一种机载LiDAR和车载LiDAR点云的自动配准方法
首发时间:2017-05-26
摘要:机载激光扫描(ALS)和车载激光扫描(MLS)是目前采集城区三维数据的重要手段。由于两者工作方式的局限性,都不能完整获取目标顶面和各个立面的数据,需要将两种数据源进行融合处理。ALS点云和MLS点云的配准是消除两者的几何偏差,实现融合处理的重要前提。本文提出了一种ALS和MLS点云自动配准的算法,首先自动提取点云中的地面和立面区域;然后根据平坦度判断,从这些区域中选取可靠的候选点,并对候选点集进行方向加权的ICP配准。利用Dundas地区采集的真实ALS和MLS数据进行实验,结果表明本算法可以有效的消除两种数据源之间的几何偏差,验证了方法的可行性和有效性。
For information in English, please click here
Automatic Registration of point clouds from ALS and TLS
Abstract:Airborne Laser Scanning (ALS) and Mobile Laser Scanning (MLS) are two wide-used technique for spatial data collection. As the limitation of their working modes, it is impossible for each one individually to capture complete data covering the object's roof and every fa?ades. Merging ALS and MLS data is a practical way to take advantages of their own merits. Due to the misalignment of these two data, data registration is the primary concern. In this paper, an method for ALS and MLS point cloud registering is presented. Firstly, the terrain area and fa?ade area are detected from ALS and MLS point cloud respectively. Then a reliable set of candidate points is extracted from there area according to their flatness. Finally, a direction-weighting ICP algorithm is employed to determine the geometric transformation between ALS and MLS point cloud. A suite of ALS and MLS datasets which acquired in Dundas is used to test our method. The result shows our method can releive the geometric discrepancy between ALS and MLS point cloud effectively.
Keywords: LiDAR point cloud registration ICP
基金:
论文图表:
引用
No.4735495120158914****
同行评议
共计0人参与
勘误表
一种机载LiDAR和车载LiDAR点云的自动配准方法
评论
全部评论0/1000