An improved FCM combining mean shift algorithm is proposed to improve the segmentation visual effects and efficiency of traditional FCM. Firstly, image is segmented into many small homogeneous regions by using mean shift pre-segmentation algorithm, and the homogeneous regions, instead of pixels are taken as new nodes. Then, image local entropy is adopted to describe the new nodes spatial and gray feature. Finally, an exponential function which can simulate well the human nonlinear visual response is used to measure the similarity between new nodes and cluster center nodes. Experimental results on both complex background and noises images show that the proposed algorithm has better robustness to the segmentation effects and efficiency.