A Fast Image Thresholding Method Based on Chaos Optimization and Recursive Algorithm for Two-Dimensional Tsallis Entropy

Abstract

The two-dimensional (2-D) maximum Tsallis entropy method often gets ideal segmentation results, because it not only takes advantage of the spatial neighbor information with using the 2-D histogram of the image, but also has some flexibility with a parameter. However, its time-consuming computation is often an obstacle in real time application systems. In this paper, a fast image thresholding method based on chaos optimization and recursive algorithm for 2-D Tsallis entropy is presented. Firstly, improve the traditional chaos optimization algorithm(COA) so that it can get global solution with lower computation load, then  propose a recursive algorithm with the stored matrix variables and prove the equation among Tsallis entropies of four parts in the 2-D histogram, finally combine the improved COA and the recursive algorithm to reduce much computational cost in the process of solving the 2-D maximum Tsallis entropy problem. Experimental results show the proposed approach can get better segmentation performance and has much higher speed.