Optimizing Image Segmentation: A Multilevel Thresholding Based on Differential Evolution

Thresholding is a popular image segmentation technique for converting grey-level images to binary images. A Multilevel Thresholding Algorithm for image segmentation based on Differential Evolution is to designed an image segmentation module using Multilevel Thresholding (MTH) which separates pixels into discrete zones with respect to the image’s objects and is the most commonly used option for segmenting real-time images. From the image’s histogram, we will determine the optimum threshold value based on Otsu’s method in Multilevel Thresholding (MTH) based on Differential Evolution. After using Otsu’s techniques, we examine the system by comparing Multilevel Thresholding (MTH) performance with different optimization using the Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index (SSIM). The outcome demonstrates that the Differential Evolution algorithm almost matches the performance of the Harmony Search (HS), Artificial Bee Colony (ABC), and Particle Swarm Optimization (PSO) approaches. This is evident from the PSNR and SSIM values, which reveal little differences across the four techniques. It can be concluded that the Differential Evolution algorithm performs well and achieves nearly the same PSNR and SSIM value as other image segmentation techniques. Due to the numerous multilevel thresholding techniques, the PSNR and SSIM values from the obtained data show substantial variations.