十进制线性四叉树的点云数据格网索引方法研究

Research on point cloud data management based on spatial index and database

三维激光点云数据是海量三维点的集合,导致数据量庞大,组织和管理困难,不仅增加了系统负荷,而且大大降低了点云数据后续处理效率。该文针对海量点云数据的组织与管理中遇到的加载和显示效率低、建立索引困难、不能实时动态显示等问题,提出了基于十进制线性四叉树的点云数据格网索引方法,该方法用四叉树结构分割点云数据和用SQL Server数据库存储,采用Morton码或矩形区域对点云数据进行分块空间索引,结合空间索引和数据库的优势对点云数据进行高效、动态、智能管理。实验结果表明,该方法较好地解决海量点云数据的组织与管理效率低下,不能实时动态显示的问题。

Laser point set has a massive quantity of data,which not only increases system load,but also greatly reduces the follow-up processing efficiency as well.Considering the difficulty of mass point cloud data organization and management,grid index of point cloud data based on the decimal linear quadtree was proposed in this paper,point cloud data was segmented and encoded by means of linear quadtree based on decimal system,which would be stored in SQL Server.Then,point cloud could be blocked and indexed in terms of Morton code or rectangular region,combined with the advantages of both spatial index and database to manage point cloud efficiently and safely.The results showed that the proposed method ideally solved the problem of point cloud data organization and management.