基于最优邻域局部熵的点云精简算法  

Point cloud simplification algorithm based on local entropy of optimal neighborhood

作　　者：林松 田林亚[1] 毕继鑫 施贵刚 朱依民 闻亚 LIN Song;TIAN Linya;BI Jixin;SHI Guigang;ZHU Yimin;WEN Ya(School of Earth Science and Engineering, Hohai University, Nanjing 211100,China;Zhejiang Huadong Surveying and Security Technology Co., Ltd,Hangzhou 310014,China;College of Civil Engineering,Anhui Jianzhu University, Hefei 230099,China;Key Laboratory of Unmanned Aerial Vehicle Development & Data Application of Anhui Higher Education Institutes,Ma'anshan 243031,China)

机构地区：[1]河海大学地球科学与工程学院,江苏南京211100 [2]浙江华东测绘与工程安全技术有限公司,浙江杭州310014 [3]安徽建筑大学土木工程学院,安徽合肥230099 [4]安徽省教育厅无人机开发及数据应用重点实验室,安徽马鞍山243031

出　　处：《测绘工程》2021年第5期12-17,共6页Engineering of Surveying and Mapping

基　　金：安徽省教育厅高校自然科学研究重大项目(KJ2019ZD53);华东勘测设计研究院有限公司科技项目(KY2016-02-11-W1)。

摘　　要：针对传统的点云精简算法中不能良好保留细节特征的问题,提出一种基于最优邻域局部熵的点云精简算法。首先利用点云局部邻域协方差矩阵的3个特征值构造的维度特征,构建局部邻域信息熵函数,其次依据局部熵值最小原则确定最优邻域,然后根据最优邻域下计算的特征值间的关系,以及局部信息熵来剔除平坦区域数据点。通过模拟数据和实例扫描数据精简实验,结果表明该方法能较好的保留细节特征。To solve the problem that the traditional point cloud simplification algorithm can’t keep the detail features well,a point cloud simplification algorithm based on the optimal neighborhood local entropy is proposed.Firstly,the dimensional features of the three eigenvalues of the local neighborhood covariance matrix of the point cloud are used to construct the information entropy function of the local neighborhood.Secondly,the optimal neighborhood is determined according to the principle of the minimum of the local entropy.Finally,the data points in the flat region are eliminated according to the local information entropy and the relations between the eigenvalues calculated under the optimal neighborhood.Through the simplifying experiment of the simulated data and the example scan data,the result shows that the method can retain the detail features better.

关 键 词：点云精简 主成分分析 最优邻域 信息熵 

分 类 号：P234.4[天文地球—摄影测量与遥感]