Abstract:

LiDAR point cloud is a 3D point set composed of massive discrete laser dots which exist in both plane and vertical directions. Because of lacking space topological relations among the discrete dots of LiDAR point cloud, it is important to establish an appropriate data structure for LiDAR point cloud as the foundation of LiDAR processing. According to the structural characteristics of LiDAR point cloud data, a two-level data structure with “non-null” regular cube grid and K-D tree is established for the organization and management LiDAR point cloud using red-black tree and K-D tree to build. The structure could reduce the structural redundancy and improve indexing efficiency.

Key words: LiDAR, red-black tree, K-D tree, data structure, data organization, regular cube grid