Histogram based fuzzy C-mean algorithm for image segmentation

Since a real image is usually very complex, there must be some uncertainties and errors in image segmentation. The fuzzy C-mean (FCM) algorithm can overcome this problem, but the cost will be a large amount of computation time. The authors present an improved FCM algorithm which uses a histogram, instead of the gray function, to find centers of the gray level. Theoretical analysis and experiment results show that it can reduce the computing time significantly.<<ETX>>