Image Segmentation Based on GA-FCM Clustering and Probability Relaxation

Image segmentation is an essential approach for image processing.For some complex images,the extracted objects usually have a problem of incomplete edge or broken boundary.To solve this problem,we first apply fuzzy C-means clustering method based on genetic algorithm(GA-FCM) to segment the image pixels into different regiments.Besides background pixels and definite object pixels,union operation shows that the pixels around broken area are uncertain to be classified into object or background.We present probability relaxation(PR) algorithm to further segment the uncertain pixels according to their statistic properties.Experimental results indicate that this algorithm well solved the above problem and is effective for image segmentation and object extraction.