Fuzzy entropy based multilevel image thresholding using modified gravitational search algorithm

Abstract

Multilevel image thresholding is a powerful and commonly used technique in image analysis. Conventional image segmentation methods suffer a large amount of computation time and unstable segmentation results. In this paper, we present a multilevel image thresholding method based on fuzzy entropy and modified gravitational search algorithm. Fuzzy entropy based image thresholding is extended to multilevel, and a modified gravitational search algorithm (mGSA) is proposed to accelerate the multilevel fuzzy entropy maximization process. Experimental results show that the proposed method can obtain optimal multilevel thresholds and the proposed mGSA has more accurate and stable segmentation results compared with firefly algorithm (FA) and cuckoo search (CS).