Improved fuzzy c-means algorithm for image segmentation

In order to preserve more image details and enhance its robustness to noise for image segmentation, an improved fuzzy c-means algorithm (FCM) for image segmentation is presented by incorporating the local spatial information and gray level information in this paper. The modified membership function and clustering center function are more mathematically reasonable than those of the FLICM, so the iterative sequence can converge to a local minimum value of the improved objective function. The new fuzzy factor grants the algorithm a novel balance between robustness to noise and effectiveness of preserving the details. The revised algorithm flow has significantly accelerated the processing procedure. Through these improvements, the experiments on the artificial and real images show that the proposed algorithm is very effective.