基于改进麻雀搜索算法的二维Otsu多阈值分割  

Two-dimensional Otsu Multi-threshold Segmentation Based on Improved Sparrow Search Algorithm

作　　者：黄聪 HUANG Cong(Yueyang Vocational Technical College,Yueyang,Hunan 414000)

机构地区：[1]岳阳职业技术学院,湖南岳阳414000

出　　处：《岳阳职业技术学院学报》2025年第1期78-82,共5页Journal of Yueyang Vocational and Technical College

基　　金：2019年度湖南省教育厅科学研究一般课题“基于移动终端的微课平台的开发与应用”(16C1643)。

摘　　要：本文针对现有二维Otsu多阈值分割方法存在的分割精度较低、分割速率较慢等问题,提出了一种基于改进麻雀搜索算法的二维Otsu多阈值分割方法。在初始化阶段,引入Logistic混沌映射增强种群的多样性;在局部搜索阶段,分别应用莱维飞行策略、柯西变异策略更新麻雀种群中发现者和加入者的位置,以解决种群陷入局部最优的问题;最后,通过改进麻雀搜索算法求解二维Otsu算法的分割阈值。在BSDS500分割数据集上与5种群体智能优化算法优化的二维Otsu算法进行全面比较,在结构相似性和计算效率2个量化指标上的综合实验结果表明:该方法在分割精度和计算效率方面明显优于相比较的其他5种方法。To solve the problems of low segmentation accuracy and slow segmentation speed in existing two-dimensional Otsu multi-threshold image segmentation methods,a two-dimensional Otsu multi-threshold image segmentation method based on an improved sparrow search algorithm is proposed.Firstly,in the initialization stage,Logistic chaotic mapping is introduced to enhance the diversity of the population.Secondly,in the local search stage,Lévy flight strategy and Cauchy mutation strategy are applied to update the positions of the discoverer and joiner in the sparrow population,respectively,to solve the problem of the population falling into local optima.Finally,the improved sparrow search algorithm is used to solve the segmentation thresholds of the two-dimensional Otsu algorithm.A comprehensive comparison is conducted with 5 population intelligence optimization algorithms for optimizing the two-dimensional Otsu algorithm on the BSDS500 segmentation dataset.Comprehensive experimental results in two quantitative indicators,including structural similarity and computational efficiency,indicate that the proposed method significantly outperforms the 5 methods in terms of both segmentation accuracy and computational efficiency.

关 键 词：图像分割 二维Otsu算法 多阈值 改进麻雀搜索算法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]