Hybrid strategy improved dung beetle optimization algorithm based 2D kapur entropy image segmentation method

As one of the key technologies in image processing, multi-threshold image segmentation has been widely applied in various image analysis tasks. However, how to improve computational efficiency while ensuring segmentation accuracy remains a current research challenge. To address this, this paper proposes a hybrid strategy-based Improved Dung Beetle Optimization (IDBO) algorithm and applies it to Kapur multi-threshold image segmentation, then hybrid strategy Improved Dung Beetle Optimization algorithm based 2D Kapur entropy image Segmentation (IDBOKS) method is obtained. The proposed algorithm first initializes the population with the SPM-DE mechanism, which integrates the SPM sequence and differential evolution (DE) strategy to ensure a uniform distribution of initial solutions in the search space and enhance population diversity. Second, a rolling dung beetle group information sharing strategy is introduced, a dynamic scaling factor is employed for simulating dung beetle cooperative behavior to improve information exchange among individuals and boost search efficiency. Finally, a global quadratic interpolation mechanism is adopted to optimize the position update process to further enhance the algorithm's ability to escape local optima and accelerate convergence speed. To validate the effectiveness of the proposed algorithm, FSIM, SSIM, and PSNR are selected as evaluation metrics, and comparative experiments are conducted against several recent swarm intelligence-based image segmentation algorithms. The experimental results show that the proposed IDBOKS demonstrates superior segmentation performance and stronger robustness when dealing with complex images.