Multi-Level Thresholding for Image Segmentation on Medical Images Using Multi Otsu and Sine Cosine Optimization Algorithm

Abstract

Image thresholding is deciding stage in many image processing algorithms. It helps to segment the image, which results in successful prediction correctly. The Sine Cosine is a meta heuristic optimization algorithm that outperforms many conventional algorithms because of its unique principle. Since lung cancer is the most deadly illness, creating a significant number of deaths worldwide, this research used lung cancer CT images. On lung cancer Computed Tomography (CT) images, this paper illustrates how to use the Otsu multi thresholding objective function and Sine Cosine nature-inspired optimization algorithms. The proposed approach uses the Otsu multi thresholding technique on a CT image as an objective function in the Sine Cosine algorithm (SCA).  It aids in selecting elite solutions by measuring fitness for a given range of candidate solutions. When the Algorithm's output was evaluated using PSNR and SSIM, as well as calculation time, it was observed that the proposed approach performed higher.