This paper utilizes histogram based FCM clustering algorithm [1] for segmentation of images with 3-D transform. The spatial probability of neighboring pixels is incorporated in the objective function of FCM to increase the robustness against noise. The proposed approach makes use of histogram based Fuzzy C-Means clustering with de-noising & spatial probability for the segmentation of noisy images, which will give better segmentation accuracy. The experimental result at the end shows that this approach achieves consistent segmentation accuracy regardless of noise levels.

Keywords : Image Segmentation, Fuzzy Logic, De-Noising, Clustering and Spatial Probability.