Texture Attributes Extraction Library

Description

This toolbox contains functions to compute texture descriptors based on histogram, gray level gradient, co-occurence matrix and run length matrix.

Extraction Functions

  • histstat - Computes histogram statistics from an image
  • glcmdesc- Computes descriptors from the co-occurence matrix
  • iarl - Run Length Matrix
  • rldesc- Computes descriptors from the run length matrix
  • gradstat- Computes statistics from a gradient histogram
  • graycomatrix - computes the glcm matrix given an image

Pipeline Functions

  • textpipe - computes a set of texture descriptors given an image, roi and a quantization rate
  • textComp - Compiles the attributes by computing the mean of different orientation attributes

Auxiliary Functions

  • preproc3sigma - pre-processing [mu-3sigma,mu+3sigma]
  • quantimg - quantize an image given and bits/pixel rate
  • readata - reads .txt files containing feats and labels

Pre-processing Functions

  • norm3sigma - normalization [mu-3sigma,mu+3sigma]
  • normmean - intensity normalization by fixing MEAN value over a database
  • normmax- intensity normalization by fixing MAX value over a database
  • normhisteq - intensity normalization by applying histogram equalization
  • normhiststret - intensity normalization by applying histogram stretching
  • normgausker - intensity normalization by applying gaussian kernel

Extras

  • textpipe2 - computes a set of texture descriptors given an image, roi and a quantization rate using only 1 distance
Recently modified pages
More