多目标轮换策略引导的信息熵阈值分割算法  

Information entropy thresholding algorithm guided by the multi-objective swapping strategy

作　　者：雷博 王宁宁 李金明 LEI Bo;WANG Ningning;LI Jinming(School of Communications and Information Engineering,Xi’an University of Posts and Telecommunications,Xi’an 710121,China;Xi’an Key Laboratory of Image Processing Technology and Applications for Public Security,Xi’an 710121,China)

机构地区：[1]西安邮电大学通信与信息工程学院,陕西西安710121 [2]西安市公共安全图像处理技术及应用重点实验室,陕西西安710121

出　　处：《西安邮电大学学报》2023年第5期27-39,共13页Journal of Xi’an University of Posts and Telecommunications

基　　金：陕西省自然科学基础研究计划项目(2022JM-370,2017JM6107)。

摘　　要：为了提高信息熵阈值分割算法图像分割的准确性,提出了一种多目标轮换策略引导的信息熵阈值分割算法。综合利用图像的Kapur熵、交叉熵、Renyi熵和Masi熵作为目标函数进行轮换,采用混沌粒子群优化算法寻优,得到一个由4个目标函数生成的解集,利用改进的Xie-Beni指数来测度、选择解集中的最优解,最终利用最优解实现图像分割。实验结果表明,多目标轮换算法的分割准确率能够达到66%~99%。与相关算法相比,所提算法对图像分割的准确性较高。In order to improve the accuracy of information entropy thresholding algorithm,an information entropy thresholding algorithm guided by multi-objective swapping strategy is proposed.The Kapur entropy,cross entropy,Renyi entropy and Masi entropy of an image are comprehensively used as the objective function to swap,and the chaotic particle swarm optimization algorithm is adopted to optimize,and a solution set is generated by the four fitness functions.Finally,the optimal solution of the solution set is measured and selected by the improved Xie-Beni index,and the optimal solution is utilized to achieve image segmentation.Experimental results show that the segmentation accuracy of the multi-objective swapping algorithm can reach 66%~99%.Compared with related algorithms,the proposed algorithm is more accurate in image segmentation.

关 键 词：阈值分割 信息熵 多目标轮换策略 混沌粒子群优化 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]