A multi-level thresholding image segmentation method using hybrid Arithmetic Optimization and Harris Hawks Optimizer algorithms
Li Qiao,Kai Liu,Yanfeng Xue,Weidong Tang,Taybeh Salehnia
DOI: https://doi.org/10.1016/j.eswa.2023.122316
IF: 8.5
2024-05-01
Expert Systems with Applications
Abstract:Today, image segmentation methods are widely used for various applications, including object detection. Multilevel Thresholding Image Segmentation (MTIS) methods are among the efficient methods for image segmentation. In MTIS methods, it is very important to find the thresholds that gives the best performance for the MTIS and better separate and detect the objects on the image from the image background. Meta-Heuristic (MH) algorithms are among the strategies that can achieve good results in obtaining optimal thresholds to solve this problem. In this paper, we use the combination of Arithmetic Optimization Algorithm (AOA) and Harris Hawks Optimizer (HHO) to improve AOA in exploitation phase, and achieve an optimal threshold vector for MTIS. Therefore, our new hybrid AOA-HHO algorithm solves the MTIS problem with better quality than both AOA and HHO algorithms and some other MH algorithms, and can obtain better thresholds that increase the performance of the MTIS system than AOA and HHO. AOA is powerful in the exploration, and HHO in exploitation phase is powerful. Therefore, AOA-HHO uses the features of both algorithms to search the entire search space locally and globally to find the best find the solution, the high power of the AOA exploration phase, and the high power of the HHO exploitation phase. Also, we use a mathematical equation as the fitness function, that is obtained by using image features. A series of experiments were performed using seven different threshold levels on the test images. Experiments show that AOA-HHO method is better than the compared algorithms and even HHO and AOA in terms of image segmentation accuracy, fitness function value, Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), and execution time.
computer science, artificial intelligence,engineering, electrical & electronic,operations research & management science