Combining GLCM Features and Markov Random Field Model for Colour Textured Image Segmentation

In this paper, we propose a new approach for color textured image segmentation. It is a two stage technique, where in the first stage, textural features using gray level co-occurrence matrix (GLCM) are computed for regions of interest (ROI)considered for each class. ROI act as ground truths for the classes. Ohta model (I1, I2, I3) is the colour model used for segmentation. Mean at inter pixel distance (IPD) 1 of I2 component was found to be the optimized textural feature for further segmentation. In the second stage, the feature matrix obtained is assumed to be the degraded version of the image labels and Markov random field model is employed to model the unknown image labels. The labels are estimated through maximum a posteriori estimation criterion using iterated conditional modes algorithm. The performance of the proposed approach is compared with that of using GLCM and Maximum Likelihood classifier and with the one which uses GLCM and MRF in RGB colour space. The proposed method is found to be better in terms of accuracy than the other two methods.