渐进加密的点云滤波算法效率分析与优化  

Efficiency Analysis and Optimization of Point Cloud Data Filter Algorithm for Progressive Encryption

作　　者：陈性义[1] 陶思然 毛君亚 黄迟 CHEN Xingyi;TAO Siran;MAO Junya;HUANG Chi(China University of Geosciences(Wuhan),Wuhan 430074,China;Tianjin Surveying and Mapping Institute,Tianjin 300381,China;Xi’an Love Drone Technology Company Limited,Xi’an 710065,China)

机构地区：[1]中国地质大学(武汉)信息工程学院,武汉430074 [2]天津市测绘院,天津300381 [3]西安爱生无人机技术有限公司,西安710065

出　　处：《遥感信息》2018年第5期106-111,共6页Remote Sensing Information

基　　金：国家863计划资助项目(2007AA092102)

摘　　要：滤波是机载LiDAR点云数据处理中极为重要的步骤,也是时间消耗较大的环节。该文分析了基于TIN渐进加密滤波算法各个阶段的效率,针对其中最为耗时的点定位阶段,比较了使用不同点定位方法的滤波效率,提出了一种基于二级格网的逆向点定位方法。在不同层次的迭代滤波过程中,该方法将待判脚点以大小不同的格网进行组织,然后从各个三角形面片出发,判定三角形中包含的待判点。实验表明,该方法能很好地适应于机载LiDAR数据从大量三角形中快速定位大量待判点的问题,滤波效率得到明显优化。对研究基于TIN渐进加密的点云数据滤波效率提升有一定实用意义,也可为研究TIN快速构建算法提供参考。Filtering for point clouds is a crucial and high time cost step in the processing of airborne LiDAR data.This paper analyzed the efficiency in each stages of filtering algorithm based on progressive TIN densification.According to one of the most time consuming stage which is called point locating,this paper compared the filtering efficiency by using different point locating method,and presented a new reverse point locating method based on two levels of grid.Firstly,the new method organizes data in different sizes of grid in different iteration of filtering processes,and then it calculates the points which contained by triangle from the perspective of each triangle.The test results show that the algorithm can be well adapted to the airborne LiDAR point cloud data which need to locate massive points from massive triangles,and the efficiency optimization is obvious.It has the practical significance to the study of the efficiency optimization of filtering algorithm based on progressive TIN densification for airborne LiDAR point cloud data,and it can provide reference for research on rapid TIN construction algorithm,too.

关 键 词：TIN渐进加密 点云 滤波 点定位 两级格网 LIDAR 

分 类 号：P237[天文地球—摄影测量与遥感]