改进的鲸鱼优化算法与图像阈值分割  

Improved whale optimization algorithm and image threshold segmentation

作　　者：郭松林 巴艳坤 Guo Songlin;Ba Yankun(School of Electrical&Control Engineering,Heilongjiang University of Science&Technology,Harbin 150022,China)

机构地区：[1]黑龙江科技大学电气与控制工程学院,哈尔滨150022

出　　处：《黑龙江科技大学学报》2022年第5期672-678,共7页Journal of Heilongjiang University of Science And Technology

摘　　要：为解决鲸鱼优化算法全局搜索能力弱和收敛性能差等问题,提出了融合Tent混沌映射的改进鲸鱼算法,将算法中的线性递减策略替换为非线性过渡策略,选取8个测试函数测试改进算法的性能,以图像分割为例验证其工程实用性。结果表明,改进后的鲸鱼优化算法不仅全局搜索能力有所提高,收敛性能显著改善,图像阈值分割结果与大津算法相比的平均误差仅为0.15%,符合工程领域应用的标准。This paper proposes an improved whale algorithm fusing Tent chaotic shining to solve the problem of the poor global searching ability and the poor convergence performance.The study involves replacing the linear decreasing strategy to a nonlinear transiting strategy;testing the improved algorithm by selecting eight benchmark functions and verifying the engineering practicability by taking the image segmentation as example.The results show that the improved whale optimization algorithm improves both the global search ability and the convergence performance.The average error is only 0.15%by comparision of the image threshold segmentation result and Otsu algorithm,which meets the standard of engineering application.

关 键 词：鲸鱼优化算法 局部最优 Tent混沌 非线性过渡参数 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP391.413[自动化与计算机技术—控制科学与工程]