# unu-dist - Man Page

Euclidean distance transform

## Synopsis

**unu dist -th**,**--thresh** <*val*> [**-b**,**--bias** <*val*>] [**-t**,**--type** <*type*>] [**-sgn**] [**-inv**] [**-i**,**--input** <*nin*>] [**-o**,**--output** <*nout*>]

## Description

Euclidean distance transform. Based on “Distance Transforms of Sampled Functions” by Pedro F. Felzenszwalb and Daniel P. Huttenlocher, *Cornell Computing and Information Science* TR2004-1963. This function first thresholds at the specified value and then does the distance transform of the resulting binary image. The signed distance (negative values inside object) is also available. Distances between non-isotropic samples are handled correctly.

- Uses
**nrrdDistanceL2**or**nrrdDistanceL2Signed**

## Options

- -th <
*val*> , --thresh <*val*> threshold value to separate inside from outside (

*double*)- -b <
*val*> , --bias <*val*> if non-zero, bias the distance transform by this amount times the difference in value from the threshold (

*double*); default: “**0.0**”- -t <
*type*> , --type <*type*> type to save output in; default: “

**float**”- -sgn
also compute signed (negative) distances inside objects, instead of leaving them as zero

- -inv
values

**below**threshold are considered interior to object. By default (not using this option), values above threshold are considered interior.- -i <
*nin*> , --input <*nin*> input

*nrrd*- -o <
*nout*> , --output <*nout*> output

*nrrd*(*string*); default: “**-**”