摘要
近年来利用三维激光扫描技术进行的国内各地城市数字化发展迅速,随着硬件技术的升级和扫描范围的逐渐扩大,获得的相应的三维数据可达TB级。因此,如何合理的建立点云索引管理机制,是解决海量点云数据组织和管理的关键问题。本文首先对传统八叉树数据结构的索引方法进行了优化,然后对三维点云分块,建立八叉树索引数据文件,同时用LOD方法对其进行分层抽稀操作,通过建立改进的八叉树与LOD方法相结合的索引,来降低内存的消耗、提高查询的效率,最后根据屏幕显示范围与视角变化实时读取释放点云索引数据,从而实现海量点云数据的可视化。
The using of 3D scanning technology in digital city construction development rapidly, with the upgrade of hardware, the scanned 3D data up to TB.How to establish a point cloud index management mechanism as the key to solving the problem of massive point cloud data organization and management.This paper proposed a optimized index management method based on traditional Octree structure,and using LOD method to divided the massive point cloud data into blocks,then building the corresponding index data files for point cloud data blocks,in order to improve the performance and reduce the memory consumption of visualization of the massive point cloud data.
出处
《软件》
2016年第3期114-117,共4页
Software
