以倾斜摄影技术三维重建恢复的点云为参考点云，以激光雷达点云为配准点云，

Inclined photography reconstruction recovery point cloud as a reference point cloud, laser radar point cloud as a registration point cloud, obvious feature points in the manually selected overlapping region, such as roofing, intersection, etc., after the registration level of error control at 0.2m within vertical errors are less than 0.1M, in line with the requirements of individual tree extraction. Go after the registration of a point cloud shown in Figure 3.

To tilt photography technology three-dimensional reconstruction recovery point cloud as a reference point cloud, lidar point cloud as the registration point cloud, manual selection of overlapping areas of obvious feature points, such as roofs, intersections, etc., after the registration of horizontal error control within 0.2m, vertical error control within 0.1m, in line with the single wood parameter extraction requirements. The normalized point cloud after registration is shown in Figure 3.

Taking the point cloud recovered from the 3D reconstruction of tilt photography technology as the reference point cloud, the LIDAR point cloud as the registration point cloud, and manually selecting the obvious feature points in the overlapping area, such as the roof, intersection, etc., after registration, the horizontal error is controlled within 0.2m, and the vertical error is controlled within 0.1M, which meets the requirements of single wood parameter extraction. The normalized point cloud after registration is shown in Figure 3.<br>