Spatial Multiple Criteria Fuzzy Clustering for Image Segmentation

In this paper, we propose a spatial fuzzy clustering method based on multiple criteria optimization method. Our fuzzy clustering method is an enhanced version of fuzzy c-means (FCM) with consideration of multiple criteria. We initiate our multiple criteria optimization approach with two criteria in term of the features of an image: spatial information and intensity. Our spatial clustering attempts to overcome the limitation of conventional FCM. We have tested the proposed method on synthetic and coins images with added Gaussian and Salt and Pepper noises. Besides, we also experiment on real medical images (Magnetic Resonance Images (MRI)) for brain and osteosarcoma (bone tumour). The result reported was encouraging.