IMAGE SEGMENTATION USING PARTICLE SWARM OPTIMIZATION BASED INTUITIONISTIC FUZZY CLUSTERING

Abstract

Image segmentation leads to the identification of several patterns from an image. The main aim of segmentation is to remove the noise and fuzziness from the image and identify certain regions of interest by grouping the pixels into several clusters based on their features. This study presents a novel method for segmenting satellite images using Particle Swarm optimization and Intuitionistic fuzzy clustering algorithm by extracting several statistical features from the image. The performance of the proposed methodology is evaluated over images from QuickBird satellite and the regions are clustered based on the spectral contrast. The optimal number of clusters is evaluated using silhouette coefficient.