DEM辅助的机载LiDAR点云滤波改进算法研究  

Research on improved algorithm of DEM-assisted airborne LiDAR point cloud filtering

作　　者：周薇薇 彭泊涵 张辉松 李兵[2,3] 蔺志永 ZHOU Weiwei;PENG Bohan;ZHANG Huisong;LI Bing;LIN Zhiyong(Jiangsu Ocean University,Lianyungang,Jiangsu 100120,China;Beijing Institute of Surveying and Mapping,Beijing 100038,China;Beijing Key Laboratory of Urban Spatial Information Engineering,Beijing 100038,China;Wuhan University,Wuhan 430072,China)

机构地区：[1]江苏海洋大学,江苏连云港222005 [2]北京市测绘设计研究院,北京100038 [3]城市空间信息工程北京市重点实验室,北京100038 [4]武汉大学,武汉430072

出　　处：《测绘科学》2022年第5期91-98,共8页Science of Surveying and Mapping

基　　金：江苏省高等学校自然科学研究面上项目(20KJB420003)

摘　　要：针对多源数据辅助LiDAR点云滤波的问题,提出了以经典渐进加密不规则三角网(TIN)算法为基础,对该算法中种子点选取和地面点判断准则两个过程进行分析,在DEM数据辅助下分别对这两个过程进行改进。同时考虑往期DEM与数据之间的变化情况并进行处理,实现地形自适应的点云滤波。该文选择的两块地形条件不同的实验数据进行实验,结果表明,不同地形条件下,使用本文方法与传统TIN加密滤波算法对比滤波精度明显提高。此外,该文还选用5种不同分辨率的DEM数据进行辅助滤波实验,实验结果表明DEM分辨率变化会导致滤波精度变化,但与传统TIN滤波方法相比,各分辨率DEM数据辅助下的点云滤波结果精度都有提高。Using multi-source data to assist LiDAR point cloud filtering is an important research direction of LiDAR point cloud filtering.In this paper,based on the incrementally encrypted triangulated irregular network(TIN)algorithm,the two processes of seed point selection and ground point judgment criteria in the algorithm are elaborately analyzed and greatly improved with the aid of DEM data by our method.Meanwhile,the changes between DEM and data in the past are considered and processed to realize the terrestrially adaptive point cloud filtering.Finally,this paper selects three pieces of experimental data with different terrain conditions for filtering.The results illustrate that,under different terrain conditions,the filtering accuracy of this method is significantly improved compared with the traditional TIN encryption filtering algorithm.In addition,we also select five kinds of DEM data with different resolutions for auxiliary filtering experiments.Experimental results show that changes in DEM resolution will lead to changes in filtering accuracy,but these accuracies are all better than that obtained using traditional TIN filtering method,further validating the effectiveness of the proposed method.

关 键 词：DEM 点云滤波 机载LIDAR 

分 类 号：TN713[电子电信—电路与系统] P237[天文地球—摄影测量与遥感]