城区LiDAR点云自适应坡度的滤波算法  

Adaptive Slope Filtering Algorithm for LiDAR Point Clouds in Urban Areas

作　　者：何培培[1] 万幼川[2] 黄桂平[1] 马开锋[1] 

机构地区：[1]华北水利水电大学资源与环境学院,郑州450052 [2]武汉大学遥感信息工程学院,武汉430079

出　　处：《遥感信息》2017年第5期62-68,共7页Remote Sensing Information

基　　金：国家科技支撑计划(2014BAL05B07);河南省工程实验室开放基金资助项目(502002-B07)

摘　　要：针对机载LiDAR系统所获取的离散点云中地面点与非地面点有效分离的问题,提出了一种适合城区分类的自适应坡度滤波算法。首先,充分考虑点云数据组织形式以及地表形态,借鉴基于坡度滤波的基本原理,实现坡度值的自适应更新。其次,该算法也是一个迭代更新的过程,在该过程中不仅可以更新坡度值,同时可以更新DTM,能够很好地顾及大尺度数据整体的地形起伏和局部细节。最后,采用3组不同类型的数据对算法进行定性或定量方面的分析,以验证该算法性能的可行性。In order to effectively separate ground point and non-point ground, which are acquired by the airbone LiDAR system, the paper proposes an adaptive filtering algorithm suitable for slope urban classification. Firstly, this algorithm fully takes the point organization and surface morphology into account,and updates the slope value using the basic principle of the gradient filtering. And then, since this filtering method is an iterative process of updating,it not only updates the slope value in this iterative process, but also can update the DTM simultaneously. Besides, this method takes the large-scale data overall topography and local details into account. Finally, experiments with qualitative or quantitative analysis of three different types of point clouds show the feasibility of the filtering algorithm.

关 键 词：机载LIDAR 点云滤波 自适应坡度 DTM 迭代更新 

分 类 号：P231.5[天文地球—摄影测量与遥感]