基于高程统计方法的机载LiDAR点云数据滤波  

A Filtering Method of Airborne LiDAR Point Cloud Data Based on Statistics of Elevation Distribution

作　　者：龚亮[1,2] 张永生[1] 施群山[1] 徐国华[1] 

机构地区：[1]信息工程大学测绘学院,河南郑州450052 [2]江西省数字国土重点实验室,江西抚州344000

出　　处：《测绘与空间地理信息》2012年第2期42-45,共4页Geomatics & Spatial Information Technology

基　　金：江西省数字国土重点实验室开放研究基金项目(DLLJ201111)资助

摘　　要：提出了一种基于高程统计的LiDAR数据滤波方法,该方法以高程分布统计量为依据,通过迭代的方法逐个滤除,是一种基于原始点云的非监督滤波算法。采用北京凤凰岭地区某块机载LiDAR数据进行实验,并做出了实验改进。实验表明,该方法能够准确有效地滤除地物点,既能较好地保持数据信息量与精度,又不依赖于其他数据源,是一种简单有效的滤波算法。The authors put forward a LiDAR data filtering method based on the statistics of elevation distribution in this research. The proposed method ,which is unsupervised and bases itself on raw points cloud, filters out object points via iteration on the basis of the statistics of elevation distribution. An experiment is conducted and improved with a group of LiDAR data of Fenghuangling area in Beijing. The results show that the proposed algorithm is simple and effective in filtering out object points, keeping data quantity and precision at the same time without relying on other data sources.

关 键 词：机载激光扫描 滤波 高程分布 第三势差 

分 类 号：P228.4[天文地球—大地测量学与测量工程]