一种利用面元拟合的地面点云数据三维R树索引方法

3D R-Tree Point Clouds Data Indexing Method Based on Plane Fitting

针对地面激光扫描的真三维点云数据,提出基于面元拟合的三维R树索引方法,在对点云进行递归分割和面元拟合的过程中形成三维R树节点的包围盒,通过对传统R树的改进,使之更好地保留平面特征,以适应点云数据的管理。通过实验实现了地面点云数据的高效管理与查询检索,证明了方法的有效性。

This paper proposed a data structure based on plane feature and R-tree indexed for terrestrial laser scanning point clouds organization. We construct bounding box of R-tree node by using Recursive partitioning algorithm and considering plane feature as well to or- ganize TLS data. The revised R-tree method is more efficient for TLS data management and query. The paper also shows an experiment to prove effectiveness of the method.