A new gray image segmentation algorithm using cat swarm optimization

摘要

This paper proposes a new method for image segmentation using Cat Swarm Optimization (CSO) by multilevel thresholding. In the algorithm, each cat contains a set of threshold values that helps to get a given image partitioned into regions. The fitness value, position and velocity information of each cat is used to update the threshold values of each cat. The evaluation function deals with the segmented images as a set of regions, a cluster of associated pixels having gray levels within a given range by using the gray level of image pixel elements. The method for proposal has been compared with the results obtained using Particle Swarm Optimization (PSO) Algorithm on various benchmark images and the outcome shows its potency.