Kernelized Intuitionistic Fuzzy C-Means algorithms fused with firefly algorithm for image segmentation

Data clustering methods have been used extensively for image segmentation in the past decade. In our previous work, we had established that combining the traditional clustering algorithms with a meta-heuristic like Firefly Algorithm improves the stability of the output as well as the speed of convergence. In this paper, we have replaced the Euclidean distance formula with kernels. We have combined Intuitionistic Fuzzy C-Means (IFCM) with Firefly algorithm and replaced Euclidean distance with Gaussian, Hyper tangent and Radial Basis Function Kernels. The intuition behind using Kernels instead of Euclidean or Manhattan distance is explained in this paper. In order to prove that the kernel based counter part of IFCM performs better than its corresponding Euclidean distance based algorithm, we have used standard performance indices such as DB index and Dunn index.