Modified FCM approach for MR brain image segmentation

摘要

Image segmentation is defined as action of splitting an image into band of pixels. Aim of this process is to group pixels into regions. This is used in locating objects in satellite images, face recognition, iris recognition, agricultural imaging, medical imaging. Clustering is process that make group of similar objects and is widely used technique for brain tumour detection. Fuzzy C means clustering is testified to be good as compared to other technique. Drawback of this method is time required is large for convergence. So this time can be minimized by improving cluster center and membership value updating criteria. Magnetic Resonance brain images various types like Craniopharyngioma, Ependymoma, CNS-lymphoma, Pineal tumour, germinoma, Glioblastoma can be used.