改进鲸鱼算法优化Ostu多阈值图像分割

Improved Whale Algorithm Optimizing Ostu Multi-threshold Image Segmentation

针对传统Ostu多阈值图像分割方法确定最优阈值效率低、分割效果差的不足,提出基于改进鲸鱼算法优化Ostu多阈值的图像分割算法。利用鲸鱼优化算法中种群对食物源的启发式搜索机制确定图像分割的最优阈值有效降低了算法时间复杂度。在传统鲸鱼优化算法中引入伪对立学习和混沌Tent映射提高初始种群多样性,设计非均匀衰减收敛因子加快算法收敛,利用自适应惯性权重位置更新机制平衡全局搜索和局部开发,进而有效提升算法寻优精度。

Aiming at the disadvantages of low efficiency and poor segmentation effect of traditional Ostu multi-threshold image segmentation methods,an image segmentation algorithm based on improved whale algorithm was proposed to optimize Ostu multi-threshold image segmentation.The heuristic search mechanism of population to food source in whale optimization algorithm is used to determine the optimal threshold of image segmentation,which effectively reduces the time complexity of the algorithm.In the traditional whale optimization algorithm,pseudo-opposition learning and chaotic Tent mapping are introduced to improve the initial population diversity,non-uniform attenuation convergence factor is designed to accelerate the algorithm convergence,and the adaptive inertial weight position update mechanism is used to balance the global search and local development,so as to effectively improve the optimization accuracy of the algorithm.