建筑物点云数据拼接模型及拼接精度分析

Splicing model and precision analysis of building point cloud data

激光雷达进行扫描时,将扫描目标的形态信息记录下来形成点云数据。在实际扫描目标建筑时,通常受到视角和工作环境的限制,因此会在不同角度对目标建筑进行多次扫描,并将多视角点云信息进行拼接后,得到目标建筑的完整点云信息,点云数据的拼接质量对三维重建模型的精度有直接影响。在使用车载激光雷达扫描时,由于GPS可以实时获取大地坐标信息,因此,利用此信息可以将不同视角的点云数据整合到全局坐标系中,从而完成较高精度的点云拼接工作。

When the laser radar is scanning,the morphological information of the scanned target is recorded to form point cloud data.In the actual scanning of the target building,it is usually limited by the perspective and working environment,so the target building will be scanned multiple times from different angles,and the multi-view point cloud information will be spliced to obtain the complete point cloud information of the target building.The splicing quality of point cloud data has a direct impact on the accuracy of the three-dimensional reconstruction model.When using vehicle-mounted LIDAR scanning,since GPS can obtain geodetic coordinate information in real time,it can use this information to integrate point cloud data from different perspectives into the global coordinate system,so as to complete the high-precision point cloud stitching work.