基于分块ROAM算法的大规模路面点云重建  

Large-scale Point Cloud of Pavement Reconstruction based on Blocked ROAM Algorithm

作　　者：吴启芳[1] 唐好选[1] 顾海燕[1] 

机构地区：[1]哈尔滨工业大学计算机科学与技术学院,哈尔滨150001

出　　处：《智能计算机与应用》2015年第5期79-83,共5页Intelligent Computer and Applications

基　　金：国家自然基金(51138003)

摘　　要：本文主要工作是根据车载激光获取的路面点云数据,在尽量保留路面特征的基础上,建立路面细节层次模型。通过对比相关技术,研究了基于分块的ROAM算法,实现了实时调度大规模路面点云并建模的工作。首先将点云数据通过内存映射读入内存,选取进行建模的初始点云数据,对建模数据进行分块,根据与视点的距离建立两个不同评价准则,建立二元三角树,实现了三维路面可视化及漫游,并利用文件数据动态更新显示的数据。实验结果表明,文中算法能够很好地处理大规模点云数据,并且能保留路面特征。The main work of this paper is to establish pavement level of detail model according to pavement surface point cloud data obtained by the vehicle—borne laser,based on retain pavement features as much as possible. Through the comparison of related technology,the paper analyzes the ROAM algorithm based on block,realizes the real- time scheduling and modeling of large- scale pavement point cloud. First the paper reads point cloud data into memory through memory mapped,then selects the initial point cloud data modeling. After that,the modeling data are divided into blocks,and according to the distance from viewpoint two different evaluation criteria are set up. At last,establishes Triangle Bintree,achieves The 3D road visualization and the roaming,therefore uses the data dynamic to update data. The experimental results show that this algorithm can deal with large- scale point cloud data,and can keep the pavement characteristics.

关 键 词：路面重建 点云 细节层次 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]