Social spider algorithm employed multi-level thresholding segmentation approach

Abstract

Multi-level based thresholding is one of the most imperative techniques to realize image segmentation. In order to determine the threshold values automatically, approaches based on histogram are commonly employed. We have deployed histogram based bi-modal and multi-modal thresholding for gray image using social spider algorithm (SSA). We have employed Kapur’s and Otsu’s functions and in order to maximize its value, we have employed social spider algorithm (SSA). We have used the standard pre-tested images. Results have shown that the social spider algorithm has out-performed the results obtained by Particle Swarm Optimization (PSO) as far as optimal threshold values and computational time are concerned.