Image segmentation based on modified centroid weight particle swarm optimization and spatial fuzzy C-means clustering algorithm

Abstract

An ordinary FCM algorithm does not completely utilize the spatial information in the image. In this paper, we exhibit a fuzzy c-means (FCM) method that integrates spatial information into the membership function for clustering. The spatial function is a summation of membership function in neighborhood of every pixel under consideration. In image segmentation fuzzy C-means (FCM) clustering method making it effortlessly traps into local optimum and huge calculation, image segmentation algorithm based on the modified centroid weight particle swarm optimization (MPSO) and Spatial FCM clustering algorithm is proposed. This technique is a powerful method for medical image and multimedia image segmentation with spatial information. The simulation outcomes and the comparison among the proposed method, FCM and K-means algorithm indicate that the proposed method can achieve better segmentation and excel the existing FCM algorithm in several performances, such as the average error and cluster centroid.