一种联合实虚靶点的点云配准方法、设备及存储介质
By combining physical target spheres and virtual target points in cross-river ultra-long-distance surface deformation monitoring, the problems of sparsity and error amplification in point cloud registration are solved, achieving high-precision point cloud registration. This method is suitable for surface deformation monitoring in high and steep mountain reservoirs and along rivers.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SANXIA JINSHAJIANG YUNCHUAN HYDROPOWER DEV CO LTD
- Filing Date
- 2026-04-14
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies for cross-river ultra-long-distance surface deformation monitoring suffer from problems such as sparse target spheres, large registration errors, and a lack of natural landform features in point cloud registration, resulting in insufficient monitoring accuracy and reliability.
A point cloud registration method combining physical target spheres and virtual target points is adopted. By deploying physical target spheres near the shore and constructing virtual target points in the monitoring area, a long control baseline is formed. Point cloud registration is then performed by combining singular value decomposition and rotation matrix calculation.
It breaks through the distance limitation, significantly improves the accuracy and reliability of point cloud registration, and is suitable for monitoring large-span natural scenes such as reservoirs in high and steep mountainous areas and riverbanks, reducing registration errors and improving robustness.
Smart Images
Figure CN122049008B_ABST