Parallel Whale Optimization Algorithm-Kmeans for Image Segmentation using Multiprocessing

Image segmentation is an important step in inter-preting and analyzing images, allowing computers and algorithms to process the content of an image accuratly and more effectively. Image segmentation is a crucial step in image preprocessing, while Kmeans is a widely-used clustering algorithm for im-age segmentation. However, the Whale Optimization Algorithm (WOA) is a nature-inspired metaheuristic optimization tech-nique known for its simplicity and global search capability. In this study, a hybrid approach was proposed, which combined WOA with Kmeans algorithm (WOA-Kmeans) to optimize image segmentation process, in the other hand, a parallel strategy was implemented for enhancing both segmentation quality and execution time. Traditional Kmeans suffers from sensitivity to initial cluster centroids, often leading to suboptimal results. By integrating WOA, the centroids initialization is improved, guiding Kmeans towards better convergence. Additionally, the parallelization of WOA-Kmeans using multiprocessing enables faster computation by distributing the work among multiple processes, making it suitable for high-resolution image processing tasks. This combination of WOA-Kmeans in a multiprocess environment ensures efficient and accurate image segmentation. The results demonstrate that the Parallel WOA-Kmeans outperform the sequentiel WOA-Kmeans in term of several performance measures, such as, accuracy, Root Mean Squar Error (RMSE), Peak signal to noise ratio (PSNR), and structural similarity index (SSIM).