An Enhanced Bats Algorithm for Optimization Thresholds in Medical Image Segmentation

Abstract

The key to threshold segmentation is choosing the thresholds that will decide the segmentation's outcome— as the number of thresholds grows, so does the calculation complexity, which causes challenges with traditional methods. This study introduces an enhanced Bats algorithm (EBA) for selecting optimal picture segmentation thresholds and applying it to the global optimization issue of the segmentation function's objective function. The EBA blends well with picture segmentation that produces exceptional computation in global convergence and robustness and prevents trapping into local optimization. It is particularly well suited to solving complex functions with high dimensions and multiple peaks. Compared to GA and PSO, extensive theoretical study and simulation results reveal that EBA has greater efficacy, efficiency, stability of the range of thresholds, and quality in multi-image and multi-thresholds segmentation.