Sine cosine algorithm for underwater multilevel thresholding image segmentation

As an important and effective method, the multilevel thresholding not only attractes the attention of many scholars in recent years, but also is widely used to solve the image segmentation problem. However, the computational complexity becomes large when the threshold levels increase. To overcome this shortcoming, the sine cosine algorithm (SCA) based on Kapur's entropy method is proposed to solve the underwater multilevel thresholding image segmentation problem, which can balance exploration and exploitation to obtain the global optimal solution. To verify the effectiveness and feasibility of the SCA, the segmentation results are compared with other algorithms including bat algorithm (BA), flower pollination algorithm (FPA), particle swarm optimization (PSO), whale optimization algorithm (WOA) by maximizing fitness value of Kapur's entropy method. The fitness value, execution time, peak signal to noise ratio (PSNR), structure similarity index (SSIM) and Wilcoxon's rank-sum test are important evaluation indicators. The experimental results indicate that the SCA has a shorter execution time, higher segmentation accuracy and stronger robustness.