HISTHOMOGENEITY - Homogeneity feature derived from (co)occurrence matrices.

Contents

Description
Syntax
References
See also
Function implementation

Description

Calculate the homogeneity feature defined in [HSD73] from (co)occurrence matrices.

This feature measures the local homogeneity H in an image as [ST99]:

GLCM case - with $P_{ij}$ the joint probability of the greylevel pair : $H = \sum_{i,j} \frac{P_{ij}}{(1+(i-j)^2)}$

GLSDV case - with the probability of the greylevel difference : $H = \sum_d \frac{P_d}{(1+d^2)}$

Syntax

   H = HISTHOMOGENEITY(Pij, i, j);
   H = HISTHOMOGENEITY(Pd, d);

References

See HISTCONTRAST.

See also

HISTCONTRAST, HISTENERGY, HISTVARIANCE, HISTENTROPY2, HISTENTROPY10, HISTMAXIMUM, HISTMEAN, HISTDISSIMILARITY, HISTIDIFFERENCE, HISTCORRELATION, LOCALGLCM2D, LOCALGLOV2D, LOCALGLSDV2D.

Function implementation

function H = histhomogeneity(pIJ,I,varargin)

H = I;
if nargin>2
    H = H - varargin{1};
end;

H = (1 + H .* H);
H = sum (pIJ(:) ./ H(:));
end % end of histhomogeneity