Brain Tumor Image Segmentation Using Kernel Based Fuzzy C Means Clustering (KFCM) Algorithm

Abstract

The kernel based fuzzy c means clustering is proposed in this article for segmentation of MR brain image. To alleviate the problem of drawback of computation cost of segmentation in the Fuzzy C Means is overcome by this kernel based FCM algorithm. The FCM algorithm provides good accuracy in the absence of noise; but in the presence of noise it doesn’t give good accuracy. In Kernal Based Fuzzy C Means, First, Enhanced Non Local mean Filter is applied on MR brain image for removal of noise and it replace the gray scale of the denoised image by the average, median filter. The Gaussian Radial basis function is used as a kernel function instead of Euclidean distance.