改进的渐进加密三角网机载点云滤波方法  

An improved progressive TIN densification filtering method for airborne point clouds

作　　者：王效盖 王健[1] 刘翔宇 曹一 WANG Xiaogai;WANG Jian;LIU Xiangyu;CAO Yi(College of Geomatics,Shandong University of Science and Technology,Qingdao Shandong 266590,China)

机构地区：[1]山东科技大学测绘与空间信息学院,山东青岛266590

出　　处：《北京测绘》2023年第2期178-184,共7页Beijing Surveying and Mapping

基　　金：高端外国专家引进计划(G2021025006L)。

摘　　要：针对渐进加密三角网滤波算法在林区机载点云滤波中存在种子点选取困难和精度较低的问题,提出了一种适合林区点云数据的改进渐进加密三角网滤波方法。该方法首先使用去噪算法(SOR)对离群点进行剔除,然后采用布料模拟和局部薄板样条插值方法获取大量均匀可靠的地面种子点,最后利用改进的渐进加密三角网滤波方法进行滤波,迭代运算进而得到地面点。使用6组标准数据和3组林区数据进行实验,标准数据的平均总误差和Kappa系数分别为2.16%和84.96,林区数据的平均总误差为4.62%。实验结果表明,改进方法适用于复杂的林区机载点云滤波,且提高了滤波精度。Aiming at the problems of difficult seed point selection and low accuracy of the progressive encrypted triangular mesh filtering algorithm in the airborne point cloud filtering in forest areas,a progressive encrypted triangular mesh filtering method suitable for forest area point cloud data was proposed.The improved method was designed to firstly explore the statistical outlier removal(SOR)algorithm to reject outlier points,then obtain a large number of uniform and reliable ground seed points with the help of fabric simulation and local thin-slab sample interpolation method,and finally apply the improved progressive encrypted triangular mesh filtering method to filter and iterate to obtain ground points.The average total error and Kappa coefficient of the standard data were 2.16% and 84.96%,respectively,and the average total error of the forest data was 4.62%.Comparing with other methods,the results showed that the improved method could be applied to forest area airborne point cloud data filtering and improve the filtering quality.

关 键 词：机载激光雷达(LiDAR) 滤波 渐进加密三角网(TIN) 布料模拟 

分 类 号：P237[天文地球—摄影测量与遥感]