Multilevel thresholding Aerial image segmentation using comprehensive learning-based Snow ablation optimizer with double attractors

Aerial photography is a remote sensing technique used for target detection, enabling both qualitative and quantitative analysis. The segmentation process is considered one of the most important processes to improve the analysis of Aerial images. In this study, we introduce an alternative multilevel threshold image segmentation based on a modified Snow ablation optimizer (SAO) algorithm. This modification is conducted using the strengths of Comprehensive learning and Double attractors which aims to enhance the exploration and exploitation abilities of the SAO during the process of discovering the optimal threshold levels that are used to segment the Aerial photography image. To validate the quality of the modified version of SAO, named DCSAO, a set of experimental series is conducted using the CEC2022 benchmark function and sixteen Aerial images at different threshold levels. In addition, we compared the results of DCSAO with different well-known Metaheuristic techniques. The results show the superior performance of DCSAO in comparison to other algorithms according to the performance metrics.