An image segmentation method based on improved whale optimization algorithm

This study presents a Whale Optimization Algorithm (LWOA) that utilizes Logistic-Tent chaotic mapping for OTSU threshold segmentation. Initially, the Logistic-Tent chaotic mapping and Simulated Binary Crossover operation are applied to initialize the whale population, enhancing ergodicity and randomness, which boosts global search capabilities. Adaptive weights are then introduced to help the algorithm more effectively avoid local minima. Finally, experiments with various algorithms on standard test images demonstrate that the proposed LWOA method reduces overall running time by 25% to 35% while maintaining segmentation accuracy compared to other threshold segmentation methods.