Airborne laser radar point cloud data conversion method and system based on attribute matching
By decomposing and dynamically matching airborne LiDAR point cloud data, and utilizing GPU acceleration and KD tree proximity retrieval, the problems of uneven accuracy and large edge difference in airborne LiDAR point cloud data conversion were solved, achieving high-precision point cloud conversion.
Patent Information
- Application Number
- CN202610110583.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-01-27
- Publication Date
- 2026-05-26
AI Technical Summary
In the existing technology, the conversion methods between airborne LiDAR point cloud data obtained by different calculation methods have uneven accuracy, large differences in the edge connection of point clouds between flight strips, and poor processing effect in areas with abrupt changes in elevation, which cannot meet the requirements for accurate conversion.
The airborne LiDAR point cloud data conversion method based on attribute matching decomposes the point cloud data into three types: deduplicated point cloud, duplicate point cloud, and point cloud with missing attributes. It uses GPU-accelerated row-by-row matching and selective KD-tree neighbor retrieval to perform dynamic matching, and finally achieves accurate conversion of all airborne LiDAR point clouds.
It achieves accurate conversion of airborne LiDAR point cloud data, with a conversion success rate of 100%, and lossless conversion of 99.9999% of the complete point cloud data with attributes. The conversion accuracy is higher and meets real-world requirements.
Smart Images

Figure CN122085247A_ABST