This paper presents a texture segmentation algorithm based on directional Gabor filters when orientation characteristics of textures are taken into account.Incorporating into the human visual characteristics

a design approach of optimal directional Gabor filters is proposed.Each texture can be thought of as containing a narrow range of frequency and orientation components.By filtering an input texture image with Gabor filters tuned to the dominant frequency and orientation component of the textures

it is possible to locate each texture.The magnitude of the channel ouput should be large when the texture exhibits the frequency and orientation characteristics to which the channel's Gabor filter is tuned

vice versa.There are weak intraclass dispersion and strong interclass dispersion in the filtered image

and the issue of texture segmentation is translated into that of traditional image segmentation.Experimental results indicate that the proposed algorithm outperforms conventional approaches in terms of both objective measurements and visual evaluation.