A Hybrid Adaptive Quantum Behaved Particle Swarm Optimization Algorithm Based Multilevel Thresholding for Image Segmentation

Thresholding is a frequently used method in image processing because of its consistency and low computational cost. Kapur's method is an important threshold segmentation method. However, it is computationally expensive when extended to multilevel thresholding since it exhaustively searched the optimal thresholds to optimize the objective functions. Recently, metaheuristic algorithms have been successfully applied for thresholding problems. A multi-threshold segmentation of 2D Kapur's entropy based on hybrid adaptive quantum behaved particle swarm optimization (HAQPSO) algorithm is proposed. Then, the Gaussian chaotic map model and the Levy flight are employed to increase the search ability of HAQPSO algorithm and balance the exploitation and exploration. The HAQPSO algorithm optimizes the Kapur's multi-threshold method to conduct experiments on standard images, satellite images and sport images. The experimental results show that HAQPSO is an effective image segmentation method, with high segmentation accuracy, good convergence, strong anti-noise and certain engineering practicability.