Abstract: In view of medical image segmentation greatly influenced by noise and low segmentation accuracy, a image segmentation algorithm based on genetic clustering is put forward. Firstly, according to sliding window we divide image into several sub-domains and using genetic algorithm searches the optimal population in each sub-domain. The following thing is the most optimal population is treated as clustering initial value of fuzzy c-means clustering algorithm for image segmentation. Finally, we obtain a number of better result of image segmented. Experimental results show that the algorithm has significant improvement to image segmentation relative to Otsu algorithm.