Homogeneity
(HOM) (also called
the "Inverse Difference Moment")
Dissimilarity and Contrast result in larger numbers for more contrasty windows. If weights decrease away from the diagonal, the result will be larger for windows with little contrast.
Homogeneity weights values by the inverse of the Contrast weight, with weights decreasing exponentially away from the diagonal:
Homogeneity equation
Exercise: Calculate the homogeneity value for the horizontal GLCM and compare it with the Dissimilarity value.
Self test: The weight used in contrast is (i-j)2. The weight in homogeneity is 1/[1+(i-j)2].
- What would happen if the contrast weight were [1+(i-j)2]?
- Would this be a bad thing?
- What degree is this measure?
Exercise: Do-it-yourself "Similarity" texture (also called Inverse Difference):
Homogeneity is the most commonly used measure that increases with less contrast in the window. However, it would be easy to use the above model to construct a first degree "similarity" measure. Write the equation and perform the calculation for "similarity" using the horizontal GLCM