关键词: 图像分割，模糊G均值，模糊隶属度

Abstract: Abstract I}he conventional FCM algorithm is sensitive to noise. I}o overcome the defect of FCM algorithm, we proposed a novel regularized fuzzy o means algorithm for image segmentation based on the spatial restrained fuzzy membership. This approach introduced the relation of space restrained fuzzy membership into the modified FCM objective function. In the new objective function, the membership generated by the proposed algorithm is a product of two terms, the first term is the standard FCM membership responsible for data partitioning, and the second term is a robust constraint of neighborhood membership. Due to the introduction of the spatial information and the relation of neighborhood membership during the process of clustering, the algorithm has good performance in resisting noises.

Key words: Image segmentation,FCM,Fuzzy membership