基于八叉树及KD树的混合型点云数据存储结构  

Integrated Point Cloud Storage Structure Based on Octree and KDTree

作　　者：廖丽琼[1] 白俊松[1] 罗德安[1] 

机构地区：[1]北京建筑工程学院测绘工程系,北京100044

出　　处：《计算机系统应用》2012年第3期87-90,共4页Computer Systems & Applications

基　　金：国家自然科学基金(40871196;40901192)

摘　　要：通过对现有点云数据存储结构进行综合分析及比较,提出了一种基于八叉树及KD树的混合型点云数据存储结构模型,文中对该模型的基本原理、实现步骤及快速索引的建立等进行了全面的论述,最后以一组实测数据为例,比较了KD树、八叉树和本文提出的混合结构三种不同数据组织方式的检索效率,证明了所提出存储结构的有效性及实用性。Based on analysis and comparison to existing point cloud storage structures, presents an integrated point cloud storage structure base on Octree and KDTree, describes its basic theory, work flow and index building method, at last a group of real data are employed to check and test the efficiency of Octree, KDTree and presented data structure, and the results shows that presented data structure has higher search efficiency and more suitable for point cloud data storage than the others.

关 键 词：海量点云 数据结构 三维激光扫描 存储结构 索引 

分 类 号：TP333[自动化与计算机技术—计算机系统结构]