Features for texture segmentation using Gabor filters

This work presents a method of extracting texture features from a Gabor transform data block and the application of these features for texture segmentation by clustering feature vectors. For a given image, 16 Gabor features using Gabor kernels with four scales and four orientations are computed. Filtered images are computed by using a Gabor filter bank on a 32/spl times/32 windowed neighborhood for each pixel of the image. Texture features are obtained by computing the 'energy' in the window for each pixel from the filtered images. A clustering algorithm is used to group the vectors based on their distribution in feature space. By clustering Gabor features, it is possible to segment an image into uniform regions. Experimental results demonstrate that features extracted using the proposed approach have excellent discriminating power.