摘要
大多数图像分割方法存在分割精度低、计算复杂度高等传统缺陷.将多混沌系统、自适应动态调整策略融入基本差分进化中,结合最大二维熵原理,给出一种基于新型混合智能算法的图像分割方法(IHIAIS).利用初始化阶段多混沌系统优点、采用控制参数少、收敛速度快、收敛精度高的基本差分进化算法,搜索最优阈值,最终实现图像分割.实验结果表明:该算法在图像分割中具有良好的效果和较快的收敛速度,优于大多数图像阈值分割方法.
Most image segmentation methods have the traditional defects of low segmentation accuracy and high computational complexity.In this paper,an image segmentation method based on a new hybrid intelligent algorithm(IHIAIS)is proposed by integrating the multi-chaotic system and the adaptive dynamic adjustment strategy into the basic differential evolution and combining with the principle of maximum two-dimensional entropy.By using the advantages of multiple chaotic system in initialization stage,the basic differential evolution algorithm with few control parameters,fast convergence speed and high convergence precision is adopted to search the optimal threshold,and finally the image segmentation is realized.Experimental results show that this algorithm has good effect and fast convergence rate in image segmentation,which is better than most image threshold segmentation methods.
作者
刘俊梅
马永刚
LIU Junmei;MA Yonggang(School of Mathematics and Statistics,Yulin University,Yulin 719000,Shaanxi China)
出处
《河南科学》
2022年第5期709-713,共5页
Henan Science
基金
陕西省教育厅项目(20JK1016)
榆林高新区科技计划项目(CXY—2021—61,CXY—2021—65)。
关键词
图像分割
最大二维熵
多混沌系统
差分进化算法
image segmentation
maximum two-dimensional entropy
multi-chaotic system
differential evolution algorithm
