基于莱维飞行麻雀搜索优化的Otsu多阈值图像分割

Otsu multi threshold image segmentation based on Levy flight sparrow search optimization

群智能算法优化多阈值图像分割策略,易陷入局部最优,搜索精度不高。将莱维飞行扰动策略融入麻雀种群觅食的搜索优化过程,以增加图像分割空间搜索的多样性,从而提高分割精度,避免搜索过程陷入局部最优;同时嵌入Sin混沌搜索机制,改进种群初始化策略,加强搜索性能。最终实现多阈值图像分割的应用性能提升,在保持算法全局寻优能力的基础上大幅度提高收敛速度和求解精度。对经典的图像多阈值分割的实验结果表明,本文方法相比于传统的智能优化策略,在寻优率和分割精度方面提升显著,收敛能力强。

The swarm intelligence optimization multi threshold image segmentation strategy is easy to fall into local optimization and the search accuracy is not high.Levy flight disturbance strategy is integrated into the search optimization process of sparrow population foraging to increase the diversity of image segmentation spatial search,so as to improve the segmentation accuracy and avoid falling into local optimization in the search process.At the same time,the Sin chaotic search mechanism is embedded to improve the population initialization strategy and strengthen the search performance.Finally,the application performance of multi threshold image segmentation is improved,and the convergence speed and solution accuracy are greatly improved on the basis of maintaining the global optimization ability of the algorithm.The experimental results of classical multi threshold image segmentation show that compared with the traditional intelligent optimization strategy,this method significantly improves the optimization rate and segmentation accuracy,and has strong convergence ability.