SDEXTRACT  Extract texture features

    DATA=SDEXTRACT(FILENAME,options)
    DATA=SDEXTRACT(IM,options)
    DATA=SDEXTRACT(DATA,options)

 INPUT
    FILENAME Image filename
    IM       Image matrix
    DATA     Data set with image data

 OUTPUT
    DATA  Data set with image data

 OPTIONS
    'block'    Block size (default: 8 pixels)
    'step'     Step size, (default: 1)
    'features' Feature type (default: 'moments')
    'mask',M   foreground mask, where the features get computed.
               (matrix of the same size as input image)

 DESCRIPTION
 SDEXTRACT creates a data set with local image features computed on a
 regular grid over the input image. The output data set DATA is an image
 and may be visualized using SDIMAGE. SDEXTRACT may be used also on
 arbitraryrily-shaped regions (data subsets) of the original image.  In
 this situation, SDEXTRACT computes features only in regions without
 holes.

 FEATURES
 'moments' - mean and variance in a region
 'hist' - region histogram
   'hist_bins' Number of histogram bins (def: 8)
   'data_range' - [min,max] vector with image data range (def: [min,max])
 'histfeat' - 5 features computed on the histogram
   'hist_bins' Number of histogram bins (def: 8)
   'data_range' - [min,max] vector with image data range (def: [min,max])
 'CM' - cooccurrence matrix
   'cm_bins' - number of bins used to subsample original data (def:8)
             - data set contains cm_bins^2 output features
   'cm_displ' - displacement in vertical direction (def:1);
   'data_range' - [min,max] vector with image data range (def: [min,max])

 EXAMPLES
  data=sdextract(im,'block',8,'step',2,'features','hist','hist_bins',16)
  data=sdextract(im,'block',8,'step',1,'features','cm','data_range',[0 255])

 READ MORE
 http://perclass.com/doc/guide/feature_extraction.html

 SEE ALSO
 SDIMAGE