集成虚拟格网和八叉树的点云数据管理方法

为实现大规模点云的快速组织和检索,提出了基于内外存动态调度的虚拟格网与八叉树相结合的点云数据索引和组织策略。该方法利用虚拟格网对点云进行节点分割和存储,并结合磁盘映射与内外存动态调度技术,将节点进行构八叉树以实现点云数据高效的组织与检索。实验证明,该方法可在计算机内存资源有限的情况下,以较小的内存消耗实现十亿级规模点云的快速组织和检索操作。

A Point Cloud Data Management Method Based on the Integration of Virtual Grid and Octree

In order to achieve the fast organization and retrieval of large scale point cloud,the index and organization strategy of point cloud based on the integration of virtual grid and octree is proposed in this paper.First,the point cloud segmentation and storage are conducted by using the virtual grid.Then,combining with the disk mapping and in-core and out-of-core dynamic exchange technology,the efficient organization and search of point cloud are achieved by organizing the node data in the form of the octree.The experiment result shows that this method can achieve the fast organization and retrieval operation of one billion scale point cloud in the case of limited computer memory resources.