机载LiDAR点云单栋建筑物激光脚点检测和分离方法

Laser Footprint Detection and Separation Method of One Single Building Based on the Airborne LiDAR Point Clouds

单栋建筑物的激光脚点检测是基于机载LiDAR点云建筑物三维重建的一个重要步骤。本文提出了一种基于TIN的联通成分分析方法，它利用TIN（不规则三角网）组织建筑物激光脚点数据，建立点之间的空间拓扑关系，然后删除TIN中水平距离和三维距离过大的边，对剩余的TIN进行连通成分分析，以此来分离出单栋建筑物的激光脚点。实验表明，该方法对于简单场景中的单层建筑物和复杂场景中的多层建筑物都能很好地实现自动分离。

Laser footprint detection and separation of one single building is an important step in the reconstruction of 3D building models based on the airborne LiDAR point clouds. A TIN - based connected - component analysis is proposed in this paper. TIN is used to organize the laser footprint data of the building and establish the spatial topological relationship of points. Then the overlong edges both in horizontal distance and 3D space are deleted in the TIN, and connected-component analysis is performed to the remaining TIN. In this way, the laser footprints of one single building can be separated. The experiment results show that the method is very effective in automatic separation for both the one - story building in simple scene and muhistory building in complex scene.