Abstract: To improve the ability of fuzzy Cmeans clustering algorithm (FCM) for complex texture structure images， a new fuzzy Cmeans clustering algorithm (EnFCM) was proposed by combining image structure features. Firstly， the input image was meanfiltered， and the filtered image was added to the original image to form the new image for the subsequent operations. Secondly， the 2D Gabor filtering function was adopted to extract texture structure feature for the new images to replace the gray level similarity measurement in the traditional FCM algorithm. Finally， a new distance measure function was proposed to calculate the distance between the nodes and the clusters. The simulation results showed that more precise segmentation results could be obtained from complicated texture structure images using the presented algorithm.

Key words: image segmentation, FCM clustering, mean filter, texture feature, 2D Gabor filtering