Objectives In order to solve the problem of uneven distribution of laser point cloud data and low efficiency of point cloud data neighbor query, this paper proposes a neighbor query method based on laser point cloud index structure.

Methods Based on the Z curve, the method first transfers the information of laser point cloud data to the KD_ZoC index structure composed of local KD tree and local octree. When the laser point cloud data is divided, the minimum cube is formed according to the Z address of the node, and the space is further divided into subspaces. Second, in order to achieve fast data retrieval, leaf nodes are grouped by local KD tree and local octree, and the number of data points is further reduced according to the proposed octant domain. Finally, the data is stored in the linked list, and the Z address of the data is combined to achieve efficient nearest neighbor query.

Results This paper makes the following contributions: (1) Aiming at the problem that the single index structure is hard to effectively process laser point cloud data, this paper proposes a new index structure, the KD_ZoC index structure, based on the Z curve. It constructs a hybrid index structure of local octrees and local KD trees, and introduces threshold values for the data division of both components. This effectively reduces laser point cloud data, solving the problem of difficult data processing when the laser point cloud data volume is large. (2)To address the problem of comparing multi-dimensional data, this paper uses the Z curve for dimensionality reduction of point cloud data. It provides a specific conversion formula to transform the data into binary data, which is stored in the leaf nodes of the leaf octree. This effectively reduces the comparison time for nearest neighbor queries and saves storage space. (3)To improve the efficiency of nearest-neighbor queries for point cloud data, this paper proposes an octant circle neighborhood search method based on the KD_ZoC index structure. By utilizing symmetry for effective data reduction, it reduces computational time costs and presents an efficient nearest - neighbor query algorithm.

Conclusion Theoretical research and experimental results show that the proposed method can effectively deal with the nearest neighbor query problem under the laser point cloud, and realize the efficient retrieval and management of data points.