arguments or not.

 onfail (+):

   Execute following commands when an error is encountered in the body of the 'local...endlocal' block.
   The status value is set with the corresponding error message.

   Example:
     [#1] image.jpg +local blur -3 onfail mirror x endlocal

 parallel (+):
     _wait_threads,"command1","command2",...

   Execute specified commands in parallel, each in a different thread.
   Parallel threads share the list of images.
   'wait_threads' can be { 0=when current environment ends | 1=immediately }.

   Default value: 'wait_threads=1'.

   Example:
     [#1] image.jpg [0] parallel "blur[0] 3","mirror[1] c"

 progress (+):
     0<=value<=100 |
     -1

   Set the progress index of the current processing pipeline.
   This command is useful only when G'MIC is used by an embedding application.

 q (+):
     Shortcut for command 'quit'.

 quit (+):

   Quit G'MIC interpreter.
   (equivalent to shortcut command 'q').

 repeat (+):
     nb_iterations,_variable_name

   Start 'nb_iterations' iterations of a 'repeat...done' block.
   'nb_iterations' is a mathematical expression that will be evaluated.

   Example:
     [#1] image.jpg split y repeat $!,n shift[$n] $<,0,0,0,2 done append y
     [#2] image.jpg mode3d 2 repeat 4 imagecube3d rotate3d 1,1,0,40 snapshot3d 400,1.4 done

   Tutorial: https://gmic.eu/oldtutorial/_repeat

 return (+):

   Return from current custom command.

 rprogress:
     0<=value<=100 | -1 | "command",0<=value_min<=100,0<=value_max<=100

   Set the progress index of the current processing pipeline (relatively to
   previously defined progress bounds), or call the specified command with
   specified progress bounds.

 run:
     "G'MIC pipeline"

   Run specified G'MIC pipeline.
   This is only useful when used from a shell, e.g. to avoid shell substitutions to happen in argument.

 skip (+):
     item

   Do nothing but skip specified item.

 u (+):
     Shortcut for command 'status'.

 status (+):
     status_string

   Set the current status. Used to define a returning value from a function.
   (equivalent to shortcut command 'u').

   Example:
     [#1] image.jpg command "foo : u0=Dark u1=Bright status ${u{ia>=128}}" text_outline ${-foo},2,2,23,2,1,255

 while (+):
     condition

   End a 'do...while' block and go back to associated 'do' if specified condition holds.
   'condition' is a mathematical expression, whose evaluation is interpreted as { 0=false | other=true }.

 12.14. Neural Networks
        ---------------

 nn_lib ::

   Return the list of library functions that has to be included in a math expression,in order to use the neural network library.

 nn_init:

   Initialize a new network.

 nn_check_layer:
     name

   Check that the layer with specified name exists in the network.

 nn_save:
     'filename.gmz'

   Save current network as a .gmz file.
   Neural network files can be only saved in .gmz format.

 nn_load:
     'filename.gmz'

   Load and initialize network saved as a .gmz file.
   Neural network files can be only loaded in .gmz format.

 nn_layer_input:
     name,width,_height,_depth,_spectrum

   Add an 'input' layer to the network.

   Default values: 'height=1', 'depth=1' and 'spectrum=1'.

 nn_layer_add:
     name,in0,in1

   Add an 'add' layer to the network.

 nn_layer_append:
     name,in0,in1

   Add an 'append' layer to the network.

 nn_layer_avgpool2d:
     name,in

   Add a 'avgpool2d' layer (2d average pooling) to the network.

 nn_layer_batchnorm:
     name,in,learning_mode.

   Add a 'batchnorm' layer to the network.
   'learning_mode' can be { 0=no parameters learnt | 1=gamma-only | 2=beta-only | 3=gamma & beta}.

 nn_layer_clone:
     name0,name1,in

   Add a 'clone' layer to the network.

 nn_layer_conv2d:
     name,in,nb_channels>0,_kernel_size>0,_stride>0,_dilation>0

   Add a 'conv2d' layer (2D convolutional layer) to the network.

   Default values: 'kernel_size=3', 'stride=1' and 'dilation=1'.

 nn_layer_conv2dnl:
     name,in,nb_channels>0,_kernel_size>0,_stride>0,_dilation>0,_activation

   Add a 'con2dnl' (2D convolutional layer followed by a non-linearity), to the network.

   Default values: 'kernel_size=3', 'stride=1', 'dilation=1' and 'activation=leakyrelu'.

 nn_layer_fc:
     name,in,nb_channels>0

   Add a 'fc' layer (fully connected layer) to the network.

 nn_layer_fcnl:
     name,in,nb_neurons>0,_activation

   Add a 'fcnl' layer (fully connected layer followed by a non-linearity), to the network.

   Default value: 'activation=leakyrelu'.

 nn_layer_maxpool2d:
     name,in

   Add a 'maxpool2d' layer (2d max pooling) to the network.

 nn_layer_nl:
     name,in,activation

   Add a 'nl' layer (non-linearity) to the network.

   Default value: 'activation=leakyrelu'.

 nn_layer_rename:
     name,in

   Add a 'rename' layer to the network.

 nn_layer_reshape:
     name,in,width>0,height>0,depth>0,spectrum>0

   Add a 'reshape' layer to the network.

 nn_layer_split:
     name0,name1,in,nb_channels0

   Add a 'split' layer to the network.

 nn_layer_upsample2d:
     name,in,_interpolation_type = { 0=nearest-neighbor | 1=linear }

   Add a 'upsample2d' layer to the network.

 nn_loss_mse:
     name,in,ground_truth,_learning_rate,_is_adaptive_learning_rate={ 0=false | 1=true }

   Add a 'mse' loss to the network.

   Default value: 'learning_rate=1e-6' and 'is_adaptive_learning_rate=1'.

 12.15. Arrays, Tiles and Frames
        ------------------------

 array:
     M>0,_N>0,_expand_type={ 0=min | 1=max | 2=all }

   Create MxN array from selected images.

   Default values: 'N=M' and 'expand_type=0'.

   Example:
     [#1] image.jpg array 3,2,2

 array_fade:
     M>0,_N>0,0<=_fade_start<=100,0<=_fade_end<=100,_expand_type={0=min | 1=max | 2=all}

   Create MxN array from selected images.

   Default values: 'N=M', 'fade_start=60', 'fade_end=90' and 'expand_type=1'.

   Example:
     [#1] image.jpg array_fade 3,2

 array_mirror:
     N>=0,_dir={ 0=x | 1=y | 2=xy | 3=tri-xy },_expand_type={ 0 | 1 }

   Create 2^Nx2^N array from selected images.

   Default values: 'dir=2' and 'expand_type=0'.

   Example:
     [#1] image.jpg array_mirror 2

 array_random:
     Ms>0,_Ns>0,_Md>0,_Nd>0

   Create MdxNd array of tiles from selected MsxNs source arrays.

   Default values: 'Ns=Ms', 'Md=Ms' and 'Nd=Ns'.

   Example:
     [#1] image.jpg +array_random 8,8,15,10

 frame:
     Shortcut for command 'frame_xy'.

 frame_blur:
     _sharpness>0,_size>=0,_smoothness,_shading,_blur

   Draw RGBA-colored round frame in selected images.

   Default values: 'sharpness=10', 'size=30', 'smoothness=0', 'shading=1' and 'blur=3%'.

   Example:
     [#1] image.jpg frame_blur 3,30,8,10%

 frame_cube:
     _depth>=0,_centering_x,_centering_y,_left_side={0=normal | 1=mirror-x | 2=mirror-y | 3=mirror-xy},_right_side,_lower_side,_upper_side

   Insert 3D frames in selected images.

   Default values: 'depth=1', 'centering_x=centering_y=0' and 'left_side=right_side,lower_side=upper_side=0'.

   Example:
     [#1] image.jpg frame_cube ,

 frame_fuzzy:
     size_x[%]>=0,_size_y[%]>=0,_fuzzyness>=0,_smoothness[%]>=0,_R,_G,_B,_A

   Draw RGBA-colored fuzzy frame in selected images.

   Default values: 'size_y=size_x', 'fuzzyness=5', 'smoothness=1' and 'R=G=B=A=255'.

   Example:
     [#1] image.jpg frame_fuzzy 20

 frame_painting:
     _size[%]>=0,0<=_contrast<=1,_profile_smoothness[%]>=0,_R,_G,_B,_vignette_size[%]>=0,_vignette_contrast>=0,_defects_contrast>=0,0<=_defects_density<=100,_defects_size>=0,_defects_smoothness[%]>=0,
       _serial_number

   Add a painting frame to selected images.

   Default values: 'size=10%', 'contrast=0.4', 'profile_smoothness=6%', 'R=225', 'G=200', 'B=120', 'vignette_size=2%', 'vignette_contrast=400',
    'defects_contrast=50', 'defects_density=10', 'defects_size=1', 'defects_smoothness=0.5%' and 'serial_number=123456789'.

   Example:
     [#1] image.jpg frame_painting ,

 frame_pattern:
     M>=3,_constrain_size={ 0 | 1 } |
     M>=3,_[frame_image],_constrain_size={ 0 | 1 }

   Insert selected pattern frame in selected images.

   Default values: 'pattern=0' and 'constrain_size=0'.

   Example:
     [#1] image.jpg frame_pattern 8

 frame_round:
     _sharpness>0,_size>=0,_smoothness,_shading,_R,_G,_B,_A

   Draw RGBA-colored round frame in selected images.

   Default values: 'sharpness=10', 'size=10', 'smoothness=0', 'shading=0' and 'R=G=B=A=255'.

   Example:
     [#1] image.jpg frame_round 10

 frame_seamless:
     frame_size>=0,_patch_size>0,_blend_size>=0,_frame_direction={ 0=inner (preserve image size) | 1=outer }

   Insert frame in selected images, so that tiling the resulting image makes less visible seams.

   Default values: 'patch_size=7', 'blend_size=5' and 'frame_direction=1'.

   Example:
     [#1] image.jpg +frame_seamless 30 array 2,2

 frame_x:
     size_x[%],_col1,...,_colN

   Insert colored frame along the x-axis in selected images.

   Default values: 'col1=col2=col3=255' and 'col4=255'.

   Example:
     [#1] image.jpg frame_x 20,255,0,255

 frame_xy:
     size_x[%],_size_y[%],_col1,...,_colN

   Insert colored frame along the x-axis in selected images.

   Default values: 'size_y=size_x', 'col1=col2=col3=255' and 'col4=255'.
   (equivalent to shortcut command 'frame').

   Example:
     [#1] image.jpg frame_xy 1,1,0 frame_xy 20,10,255,0,255

 frame_xyz:
     size_x[%],_size_y[%],_size_z[%]_col1,...,_colN

   Insert colored frame along the x-axis in selected images.

   Default values: 'size_y=size_x=size_z', 'col1=col2=col3=255' and 'col4=255'.

 frame_y:
     size_y[%],_col1,...,_colN

   Insert colored frame along the y-axis in selected images.

   Default values: 'col1=col2=col3=255' and 'col4=255'.

   Example:
     [#1] image.jpg frame_y 20,255,0,255

 img2ascii:
     _charset,_analysis_scale>0,_analysis_smoothness[%]>=0,_synthesis_scale>0,_output_ascii_filename

   Render selected images as binary ascii art.
   This command returns the corresponding the list of widths and heights (expressed as a number of characters)
   for each selected image.

   Default values: 'charset=[ascii charset]', 'analysis_scale=16', 'analysis_smoothness=20%', 'synthesis_scale=16' and '_output_ascii_filename=[undefined]'.

   Example:
     [#1] image.jpg img2ascii ,

 imagegrid:
     M>0,_N>0

   Create MxN image grid from selected images.

   Default value: 'N=M'.

   Example:
     [#1] image.jpg imagegrid 16

 imagegrid_hexagonal:
     _resolution>0,0<=_outline<=1

   Create hexagonal grids from selected images.

   Default values: 'resolution=32', 'outline=0.1' and 'is_antialiased=1'.

   Example:
     [#1] image.jpg imagegrid_hexagonal 24

 imagegrid_triangular:
     pattern_width>=1,_pattern_height>=1,_pattern_type,0<=_outline_opacity<=1,_outline_color1,...

   Create triangular grids from selected images.
   'pattern type' can be { 0=horizontal | 1=vertical | 2=crossed | 3=cube | 4=decreasing | 5=increasing }.

   Default values: 'pattern_width=24', 'pattern_height=pattern_width', 'pattern_type=0', 'outline_opacity=0.1' and 'outline_color1=0'.

   Example:
     [#1] image.jpg imagegrid_triangular 6,10,3,0.5

 linearize_tiles:
     M>0,_N>0

   Linearize MxN tiles on selected images.

   Default value: 'N=M'.

   Example:
     [#1] image.jpg +linearize_tiles 16

 map_sprites:
     _nb_sprites>=1,_allow_rotation={ 0=none | 1=90 deg. | 2=180 deg. }

   Map set of sprites (defined as the 'nb_sprites' latest images of the selection) to other selected images,
   according to the luminosity of their pixel values.

   Example:
     [#1] image.jpg resize2dy 48 repeat 16 ball {8+2*$>},${-rgb} mul[-1] {(1+$>)/16} done map_sprites 16

 pack:
     is_ratio_constraint={ 0 | 1 },_sort_criterion

   Pack selected images into a single image.
   The returned status contains the list of new (x,y) offsets for each input image.
   Parameter 'is_ratio_constraint' tells if the resulting image must tend to a square image.

   Default values: 'is_ratio_constraint=0' and 'sort_criterion=max(w,h)'.

   Example:
     [#1] image.jpg repeat 10 +resize2dx[-1] 75% balance_gamma[-1] ${-rgb} done pack 0

 puzzle:
     _width>0,_height>0,_M>=1,_N>=1,_curvature,_centering,_connectors_variability,_resolution>=1

   Input puzzle binary mask with specified size and geometry.

   Default values: 'width=height=512', 'M=N=5', 'curvature=0.5', 'centering=0.5', 'connectors_variability=0.5' and 'resolution=64'.

   Example:
     [#1] puzzle ,

 quadratize_tiles:
     M>0,_N>0

   Quadratize MxN tiles on selected images.

   Default value: 'N=M'.

   Example:
     [#1] image.jpg +quadratize_tiles 16

 rotate_tiles:
     angle,_M>0,N>0

   Apply MxN tiled-rotation effect on selected images.

   Default values: 'M=8' and 'N=M'.

   Example:
     [#1] image.jpg to_rgba rotate_tiles 10,8 drop_shadow 10,10 display_rgba

 shift_tiles:
     M>0,_N>0,_amplitude

   Apply MxN tiled-shift effect on selected images.

   Default values: 'N=M' and 'amplitude=20'.

   Example:
     [#1] image.jpg +shift_tiles 8,8,10

 taquin:
     M>0,_N>0,_remove_tile={ 0=none | 1=first | 2=last | 3=random },_relief,_border_thickness[%],_border_outline[%],_outline_color

   Create MxN taquin puzzle from selected images.

   Default value: 'N=M', 'relief=50', 'border_thickness=5', 'border_outline=0' and 'remove_tile=0'.

   Example:
     [#1] image.jpg +taquin 8

 tunnel:
     _level>=0,_factor>0,_centering_x,_centering_y,_opacity,_angle

   Apply tunnel effect on selected images.

   Default values: 'level=9', 'factor=80%', 'centering_x=centering_y=0.5', 'opacity=1' and 'angle=0'

   Example:
     [#1] image.jpg tunnel 20

 12.16. Artistic
        --------

 boxfitting:
     _min_box_size>=1,_max_box_size>=0,_initial_density>=0,_nb_attempts>=1

   Apply box fitting effect on selected images, as displayed the web page:
   http://www.complexification.net/gallery/machines/boxFittingImg/.

   Default values: 'min_box_size=1', 'max_box_size=0', 'initial_density=0.1' and 'nb_attempts=3'.

   Example:
     [#1] image.jpg boxfitting ,

 brushify:
     [brush],_brush_nb_sizes>=1,0<=_brush_min_size_factor<=1,_brush_nb_orientations>=1,_brush_light_type,0<=_brush_light_strength<=1,_brush_opacity,_painting_density[%]>=0,
       0<=_painting_contours_coherence<=1,0<=_painting_orientation_coherence<=1,_painting_coherence_alpha[%]>=0,_painting_coherence_sigma[%]>=0,_painting_primary_angle,
       0<=_painting_angle_dispersion<=1

   Apply specified brush to create painterly versions of specified images.
   'brush_light_type' can be { 0=none | 1=flat | 2=darken | 3=lighten | 4=full }.

   Default values: 'brush_nb_sizes=3', 'brush_min_size_factor=0.66', 'brush_nb_orientations=12', 'brush_light_type=0', 'brush_light_strength=0.25',
    'brush_opacity=0.8', 'painting_density=20%', 'painting_contours_coherence=0.9', 'painting_orientation_coherence=0.9', 'painting_coherence_alpha=1',
    'painting_coherence_sigma=1', 'painting_primary_angle=0', 'painting_angle_dispersion=0.2'

   Example:
     [#1] image.jpg 40,40 gaussian[-1] 10,4 spread[-1] 10,0 brushify[0] [1],1

 cartoon:
     _smoothness,_sharpening,_threshold>=0,_thickness>=0,_color>=0,quantization>0

   Apply cartoon effect on selected images.

   Default values: 'smoothness=3', 'sharpening=150', 'threshold=20', 'thickness=0.25', 'color=1.5' and 'quantization=8'.

   Example:
     [#1] image.jpg cartoon 3,50,10,0.25,3,16

 color_ellipses:
     _count>0,_radius>=0,_opacity>=0

   Add random color ellipses to selected images.

   Default values: 'count=400', 'radius=5' and 'opacity=0.1'.

   Example:
     [#1] image.jpg +color_ellipses ,,0.15

 cubism:
     _density>=0,0<=_thickness<=50,_max_angle,_opacity,_smoothness>=0

   Apply cubism effect on selected images.

   Default values: 'density=50', 'thickness=10', 'max_angle=75', 'opacity=0.7' and 'smoothness=0'.

   Example:
     [#1] image.jpg cubism ,

 draw_whirl:
     _amplitude>=0

   Apply whirl drawing effect on selected images.

   Default value: 'amplitude=100'.

   Example:
     [#1] image.jpg draw_whirl ,

 drawing:
     _amplitude>=0

   Apply drawing effect on selected images.

   Default value: 'amplitude=200'.

   Example:
     [#1] image.jpg +drawing ,

 drop_shadow:
     _offset_x[%],_offset_y[%],_smoothness[%]>=0,0<=_curvature<=1,_expand_size={ 0 | 1 }

   Drop shadow behind selected images.

   Default values: 'offset_x=20', 'offset_y=offset_x', 'smoothness=5', 'curvature=0' and 'expand_size=1'.

   Example:
     [#1] image.jpg drop_shadow 10,20,5,0.5 expand_xy 20,0 display_rgba

 ellipsionism:
     _R>0[%],_r>0[%],_smoothness>=0[%],_opacity,_outline>0,_density>0

   Apply ellipsionism filter to selected images.

   Default values: 'R=10', 'r=3', 'smoothness=1%', 'opacity=0.7', 'outline=8' and 'density=0.6'.

   Example:
     [#1] image.jpg ellipsionism ,

 fire_edges:
     _edges>=0,0<=_attenuation<=1,_smoothness>=0,_threshold>=0,_nb_frames>0,_starting_frame>=0,frame_skip>=0

   Generate fire effect from edges of selected images.

   Default values: 'edges=0.7', 'attenuation=0.25', 'smoothness=0.5', 'threshold=25', 'nb_frames=1', 'starting_frame=20' and 'frame_skip=0'.

   Example:
     [#1] image.jpg fire_edges ,

 fractalize:
     0<=detail_level<=1

   Randomly fractalize selected images.

   Default value: 'detail_level=0.8'

   Example:
     [#1] image.jpg fractalize ,

 glow:
     _amplitude>=0

   Add soft glow on selected images.

   Default value: 'amplitude=1%'.

   Example:
     [#1] image.jpg glow ,

 halftone:
     nb_levels>=2,_size_dark>=2,_size_bright>=2,_shape={ 0=square | 1=diamond | 2=circle | 3=inv-square | 4=inv-diamond | 5=inv-circle },_smoothness[%]>=0

   Apply halftone dithering to selected images.

   Default values: 'nb_levels=5', 'size_dark=8', 'size_bright=8', 'shape=5' and 'smoothnesss=0'.

   Example:
     [#1] image.jpg halftone ,

 hardsketchbw:
     _amplitude>=0,_density>=0,_opacity,0<=_edge_threshold<=100,_is_fast={ 0 | 1 }

   Apply hard B&W sketch effect on selected images.

   Default values: 'amplitude=1000', 'sampling=3', 'opacity=0.1', 'edge_threshold=20' and 'is_fast=0'.

   Example:
     [#1] image.jpg +hardsketchbw 200,70,0.1,10 median[-1] 2 +local reverse blur[-1] 3 blend[-2,-1] overlay endlocal

 hearts:
     _density>=0

   Apply heart effect on selected images.

   Default value: 'density=10'.

   Example:
     [#1] image.jpg hearts ,

 houghsketchbw:
     _density>=0,_radius>0,0<=_threshold<=100,0<=_opacity<=1,_votesize[%]>0

   Apply hough B&W sketch effect on selected images.

   Default values: 'density=100', 'radius=3', 'threshold=100', 'opacity=0.1' and 'votesize=100%'.

   Example:
     [#1] image.jpg +houghsketchbw ,

 lightrays:
     100<=_density<=0,_center_x[%],_center_y[%],_ray_length>=0,_ray_attenuation>=0

   Generate ray lights from the edges of selected images.

   Default values: 'density=50%', 'center_x=50%', 'center_y=50%', 'ray_length=0.9' and 'ray_attenuation=0.5'.

   Example:
     [#1] image.jpg +lightrays , + cut 0,255

 light_relief:
     _ambient_light,_specular_lightness,_specular_size,_darkness,_light_smoothness,_xl,_yl,_zl,_zscale,_opacity_is_heightmap={ 0 | 1 }

   Apply relief light to selected images.
   Default values(s) : 'ambient_light=0.3', 'specular_lightness=0.5', 'specular_size=0.2', 'darkness=0', 'xl=0.2', 'yl=zl=0.5',
   'zscale=1', 'opacity=1' and 'opacity_is_heightmap=0'.

   Example:
     [#1] image.jpg blur 2 light_relief 0.3,4,0.1,0

 linify:
     0<=_density<=100,_spreading>=0,_resolution[%]>0,_line_opacity>=0,_line_precision>0,_mode={ 0=subtractive | 1=additive }

   Apply linify effect on selected images.
   The algorithm is inspired from the one described on the webpage http://linify.me/about.

   Default values: 'density=50', 'spreading=2', 'resolution=40%', 'line_opacity=10', 'line_precision=24' and 'mode=0'.

   Example:
     [#1] image.jpg linify 60

 mosaic:
     0<=_density<=100

   Create random mosaic from selected images.

   Default values: 'density=30'.

   Example:
     [#1] image.jpg mosaic , +fill "I!=J(1) || I!=J(0,1)?[0,0,0]:I"

 old_photo:

   Apply old photo effect on selected images.

   Example:
     [#1] image.jpg old_photo

 pencilbw:
     _size>=0,_amplitude>=0

   Apply B&W pencil effect on selected images.

   Default values: 'size=0.3' and 'amplitude=60'.

   Example:
     [#1] image.jpg pencilbw ,

 pixelsort:
     _ordering={ + | - },_axis={ x | y | z | xy | yx },_[sorting_criterion],_[mask]

   Apply a 'pixel sorting' algorithm on selected images, as described in the page :
   http://satyarth.me/articles/pixel-sorting/.

   Default values: 'ordering=+', 'axis=x' and 'sorting_criterion=mask=(undefined)'.

   Example:
     [#1] image.jpg +norm +ge[-1] 30% +pixelsort[0] +,y,[1],[2]

 polaroid:
     _size1>=0,_size2>=0

   Create polaroid effect in selected images.

   Default values: 'size1=10' and 'size2=20'.

   Example:
     [#1] image.jpg to_rgba polaroid 5,30 rotate 20 drop_shadow , drgba

 polygonize:
     _warp_amplitude>=0,_smoothness[%]>=0,_min_area[%]>=0,_resolution_x[%]>0,_resolution_y[%]>0

   Apply polygon effect on selected images.

   Default values: 'warp_amplitude=300', 'smoothness=2%', 'min_area=0.1%', 'resolution_x=resolution_y=10%'.

   Example:
     [#1] image.jpg image.jpg polygonize 100,10 +fill "I!=J(1) || I!=J(0,1)?[0,0,0]:I"

 poster_edges:
     0<=_edge_threshold<=100,0<=_edge_shade<=100,_edge_thickness>=0,_edge_antialiasing>=0,0<=_posterization_level<=15,_posterization_antialiasing>=0

   Apply poster edges effect on selected images.

   Default values: 'edge_threshold=40', 'edge_shade=5', 'edge_thickness=0.5', 'edge_antialiasing=10', 'posterization_level=12' and
    'posterization_antialiasing=0'.

   Example:
     [#1] image.jpg poster_edges ,

 poster_hope:
     _smoothness>=0

   Apply Hope stencil poster effect on selected images.

   Default value: 'smoothness=3'.

   Example:
     [#1] image.jpg poster_hope ,

 rodilius:
     0<=_amplitude<=100,_0<=thickness<=100,_sharpness>=0,_nb_orientations>0,_offset,_color_mode={ 0=darker | 1=brighter }

   Apply rodilius (fractalius-like) filter on selected images.

   Default values: 'amplitude=10', 'thickness=10', 'sharpness=400', 'nb_orientations=7', 'offset=0' and 'color_mode=1'.

   Example:
     [#1] image.jpg rodilius 12,10,300,10 normalize_local 10,6
     [#2] image.jpg normalize_local 10,16 rodilius 10,4,400,16 smooth 60,0,1,1,4 normalize_local 10,16

 sketchbw:
     _nb_angles>0,_start_angle,_angle_range>=0,_length>=0,_threshold>=0,_opacity,_bgfactor>=0,_density>0,_sharpness>=0,_anisotropy>=0,_smoothness>=0,_coherence>=0,_is_boost={ 0 | 1 },_is_curved={ 0 |
       1 }

   Apply sketch effect to selected images.

   Default values: 'nb_angles=2', 'start_angle=45', 'angle_range=180', 'length=30', 'threshold=3', 'opacity=0.03', 'bgfactor=0', 'density=0.6',
    'sharpness=0.1', 'anisotropy=0.6', 'smoothness=0.25', 'coherence=1', 'is_boost=0' and 'is_curved=1'.

   Example:
     [#1] image.jpg +sketchbw 1 reverse blur[-1] 3 blend[-2,-1] overlay

 sponge:
     _size>0

   Apply sponge effect on selected images.

   Default value: 'size=13'.

   Example:
     [#1] image.jpg sponge ,

 stained_glass:
     _edges[%]>=0, shading>=0, is_thin_separators={ 0 | 1 }

   Generate stained glass from selected images.

   Default values: 'edges=40%', 'shading=0.2' and 'is_precise=0'.

   Example:
     [#1] image.jpg stained_glass 20%,1 cut 0,20

 stars:
     _density[%]>=0,_depth>=0,_size>0,_nb_branches>=1,0<=_thickness<=1,_smoothness[%]>=0,_R,_G,_B,_opacity

   Add random stars to selected images.

   Default values: 'density=10%', 'depth=1', 'size=32', 'nb_branches=5', 'thickness=0.38', 'smoothness=0.5', 'R=G=B=200' and 'opacity=1'.

   Example:
     [#1] image.jpg stars ,

 stencil:
     _radius[%]>=0,_smoothness>=0,_iterations>=0

   Apply stencil filter on selected images.

   Default values: 'radius=3', 'smoothness=1' and 'iterations=8'.

   Example:
     [#1] image.jpg +norm stencil. 2,1,4 +mul rm[0]

 stencilbw:
     _edges>=0,_smoothness>=0

   Apply B&W stencil effect on selected images.

   Default values: 'edges=15' and 'smoothness=10'.

   Example:
     [#1] image.jpg +stencilbw 40,4

 stylize:
     [style_image],_fidelity_finest,_fidelity_coarsest,_fidelity_smoothness_finest>=0,_fidelity_smoothnes_coarsest>=0,0<=_fidelity_chroma<=1,_init_type,_init_resolution>=0,init_max_gradient>=0,
       _patchsize_analysis>0,_patchsize_synthesis>0,_patchsize_synthesis_final>0,_nb_matches_finest>=0,_nb_matches_coarsest>=0,_penalize_repetitions>=0,_matching_precision>=0,_scale_factor>1,
       _skip_finest_scales>=0,_"image_matching_command"

   Transfer colors and textures from specified style image to selected images, using a multi-scale patch-mathing algorithm.
   If instant display window[0] is opened, the steps of the image synthesis are displayed on it.
   'init_type' can be { 0=best-match | 1=identity | 2=randomized }.

   Default values: 'fidelity_finest=0.5', 'fidelity_coarsest=2', 'fidelity_smoothness_finest=3', 'fidelity_smoothness_coarsest=0.5', 'fidelity_chroma=0.1',
    'init_type=0', 'init_resolution=16', 'init_max_gradient=0', 'patchsize_analysis=5', 'patchsize_synthesis=5', 'patchsize_synthesis_final=5',
    'nb_matches_finest=2', 'nb_matchesc_coarsest=30', 'penalize_repetitions=10', 'matching_precision=2', 'scale_factor=1.85', 'skip_finest_scales=0' and
    'image_matching_command'="s c,-3 transfer_pca[0] [2] b[0,2] xy,0.7 n[0,2] 0,255 n[1,2] 0,200 a[0,1] c a[1,2] c"'.

 tetris:
     _scale>0

   Apply tetris effect on selected images.

   Default value: 'scale=10'.

   Example:
     [#1] image.jpg +tetris 10

 warhol:
     _M>0,_N>0,_smoothness>=0,_color>=0

   Create MxN Andy Warhol-like artwork from selected images.

   Default values: 'M=3', 'N=M', 'smoothness=2' and 'color=20'.

   Example:
     [#1] image.jpg warhol 3,3,3,40

 weave:
     _density>=0,0<=_thickness<=100,0<=_shadow<=100,_shading>=0,_fibers_amplitude>=0,_fibers_smoothness>=0,_angle,-1<=_x_curvature<=1,-1<=_y_curvature<=1

   Apply weave effect to the selected images.
   'angle' can be { 0=0 deg. | 1=22.5 deg. | 2=45 deg. | 3=67.5 deg. }.

   Default values: 'density=6', 'thickness=65', 'shadow=40', 'shading=0.5', 'fibers_amplitude=0', _'fibers_smoothness=0', 'angle=0' and
    'curvature_x=curvature_y=0'

   Example:
     [#1] image.jpg weave ,

 whirls:
     _texture>=0,_smoothness>=0,_darkness>=0,_lightness>=0

   Add random whirl texture to selected images.

   Default values: 'texture=3', 'smoothness=6', 'darkness=0.5' and 'lightness=1.8'.

   Example:
     [#1] image.jpg whirls ,

 12.17. Warpings
        --------

 deform:
     _amplitude>=0,_interpolation

   Apply random smooth deformation on selected images.
   'interpolation' can be { 0=none | 1=linear | 2=bicubic }.

   Default value: 'amplitude=10'.

   Example:
     [#1] image.jpg +deform[0] 10 +deform[0] 20

 euclidean2polar:
     _center_x[%],_center_y[%],_stretch_factor>0,_boundary_conditions={ 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }

   Apply euclidean to polar transform on selected images.

   Default values: 'center_x=center_y=50%', 'stretch_factor=1' and 'boundary_conditions=1'.

   Example:
     [#1] image.jpg +euclidean2polar ,

 equirectangular2nadirzenith:

   Transform selected equirectangular images to nadir/zenith rectilinear projections.

 fisheye:
     _center_x,_center_y,0<=_radius<=100,_amplitude>=0

   Apply fish-eye deformation on selected images.

   Default values: 'x=y=50', 'radius=50' and 'amplitude=1.2'.

   Example:
     [#1] image.jpg +fisheye ,

 flower:
     _amplitude,_frequency,_offset_r[%],_angle,_center_x[%],_center_y[%],_boundary_conditions={ 0=dirichlet | 1=neumann | 2=periodic | 3=mirror}

     toto
   Apply flower deformation on selected images.

   Default values: 'amplitude=30', 'frequency=6', 'offset_r=0', 'angle=0', 'center_x=center_y=50%' and 'boundary_conditions=3'.

   Example:
     [#1] image.jpg +flower ,

 kaleidoscope:
     _center_x[%],_center_y[%],_radius,_angle,_boundary_conditions={ 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }

   Create kaleidoscope effect from selected images.

   Default values: 'center_x=center_y=50%', 'radius=100', 'angle=30' and 'boundary_conditions=3'.

   Example:
     [#1] image.jpg kaleidoscope ,

 map_sphere:
     _width>0,_height>0,_radius,_dilation>0,_fading>=0,_fading_power>=0

   Map selected images on a sphere.

   Default values: 'width=height=512', 'radius=100', 'dilation=0.5', 'fading=0' and 'fading_power=0.5'.

   Example:
     [#1] image.jpg map_sphere ,

 nadirzenith2equirectangular:

   Transform selected nadir/zenith rectilinear projections to equirectangular images.

 polar2euclidean:
     _center_x[%],_center_y[%],_stretch_factor>0,_boundary_conditions={ 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }

   Apply euclidean to polar transform on selected images.

   Default values: 'center_x=center_y=50%', 'stretch_factor=1' and 'boundary_conditions=1'.

   Example:
     [#1] image.jpg +euclidean2polar ,

 raindrops:
     _amplitude,_density>=0,_wavelength>=0,_merging_steps>=0

   Apply raindrops deformation on selected images.

   Default values: 'amplitude=80','density=0.1', 'wavelength=1' and 'merging_steps=0'.

   Example:
     [#1] image.jpg +raindrops ,

 ripple:
     _amplitude,_bandwidth,_shape={ 0=bloc | 1=triangle | 2=sine | 3=sine+ | 4=random },_angle,_offset

   Apply ripple deformation on selected images.

   Default values: 'amplitude=10', 'bandwidth=10', 'shape=2', 'angle=0' and 'offset=0'.

   Example:
     [#1] image.jpg +ripple ,

 rotoidoscope:
     _center_x[%],_center_y[%],_tiles>0,_smoothness[%]>=0,_boundary_conditions={ 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }

   Create rotational kaleidoscope effect from selected images.

   Default values: 'center_x=center_y=50%', 'tiles=10', 'smoothness=1' and 'boundary_conditions=3'.

   Example:
     [#1] image.jpg +rotoidoscope ,

 spherize:
     _radius[%]>=0,_strength,_smoothness[%]>=0,_center_x[%],_center_y[%],_ratio_x/y>0,_angle,_interpolation

   Apply spherize effect on selected images.

   Default values: 'radius=50%', 'strength=1', 'smoothness=0', 'center_x=center_y=50%', 'ratio_x/y=1', 'angle=0' and 'interpolation=1'.

   Example:
     [#1] image.jpg grid 5%,5%,0,0,0.6,255 spherize ,

 symmetrize:
     _x[%],_y[%],_angle,_boundary_conditions={ 0=dirichlet | 1=neumann | 2=periodic | 3=mirror },_is_antisymmetry={ 0 | 1 },_swap_sides={ 0 | 1 }

   Symmetrize selected images regarding specified axis.

   Default values: 'x=y=50%', 'angle=90', 'boundary_conditions=3', 'is_antisymmetry=0' and 'swap_sides=0'.

   Example:
     [#1] image.jpg +symmetrize 50%,50%,45 +symmetrize[-1] 50%,50%,-45

 transform_polar:
     "expr_radius",_"expr_angle",_center_x[%],_center_y[%],_boundary_conditions={ 0=dirichlet | 1=neumann }

   Apply user-defined transform on polar representation of selected images.

   Default values: 'expr_radius=R-r', 'expr_rangle=a', 'center_x=center_y=50%' and 'boundary_conditions=1'.

   Example:
     [#1] image.jpg +transform_polar[0] R*(r/R)^2,a +transform_polar[0] r,2*a

 twirl:
     _amplitude,_center_x[%],_center_y[%],_boundary_conditions={ 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }

   Apply twirl deformation on selected images.

   Default values: 'amplitude=1', 'center_x=center_y=50%' and 'boundary_conditions=3'.

   Example:
     [#1] image.jpg twirl 0.6

 warp_perspective:
     _x-angle,_y-angle,_zoom>0,_x-center,_y-center,_boundary_conditions={ 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }

   Warp selected images with perspective deformation.

   Default values: 'x-angle=1.5', 'y-angle=0', 'zoom=1', 'x-center=y-center=50' and 'boundary_conditions=2'.

   Example:
     [#1] image.jpg warp_perspective ,

 water:
     _amplitude,_smoothness>=0,_angle

   Apply water deformation on selected images.

   Default values: 'amplitude=30', 'smoothness=1.5' and 'angle=45'.

   Example:
     [#1] image.jpg water ,

 wave:
     _amplitude>=0,_frequency>=0,_center_x,_center_y

   Apply wave deformation on selected images.

   Default values: 'amplitude=4', 'frequency=0.4' and 'center_x=center_y=50'.

   Example:
     [#1] image.jpg wave ,

 wind:
     _amplitude>=0,_angle,0<=_attenuation<=1,_threshold

   Apply wind effect on selected images.

   Default values: 'amplitude=20', 'angle=0', 'attenuation=0.7' and 'threshold=20'.

   Example:
     [#1] image.jpg +wind ,

 zoom:
     _factor,_cx,_cy,_cz,_boundary_conditions={ 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }

   Apply zoom factor to selected images.

   Default values: 'factor=1', 'cx=cy=cz=0.5' and 'boundary_conditions=0'.

   Example:
     [#1] image.jpg +zoom[0] 0.6 +zoom[0] 1.5

 12.18. Degradations
        ------------

 cracks:
     0<=_density<=100,_is_relief={ 0 | 1 },_opacity,_color1,...

   Draw random cracks on selected images with specified color.

   Default values: 'density=25', 'is_relief=0', 'opacity=1' and 'color1=0'.

   Example:
     [#1] image.jpg +cracks ,

 light_patch:
     _density>0,_darkness>=0,_lightness>=0

   Add light patches to selected images.

   Default values: 'density=10', 'darkness=0.9' and 'lightness=1.7'.

   Example:
     [#1] image.jpg +light_patch 20,0.9,4

 noise_hurl:
     _amplitude>=0

   Add hurl noise to selected images.

   Default value: 'amplitude=10'.

   Example:
     [#1] image.jpg +noise_hurl ,

 pixelize:
     _scale_x>0,_scale_y>0,_scale_z>0

   Pixelize selected images with specified scales.

   Default values: 'scale_x=20' and 'scale_y=scale_z=scale_x'.

   Example:
     [#1] image.jpg +pixelize ,

 scanlines:
     _amplitude,_bandwidth,_shape={ 0=bloc | 1=triangle | 2=sine | 3=sine+ | 4=random },_angle,_offset

   Apply ripple deformation on selected images.

   Default values: 'amplitude=60', 'bandwidth=2', 'shape=0', 'angle=0' and 'offset=0'.

   Example:
     [#1] image.jpg +scanlines ,

 shade_stripes:
     _frequency>=0,_direction={ 0=horizontal | 1=vertical },_darkness>=0,_lightness>=0

   Add shade stripes to selected images.

   Default values: 'frequency=5', 'direction=1', 'darkness=0.8' and 'lightness=2'.

   Example:
     [#1] image.jpg +shade_stripes 30

 shadow_patch:
     _opacity>=0

   Add shadow patches to selected images.

   Default value: 'opacity=0.7'.

   Example:
     [#1] image.jpg +shadow_patch 0.4

 spread:
     _dx>=0,_dy>=0,_dz>=0

   Spread pixel values of selected images randomly along x,y and z.

   Default values: 'dx=3', 'dy=dx' and 'dz=0'.

   Example:
     [#1] image.jpg +spread 3

 stripes_y:
     _frequency>=0

   Add vertical stripes to selected images.

   Default value: 'frequency=10'.

   Example:
     [#1] image.jpg +stripes_y ,

 texturize_canvas:
     _amplitude>=0,_fibrousness>=0,_emboss_level>=0

   Add paint canvas texture to selected images.

   Default values: 'amplitude=20', 'fibrousness=3' and 'emboss_level=0.6'.

   Example:
     [#1] image.jpg +texturize_canvas ,

 texturize_paper:

   Add paper texture to selected images.

   Example:
     [#1] image.jpg +texturize_paper

 vignette:
     _strength>=0,0<=_radius_min<=100,0<=_radius_max<=100

   Add vignette effect to selected images.

   Default values: 'strength=100', 'radius_min=70' and 'radius_max=90'.

   Example:
     [#1] image.jpg vignette ,

 watermark_visible:
     _text,0<_opacity<1,_size>0,_angle,_mode={ 0=remove | 1=add },_smoothness>=0

   Add or remove a visible watermark on selected images (value range must be [0,255]).

   Default values: 'text=(c) G'MIC', 'opacity=0.3', 'size=53', 'angle=25', 'mode=1' and 'smoothness=0'.

   Example:
     [#1] image.jpg watermark_visible ,0.7

 12.19. Blending and Fading
        -------------------

 blend:
     [layer],blending_mode,_opacity[%],_selection_is={ 0=base-layers | 1=top-layers } |
     blending_mode,_opacity[%]

   Blend selected G,GA,RGB or RGBA images by specified layer or blend all selected images together,
   using specified blending mode.
   'blending_mode' can be { add | alpha | and | average | blue | burn | darken | difference |
   divide | dodge | edges | exclusion | freeze | grainextract | grainmerge | green | hardlight |
   hardmix | hue | interpolation | lchlightness | lighten | lightness | linearburn | linearlight | luminance |
   multiply | negation | or | overlay | pinlight | red | reflect | saturation | seamless | seamless_mixed |
   screen | shapeareamax | shapeareamax0 | shapeareamin | shapeareamin0 | shapeaverage | shapeaverage0 |
   shapemedian | shapemedian0 | shapemin | shapemin0 | shapemax | shapemax0 | softburn | softdodge |
   softlight | stamp | subtract | value | vividlight | xor }.
   'opacity' should be in '[0,1]', or '[0,100]' if expressed with a '%'.

   Default values: 'blending_mode=alpha', 'opacity=1' and 'selection_is=0'.

   Example:
     [#1] image.jpg +drop_shadow , resize2dy[-1] 200 rotate[-1] 20 +blend alpha display_rgba[-2]
     [#2] image.jpg testimage2d {w},{h} blend overlay
     [#3] command "ex : $""=arg repeat $""# +blend[0,1] ${arg{$>+1}} text_outline[-1] Mode:
      darken
     [#4] command "ex : $""=arg repeat $""# +blend[0,1] ${arg{$>+1}} text_outline[-1] Mode:
      freeze,grainextract,grainmerge
     [#5] command "ex : $""=arg repeat $""# +blend[0,1] ${arg{$>+1}} text_outline[-1] Mode:
      interpolation,lighten,lightness
     [#6] command "ex : $""=arg repeat $""# +blend[0,1] ${arg{$>+1}} text_outline[-1] Mode:
      multiply,negation,or,overlay
     [#7] command "ex : $""=arg repeat $""# +blend[0,1] ${arg{$>+1}} text_outline[-1] Mode:
      screen,shapeaverage,softburn
     [#8] command "ex : $""=arg repeat $""# +blend[0,1] ${arg{$>+1}} text_outline[-1] Mode:
      value,vividlight,xor

 blend_edges:
     smoothness[%]>=0

   Blend selected images togethers using 'edges' mode.

   Example:
     [#1] image.jpg testimage2d {w},{h} +blend_edges 0.8

 blend_fade:
     [fading_shape]

   Blend selected images together using specified fading shape.

   Example:
     [#1] image.jpg testimage2d {w},{h} 100%,100%,1,1,'cos(y/10)' normalize[-1] 0,1 +blend_fade[0,1] [2]

 blend_median:

   Blend selected images together using 'median' mode.

   Example:
     [#1] image.jpg testimage2d {w},{h} +mirror[0] y +blend_median

 blend_seamless:
     _is_mixed_mode={ 0 | 1 },_inner_fading[%]>=0,_outer_fading[%]>=0

   Blend selected images using a seamless blending mode (Poisson-based).

   Default values: 'is_mixed=0', 'inner_fading=0' and 'outer_fading=100%'.

 fade_diamond:
     0<=_start<=100,0<=_end<=100

   Create diamond fading from selected images.

   Default values: 'start=80' and 'end=90'.

   Example:
     [#1] image.jpg testimage2d {w},{h} +fade_diamond 80,85

 fade_linear:
     _angle,0<=_start<=100,0<=_end<=100

   Create linear fading from selected images.

   Default values: 'angle=45', 'start=30' and 'end=70'.

   Example:
     [#1] image.jpg testimage2d {w},{h} +fade_linear 45,48,52

 fade_radial:
     0<=_start<=100,0<=_end<=100

   Create radial fading from selected images.

   Default values: 'start=30' and 'end=70'.

   Example:
     [#1] image.jpg testimage2d {w},{h} +fade_radial 30,70

 fade_x:
     0<=_start<=100,0<=_end<=100

   Create horizontal fading from selected images.

   Default values: 'start=30' and 'end=70'.

   Example:
     [#1] image.jpg testimage2d {w},{h} +fade_x 30,70

 fade_y:
     0<=_start<=100,0<=_end<=100

   Create vertical fading from selected images.

   Default values: 'start=30' and 'end=70'.

   Example:
     [#1] image.jpg testimage2d {w},{h} +fade_y 30,70

 fade_z:
     0<=_start<=100,0<=_end<=100

   Create transversal fading from selected images.

   Default values: 'start=30' and 'end=70'.

 sub_alpha:
     [base_image],_opacity_gain>=1

   Compute the minimal alpha-channel difference (opposite of alpha blending) between the selected images
   and the specified base image.
   The alpha difference A-B is defined as the image having minimal opacity, such that alpha_blend(B,A-B) = A.

   Default value: 'opacity_gain=1'.

   Example:
     [#1] image.jpg testimage2d {w},{h} +sub_alpha[0] [1] display_rgba

 12.20. Image Sequences and Videos
        --------------------------

 animate:
     filter_name,"param1_start,...,paramN_start","param1_end,...,paramN_end",nb_frames>=0,_output_frames={ 0 | 1 },_output_filename |
     delay>0,_back and forth={ 0 | 1 }

   Animate filter from starting parameters to ending parameters or animate selected images
   in a display window.

   Default value: 'delay=30'.

   Example:
     [#1] image.jpg animate flower,"0,3","20,8",9

 apply_camera:
     _"command",_camera_index>=0,_skip_frames>=0,_output_filename

   Apply specified command on live camera stream, and display it on display window [0].
   This command requires features from the OpenCV library (not enabled in G'MIC by default).

   Default values: 'command=""', 'camera_index=0' (default camera), 'skip_frames=0' and 'output_filename=""'.

 apply_files:
     "filename_pattern",_"command",_first_frame>=0,_last_frame={ >=0 | -1=last },_frame_step>=1,_output_filename

   Apply a G'MIC command on specified input image files, in a streamed way.
   If a display window is opened, rendered frames are displayed in it during processing.
   The output filename may have extension '.avi' or '.mp4' (saved as a video), or any other usual image file
   extension (saved as a sequence of images).

   Default values: 'command=(undefined)', 'first_frame=0', 'last_frame=-1', 'frame_step=1' and 'output_filename=(undefined)'.

 apply_video:
     video_filename,_"command",_first_frame>=0,_last_frame={ >=0 | -1=last },_frame_step>=1,_output_filename

   Apply a G'MIC command on all frames of the specified input video file, in a streamed way.
   If a display window is opened, rendered frames are displayed in it during processing.
   The output filename may have extension '.avi' or '.mp4' (saved as a video), or any other usual image
   file extension (saved as a sequence of images).
   This command requires features from the OpenCV library (not enabled in G'MIC by default).

   Default values: 'first_frame=0', 'last_frame=-1', 'frame_step=1' and 'output_filename=(undefined)'.

 average_files:
     "filename_pattern",_first_frame>=0,_last_frame={ >=0 | -1=last },_frame_step>=1,_output_filename

   Average specified input image files, in a streamed way.
   If a display window is opened, rendered frames are displayed in it during processing.
   The output filename may have extension '.avi' or '.mp4' (saved as a video), or any other usual image
   file extension (saved as a sequence of images).

   Default values: 'first_frame=0', 'last_frame=-1', 'frame_step=1' and 'output_filename=(undefined)'.

 average_video:
     video_filename,_first_frame>=0,_last_frame={ >=0 | -1=last },_frame_step>=1,_output_filename

   Average frames of specified input video file, in a streamed way.
   If a display window is opened, rendered frames are displayed in it during processing.
   The output filename may have extension '.avi' or '.mp4' (saved as a video), or any other usual image
   file extension (saved as a sequence of images).
   This command requires features from the OpenCV library (not enabled in G'MIC by default).

   Default values: 'first_frame=0', 'last_frame=-1', 'frame_step=1' and 'output_filename=(undefined)'.

 fade_files:
     "filename_pattern",_nb_inner_frames>0,_first_frame>=0,_last_frame={ >=0 | -1=last },_frame_step>=1,_output_filename

   Generate a temporal fading from specified input image files, in a streamed way.
   If a display window is opened, rendered frames are displayed in it during processing.
   The output filename may have extension 'avi' or 'mp4' (saved as a video), or any other usual image
   file extension (saved as a sequence of images).

   Default values: 'nb_inner_frames=10', 'first_frame=0', 'last_frame=-1', 'frame_step=1' and 'output_filename=(undefined)'.

 fade_video:
     video_filename,_nb_inner_frames>0,_first_frame>=0,_last_frame={ >=0 | -1=last },_frame_step>=1,_output_filename

   Create a temporal fading sequence from specified input video file, in a streamed way.
   If a display window is opened, rendered frames are displayed in it during processing.
   This command requires features from the OpenCV library (not enabled in G'MIC by default).

   Default values: 'nb_inner_frames=10', 'first_frame=0', 'last_frame=-1', 'frame_step=1' and 'output_filename=(undefined)'.

 files2video:
     "filename_pattern",_output_filename,_fps>0,_codec

   Convert several files into a single video file.

   Default values: 'output_filename=output.mp4', 'fps=25' and 'codec=mp4v'.

 median_files:
     "filename_pattern",_first_frame>=0,_last_frame={ >=0 | -1=last },_frame_step>=1,_frame_rows[%]>=1,_is_fast_approximation={ 0 | 1 }

   Compute the median frame of specified input image files, in a streamed way.
   If a display window is opened, rendered frame is displayed in it during processing.

   Default values: 'first_frame=0', 'last_frame=-1', 'frame_step=1', 'frame_rows=20%' and 'is_fast_approximation=0'.

 median_video:
     video_filename,_first_frame>=0,_last_frame={ >=0 | -1=last },_frame_step>=1,_frame_rows[%]>=1,_is_fast_approximation={ 0 | 1 }

   Compute the median of all frames of an input video file, in a streamed way.
   If a display window is opened, rendered frame is displayed in it during processing.
   This command requires features from the OpenCV library (not enabled in G'MIC by default).

   Default values: 'first_frame=0', 'last_frame=-1', 'frame_step=1', 'frame_rows=100%' and 'is_fast_approximation=1'.

 morph:
     nb_inner_frames>=1,_smoothness>=0,_precision>=0

   Create morphing sequence between selected images.

   Default values: 'smoothness=0.1' and 'precision=4'.

   Example:
     [#1] image.jpg +rotate 20,1,1,50%,50% morph 9

 morph_files:
     "filename_pattern",_nb_inner_frames>0,_smoothness>=0,_precision>=0,_first_frame>=0,_last_frame={ >=0 | -1=last },_frame_step>=1,_output_filename

   Generate a temporal morphing from specified input image files, in a streamed way.
   If a display window is opened, rendered frames are displayed in it during processing.
   The output filename may have extension '.avi' or '.mp4' (saved as a video), or any other usual image
   file extension (saved as a sequence of images).

   Default values: 'nb_inner_frames=10', 'smoothness=0.1', 'precision=4', 'first_frame=0', 'last_frame=-1', 'frame_step=1' and
    'output_filename=(undefined)'.

 morph_rbf:
     nb_inner_frames>=1,xs0[%],ys0[%],xt0[%],yt0[%],...,xsN[%],ysN[%],xtN[%],ytN[%]

   Create morphing sequence between selected images, using RBF-based interpolation.
   Each argument (xsk,ysk)-(xtk,ytk) corresponds to the coordinates of a keypoint
   respectively on the source and target images. The set of all keypoints define the overall image deformation.

 morph_video:
     video_filename,_nb_inner_frames>0,_smoothness>=0,_precision>=0,_first_frame>=0,_last_frame={ >=0 | -1=last },_frame_step>=1,_output_filename

   Generate a temporal morphing from specified input video file, in a streamed way.
   If a display window is opened, rendered frames are displayed in it during processing.
   The output filename may have extension '.avi' or '.mp4' (saved as a video), or any other usual image
   file extension (saved as a sequence of images).
   This command requires features from the OpenCV library (not enabled in G'MIC by default).

   Default values: 'nb_inner_frames=10', 'smoothness=0.1', 'precision=4', 'first_frame=0', 'last_frame=-1', 'frame_step=1' and
    'output_filename=(undefined)'.

 register_nonrigid:
     [destination],_smoothness>=0,_precision>0,_nb_scale>=0

   Register selected source images with specified destination image, using non-rigid warp.

   Default values: 'smoothness=0.2', 'precision=6' and 'nb_scale=0(auto)'.

   Example:
     [#1] image.jpg +rotate 20,1,1,50%,50% +register_nonrigid[0] [1]

 register_rigid:
     [destination],_smoothness>=0,_boundary_conditions={ 0=dirichlet | 1=neumann | 2=periodic | 3=mirror }

   Register selected source images with specified destination image, using rigid warp (shift).

   Default values: 'smoothness=0.1%' and 'boundary_conditions=0'.

   Example:
     [#1] image.jpg +shift 30,20 +register_rigid[0] [1]

 transition:
     [transition_shape],nb_added_frames>=0,100>=shading>=0,_single_frame_only={ -1=disabled | >=0 }

   Generate a transition sequence between selected images.

   Default values: 'shading=0' and 'single_frame_only=-1'.

   Example:
     [#1] image.jpg +mirror c 100%,100% plasma[-1] 1,1,6 transition[0,1] [2],5

 transition3d:
     _nb_frames>=2,_nb_xtiles>0,_nb_ytiles>0,_axis_x,_axis_y,_axis_z,_is_antialias={ 0 | 1 }

   Create 3D transition sequence between selected consecutive images.
   'axis_x', 'axis_y' and 'axis_z' can be set as mathematical expressions, depending on 'x' and 'y'.

   Default values: 'nb_frames=10', 'nb_xtiles=nb_ytiles=3', 'axis_x=1', 'axis_y=1', 'axis_z=0' and 'is_antialias=1'.

   Example:
     [#1] image.jpg +blur 5 transition3d 9 display_rgba

 video2files:
     input_filename,_output_filename,_first_frame>=0,_last_frame={ >=0 | -1=last },_frame_step>=1

   Split specified input video file into image files, one for each frame.
   First and last frames as well as step between frames can be specified.

   Default values: 'output_filename=frame.png', 'first_frame=0', 'last_frame=-1' and 'frame_step=1'.

 12.21. Convenience Functions
        ---------------------

 alert:
     _title,_message,_label_button1,_label_button2,...

   Display an alert box and wait for user's choice.
   If a single image is in the selection, it is used as an icon for the alert box.

   Default values: 'title=[G'MIC Alert]' and 'message=This is an alert box.'.

 arg:
     n>=1,_arg1,...,_argN

   Return the n-th argument of the specified argument list.

 arg0:
     n>=0,_arg0,...,_argN

   Return the n-th argument of the specified argument list (where 'n' starts from '0').

 arg2img:
     argument_1,...,argument_N

   Split specified list of arguments and return each as a new image (as a null-terminated string).

 arg2var:
     variable_name,argument_1,...,argument_N

   For each i in [1...N], set 'variable_name$i=argument_i'.
   The variable name should be global to make this command useful (i.e. starts by an underscore).

 autocrop_coords:
     value1,value2,... | auto

   Return coordinates (x0,y0,z0,x1,y1,z1) of the autocrop that could be performed on the latest
   of the selected images.

   Default value: 'auto'

 average_colors:

   Return the average vector-value of the latest of the selected images.

 base642img:
     "base64_string"

   Decode given base64-encoded string as a newly inserted image at the end of the list.
   The argument string must have been generated using command 'img2base64'.

 base642uchar:
     "base64_string"

   Decode given base64-encoded string as a newly inserted 1-column image at the end of the list.
   The argument string must have been generated using command 'uchar2base64'.

 basename:
     file_path,_variable_name_for_folder

   Return the basename of a file path, and opt. its folder location.
   When specified 'variable_name_for_folder' must starts by an underscore
   (global variable accessible from calling function).

 bin:
     binary_int1,...

   Print specified binary integers into their octal, decimal, hexadecimal and string representations.

 bin2dec:
     binary_int1,...

   Convert specified binary integers into their decimal representations.

 covariance_colors:
     _avg_outvarname

   Return the covariance matrix of the vector-valued colors in the latest of the selected images
   (for arbitrary number of channels).
   Parameter 'avg_outvarname' is used as a variable name that takes the value of the average vector-value.

 dec:
     decimal_int1,...

   Print specified decimal integers into their binary, octal, hexadecimal and string representations.

 dec2str:
     decimal_int1,...

   Convert specifial decimal integers into its string representation.

 dec2bin:
     decimal_int1,...

   Convert specified decimal integers into their binary representations.

 dec2hex:
     decimal_int1,...

   Convert specified decimal integers into their hexadecimal representations.

 dec2oct:
     decimal_int1,...

   Convert specified decimal integers into their octal representations.

 fact:
     value

   Return the factorial of the specified value.

 fibonacci:
     N>=0

   Return the Nth number of the Fibonacci sequence.

   Example:
     [#1] echo ${"fibonacci 10"}

     [gmic]-0./ Start G'MIC interpreter.
     [gmic]-0./ 55
     [gmic]-0./ End G'MIC interpreter.

 file_mv:
     filename_src,filename_dest

   Rename or move a file from a location $1 to another location $2.

 file_rand:

   Return a random filename for storing temporary data.

 filename:
     filename,_number1,_number2,...,_numberN

   Return a filename numbered with specified indices.

 files (+):
     _mode,path

   Return the list of files and/or subfolders from specified path.
   'path' can be eventually a matching pattern.
   'mode' can be { 0=files only | 1=folders only | 2=files + folders }.
   Add '3' to 'mode' to return full paths instead of filenames only.

   Default value: 'mode=5'.

 fitratio_wh:
     min_width,min_height,ratio_wh

   Return a 2D size 'width,height' which is bigger than 'min_width,min_height' and has the specified w/h ratio.

 fitscreen:
     width,height,_depth,_minimal_size[%],_maximal_size[%] |
     [image],_minimal_size[%],_maximal_size[%]

   Return the 'ideal' size WxH for a window intended to display an image of specified size on screen.

   Default values: 'depth=1', 'minimal_size=128' and 'maximal_size=85%'.

 fontchart:

   Insert G'MIC font chart at the end of the image list.

   Example:
     [#1] fontchart

 fps:

   Return the number of time this function is called per second, or -1 if this info is not yet available.
   Useful to display the framerate when displaying animations.

 gcd:
     a,b

   Return the GCD (greatest common divisor) between a and b.

 hex:
     hexadecimal_int1,...

   Print specified hexadecimal integers into their binary, octal, decimal and string representations.

 hex2dec:
     hexadecimal_int1,...

   Convert specified hexadecimal integers into their decimal representations.

 hex2img:
     "hexadecimal_string"

   Insert new image 1xN at the end of the list with values specified by the given hexadecimal-encoded string.

 hex2str:
     hexadecimal_string

   Convert specified hexadecimal string into a string.
   See also: str2hex.

 img2base64:
     _encoding={ 0=base64 | 1=base64url },_store_names={ 0 | 1 }

   Encode selected images as a base64-encoded string.
   The images can be then decoded using command 'base642img'.

   Default values: 'encoding=0'.

 img2hex:

   Return representation of last image as an hexadecimal-encoded string.
   Input image must have values that are integers in [0,255].

 img2str:

   Return the content of the latest of the selected images as a special G'MIC input string.

 img2text:
     _line_separator

   Return text contained in a multi-line image.

   Default value: 'line_separator= '.

 img82hex:

   Convert selected 8bits-valued vectors into their hexadecimal representations (ascii-encoded).

 hex2img8:

   Convert selected hexadecimal representations (ascii-encoded) into 8bits-valued vectors.

 is_3d:

   Return 1 if all of the selected images are 3D objects, 0 otherwise.

 is_change:
     _value={ 0=false | 1=true }

   Set or unset the 'is_change' flag associated to the image list.
   This flag tells the interpreter whether or not the image list should be displayed when the pipeline ends.

   Default value: 'value=1'.

 is_half:

   Return 1 if the type of image pixels is limited to half-float.

 is_ext:
     filename,_extension

   Return 1 if specified filename has a given extensioin.

 is_image_arg:
     string

   Return 1 if specified string looks like '[ind]'.

 is_pattern:
     string

   Return 1 if specified string looks like a drawing pattern '0x......'.

 is_percent:
     string

   Return 1 if specified string ends with a '%', 0 otherwise.

 is_variable_name:
     "str"

   Returns 1 if specified argument can be considered as a variable name, 0 otherwise.

 is_videofilename:

   Return 1 if extension of specified filename is typical from video files.

 is_macos:

   Return 1 if current computer OS is Darwin (MacOS), 0 otherwise.

 is_windows:

   Return 1 if current computer OS is Windows, 0 otherwise.

 math_lib:

   Return string that defines a set of several useful macros for the embedded math evaluator.

 mad:

   Return the MAD (Maximum Absolute Deviation) of the last selected image.
   The MAD is defined as MAD = med_i|x_i-med_j(x_j)|

 max_w:

   Return the maximal width between selected images.

 max_h:

   Return the maximal height between selected images.

 max_d:

   Return the maximal depth between selected images.

 max_s:

   Return the maximal spectrum between selected images.

 max_wh:

   Return the maximal wxh size of selected images.

 max_whd:

   Return the maximal wxhxd size of selected images.

 max_whds:

   Return the maximal wxhxdxs size of selected images.

 median_color:

   Return the median color value of the last selected image.

 min_w:

   Return the minimal width between selected images.

 min_h:

   Return the minimal height between selected images.

 min_d:

   Return the minimal depth between selected images.

 min_s:

   Return the minimal s size of selected images.

 min_wh:

   Return the minimal wxh size of selected images.

 min_whd:

   Return the minimal wxhxd size of selected images.

 min_whds:

   Return the minimal wxhxdxs size of selected images.

 nmd (+):
     Shortcut for command 'named'.

 named (+):
     _mode,"name1","name2",...

   Return the set of indices corresponding to images of the selection with specified names.
   After this command returns, the status contains a list of indices (unsigned integers),
   separated by commas (or an empty string if no images with those names have been found).
   (equivalent to shortcut command 'nmd').

   'mode' can be { 0=all indices (default) | 1=lowest index | 2=highest index | 3 = all indices (case insensitive) | 4 = lowest index (case insensitive) | 5 = highest index (case
   insensitive)}

 normalize_filename:
     filename

   Return a "normalized" version of the specified filename, without spaces and capital letters.

 oct:
     octal_int1,...

   Print specified octal integers into their binary, decimal, hexadecimal and string representations.

 oct2dec:
     octal_int1,...

   Convert specified octal integers into their decimal representations.

 padint:
     number,_size>0

   Return a integer with 'size' digits (eventually left-padded with '0').

 path_cache:

   Return a path to store G'MIC data files for one user (whose value is OS-dependent).

 path_current:

   Return current folder from where G'MIC has been run.

 path_gimp:

   Return a path to store GIMP configuration files for one user (whose value is OS-dependent).

 path_tmp:

   Return a path to store temporary files (whose value is OS-dependent).

 remove_copymark:
     "image_name"

   Remove copy mark from names of selected images.

 reset:

   Reset global parameters of the interpreter environment.

 rgb:

   Return a random int-valued RGB color.

 rgba:

   Return a random int-valued RGBA color.

 shell_cols:

   Return the estimated number of columns of the current shell.

 size_value:

   Return the size (in bytes) of an image value.

 std_noise:

   Return the estimated noise standard deviation of the last selected image.

 str:
     string

   Print specified string into its binary, octal, decimal and hexadecimal representations.

 str2hex:
     "string"

   Convert specified string argument into a sequence of hexadecimal values (returned as a string).
   See also: hex2str.

   Example:
     [#1] hex=${"str2hex

     [gmic]-0./ Start G'MIC interpreter.
     [gmic]-0./ 48656c6c6f206d7920667269656e6473
     [gmic]-0./ End G'MIC interpreter.

 strcapitalize:
     string

   Capitalize specified string.

 strcontains:
     string1,string2

   Return 1 if the first string contains the second one.

 strlen:
     string1

   Return the length of specified string argument.

 strreplace:
     string,search,replace

   Search and replace substrings in an input string.

 strlowercase:
     string

   Return a lower-case version of the specified string.

 struppercase:
     string

   Return an upper-case version of the specified string.

 strvar:
     "string"

   Return a simplified version of the specified string, that can be used as a variable name.
   (version that creates a lowercase result, no longer than 128 chars).

 strcasevar:
     "string"

   Return a simplified version of the specified string, that can be used as a variable name.
   (version that keeps original case of specified string, no longer than 128 chars).

 strver:
     _version

   Return the specified version number of the G'MIC interpreter, as a string.

   Default value: 'version=$_version'.

 tic:

   Initialize tic-toc timer.
   Use it in conjunction with 'toc'.

 toc:

   Display elapsed time of the tic-toc timer since the last call to 'tic'.
   This command returns the elapsed time in the status value.
   Use it in conjunction with 'tic'.

 to_clutname:
     "string"

   Return simplified name that can be used as a CLUT name, from specified input string.

 uchar2base64:
     _encoding={ 0=base64 | 1=base64url }

   Encode the values of the latest of the selected images as a base64-encoded string.
   The string can be decoded using command 'base642uchar'.
   Selected images must have values that are integers in [0,255].

   Default values: 'encoding=0'.

 12.22. Other Interactive Commands
        --------------------------

 demos:
     _run_in_parallel={ 0=no | 1=yes | 2=auto }

   Show a menu to select and view all G'MIC interactive demos.

 tixy:
     "expression"

   Animate specified mathematical expression with a 16x16 grid of circles, using the rules described at https://tixy.land.

 x_2048:

   Launch the 2048 game.

 x_blobs:

   Launch the blobs editor.
   ../img/x_blobs.jpg [image: 'x_blobs']

 x_bouncing:

   Launch the bouncing balls demo.

 x_color_curves:
     _colorspace={ rgb | cmy | cmyk | hsi | hsl | hsv | lab | lch | ycbcr | last }

   Apply color curves on selected RGB[A] images, using an interactive window.
   Set 'colorspace' to 'last' to apply last defined color curves without opening interactive windows.

   Default value: 'colorspace=rgb'.

 x_colorize:
     _is_lineart={ 0 | 1 },_max_resolution={ 0 | >=128 },_multichannels_output={ 0 | 1 },_[palette1],_[palette2],_[grabber1]

   Colorized selected B&W images, using an interactive window.
   When >0, argument 'max_resolution' defines the maximal image resolution used in the interactive window.

   Default values: 'is_lineart=1', 'max_resolution=1024' and 'multichannels_output=0'.

 x_connect4:

   Launch the Connect Four game.

 xz:
     Shortcut for command 'x_crop'.

 x_crop:

   Crop selected images interactively.
   (equivalent to shortcut command 'xz').

 x_cut:

   Cut selected images interactively.

 x_fire:

   Launch the fire effect demo.

 x_fireworks:

   Launch the fireworks demo.

 x_fisheye:

   Launch the fish-eye effect demo.

 x_fourier:

   Launch the fourier filtering demo.

 x_grab_color:
     _variable_name

   Open a color grabber widget from the first selected image.
   Argument 'variable_name' specifies the variable that contains the selected color values at any time.
   Assigning '-1' to it forces the interactive window to close.

   Default values: 'variable_name=xgc_variable'.

 x_hanoi:

   Launch the Tower of Hanoi game.

 x_histogram:

   Launch the histogram demo.

 x_hough:

   Launch the hough transform demo.

 x_jawbreaker:
     0<_width<20,0<_height<20,0<_balls<=8

   Launch the Jawbreaker game.

 x_landscape:

   Launch the virtual landscape demo.

 x_life:

   Launch the game of life.

 x_light:

   Launch the light effect demo.

 x_mandelbrot:
     _julia={ 0 | 1 },_c0r,_c0i

   Launch Mandelbrot/Julia explorer.

 x_mask_color:
     _colorspace={ all | rgb | lrgb | ycbcr | lab | lch | hsv | hsi | hsl | cmy | cmyk | yiq },_spatial_tolerance>=0,_color_tolerance>=0

   Interactively select a color, and add an alpha channel containing the corresponding color mask.
   Argument 'colorspace' refers to the color metric used to compute color similarities, and can be basically
   one of { rgb | lrgb | ycbcr | lab | lch | hsv | hsi | hsl | cmy | cmyk | yiq }.
   You can also select one one particular channel of this colorspace, by setting 'colorspace' as
   'colorspace_channel' (e.g. 'hsv_h' for the hue).

   Default values: 'colorspace=all', 'spatial_tolerance=5' and 'color_tolerance=5'.

 x_metaballs3d:

   Launch the 3D metaballs demo.

 x_minesweeper:
     8<=_width=<20,8<=_height<=20

   Launch the Minesweeper game.

 x_minimal_path:

   Launch the minimal path demo.

 x_morph:
     _nb_frames>=2,_preview_fidelity={ 0=coarsest | 1=coarse | 2=normal | 3=fine | 4=finest }

   Launch the interactive image morpher.

   Default values: 'nb_frames=16' and 'preview_fidelity=3'.

 x_pacman:

   Launch pacman game.

 x_paint:

   Launch the interactive painter.

 x_plasma:

   Launch the plasma effect demo.

 x_quantize_rgb:
     _nbcolors>=2

   Launch the RGB color quantization demo.

 x_reflection3d:

   Launch the 3D reflection demo.

 x_rubber3d:

   Launch the 3D rubber object demo.

 x_segment:
     _max_resolution={ 0 | >=128 }

   Segment foreground from background in selected opaque RGB images, interactively.
   Return RGBA images with binary alpha-channels.

   Default value: 'max_resolution=1024'.

 x_select_color:
     _variable_name

   Display a RGB or RGBA color selector.
   Argument 'variable_name' specifies the variable that contains the selected color values (as R,G,B,[A])
   at any time.
   Its value specifies the initial selected color. Assigning '-1' to it forces the interactive window to close.

   Default value: 'variable_name=xsc_variable'.

 x_select_function1d:
     _variable_name,_background_curve_R,_background_curve_G,_background_curve_B

   Open an interactive window, where the user can defined its own 1D function.
   If an image is selected, it is used to display additional information :
      -  The first row defines the values of a background curve displayed on the window (e.g. an histogram).
      -  The 2nd, 3rd and 4th rows define the R,G,B color components displayed beside the X and Y axes.
   Argument 'variable_name' specifies the variable that contains the selected function keypoints at any time.
   Assigning '-1' to it forces the interactive window to close.

   Default values: 'variable_name=xsf_variable', 'background_curve_R=220', 'background_curve_G=background_curve_B=background_curve_T'.

 x_select_palette:
     _variable_name,_number_of_columns={ 0=auto | >0 }

   Open a RGB or RGBA color selector widget from a palette.
   The palette is given as a selected image.
   Argument 'variable_name' specifies the variable that contains the selected color values (as R,G,B,[A])
   at any time.
   Assigning '-1' to it forces the interactive window to close.

   Default values: 'variable_name=xsp_variable' and 'number_of_columns=2'.

 x_shadebobs:

   Launch the shade bobs demo.

 x_spline:

   Launch spline curve editor.

 x_starfield3d:

   Launch the 3D starfield demo.

 x_tetris:

   Launch tetris game.

 x_threshold:

   Threshold selected images interactively.

 x_tictactoe:

   Launch tic-tac-toe game.

 x_warp:
     _nb_keypoints_xgrid>=2,_nb_keypoints_ygrid>=2,_nb_keypoints_contours>=0,_preview_fidelity={ 0=coarsest | 1=coarse | 2=normal | 3=fine | 4=finest },_[background_image],0<=_background_opacity<=1

   Launch the interactive image warper.

   Default values: 'nb_keypoints_xgrid=nb_keypoints_ygrid=2', 'nb_keypoints_contours=0' and 'preview_fidelity=1'.

 x_waves:

   Launch the image waves demo.

 x_whirl:
     _opacity>=0

   Launch the fractal whirls demo.

   Default values: 'opacity=0.2'.

 13. Examples of Use
     ---------------

 'gmic' is a generic image processing tool which can be used in a wide variety of situations. The few examples below illustrate possible uses of this tool:  

 ### View a list of images:

   $ gmic file1.bmp file2.jpeg

 ### Convert an image file:

   $ gmic input.bmp output output.jpg

 ### Create a volumetric image from a movie sequence:

   $ gmic input.mpg append z output output.hdr

 ### Compute image gradient norm:

   $ gmic input.bmp gradient_norm

 ### Denoise a color image:

   $ gmic image.jpg denoise 30,10 output denoised.jpg

 ### Compose two images using overlay layer blending:

   $ gmic image1.jpg image2.jpg blend overlay output blended.jpg

 ### Evaluate a mathematical expression:

   $ gmic echo "cos(pi/4)^2+sin(pi/4)^2={cos(pi/4)^2+sin(pi/4)^2}"

 ### Plot a 2D function:

   $ gmic 1000,1,1,2 fill "X=3*(x-500)/500;X^2*sin(3*X^2)+if(c==0,u(0,-1),cos(X*10))" plot  

                                                   ../img/example_plot.png [image: '']

 ### Plot a 3D elevated function in random colors:

   $ gmic 128,128,1,3,"u(0,255)" plasma 10,3 blur 4 sharpen 10000 n 0,255 elevation3d[-1] "'X=(x-64)/6;Y=(y-64)/6;100*exp(-(X^2+Y^2)/30)*abs(cos(X)*sin(Y))'"

                                                   ../img/example_elevation3d.png [image: '']

 ### Plot the isosurface of a 3D volume:

   $ gmic mode3d 5 moded3d 5 double3d 0 isosurface3d "'x^2+y^2+abs(z)^abs(4*cos(x*y*z*3))'",3  

                                                   ../img/example_isosurface3d.png [image: '']

 ### Render a G'MIC 3D logo:

   $ gmic 0 text G´MIC,0,0,53,1,1,1,1 expand_xy 10,0 blur 1 normalize 0,100 +plasma 0.4 add blur 1 elevation3d -0.1 moded3d 4  

                                                   d../img/example_logo.png [image: '']

 ### Generate a 3D ring of torii:

   $ gmic repeat 20 torus3d 15,2 color3d[-1] "{u(60,255)},{u(60,255)},{u(60,255)}" *3d[-1] 0.5,1 if "{$>%2}" rotate3d[-1] 0,1,0,90 fi add3d[-1] 70 add3d rotate3d 0,0,1,18 done moded3d 3 mode3d 5  
  double3d 0  

                                                   ../img/example_torii.png [image: '']

 ### Create a vase from a 3D isosurface:

   $ gmic moded3d 4 isosurface3d "'x^2+2*abs(y/2)*sin(2*y)^2+z^2-3',0" sphere3d 1.5 sub3d[-1] 0,5 plane3d 15,15 rotate3d[-1] 1,0,0,90 center3d[-1] add3d[-1] 0,3.2 color3d[-1] 180,150,255  
  color3d[-2] 128,255,0 color3d[-3] 255,128,0 add3d  

                                                   ../img/example_vase.png [image: '']

 ### Launch a set of interactive demos:

   $ gmic demos  

 ** G'MIC comes with ABSOLUTELY NO WARRANTY; for details visit: https://gmic.eu **