Image Thresholding Based on Sailfish Optimizer Algorithm

Image segmentation plays a crucial role in various computer vision applications, and thresholding is a commonly used technique. An image can be effectively segmented into distinct regions by selecting appropriate threshold levels. Evolutionary algorithms have been widely employed to automate the process of determining optimal thresholds. This paper proposes a novel approach utilizing the Sailfish algorithm to select suitable thresholds by defining an objective function tailored for image segmentation tasks. Random threshold values within the image are considered members of the initial population in this algorithm. The objective is to iteratively refine and extract optimal threshold values through the execution of the Sailfish algorithm. Experimental results demonstrate that this proposed method achieves higher accuracy and performance than traditional approaches in detecting object edges and boundaries.