Performance Evaluation of Texture Segmentation Methods

Various methods have been developed for texture segmentation. Since all of them have their merits and drawbacks, the choice of the most suitable method becomes a nontrivial task. We present a comparative study of texture segmentation methods based on four frequently used texture feature extraction techniques: gray level co-occurrence matrices, Gaussian Markov random fields, Gabor filtering, and Laws’ masks. The methods have been tested on the large set of images synthesized especially for this investigation. The performance of these methods has been evaluated against ground truth image by using well-known performance measures – modified Pratt’s figure of merit and correct pixel classification rate.