Multilevel Image Thresholding with Glowworm Swam Optimization Algorithm based on the Minimum Cross Entropy

Abstract

The measure of cross entropy is a widely used index applied to many areas of computer science. In this article, we proposed a new multilevel thresholding algorithm for image segmentation using the glowworm swarm optimization based on the criterion of minimum cross entropy (MCE). In this paper, a new multilevel MCE thresholding (MCET) algorithm using the glowworm swam optimization (GSO) algorithm is proposed. The proposed image thresholding algorithm is called the GSO-based MCET algorithm. The five different methods including the exhaustive search, the honey bee mating optimization (HBMO), the firefly (FF) algorithm, the artificial bee colony algorithm (ABC) and the particle swarm optimization (PSO) are also implemented for performance comparison. The experimental results demonstrate that the proposed GSO-based MCET algorithm can efficiently search for multiple thresholds that are very close to the optimal ones examined by the exhaustive search method. Compared with the other five thresholding methods, the needs of computation time using the FF-based MCET algorithm is the smallest. Furthermore, the segmentation performance of GSO-based MCET algorithm is better than the PSO-based MCET algorithm, while the results of GSO-based MCET algorithm are not significant different to the other three bio-inspired computing algorithms.