车载LiDAR点云数据分割与半自动化建模方法

Method of segmentation and semi-automatic modeling for vehicle-borne LiDAR point cloud data

针对车载激光雷达(light detection and ranging,LiDAR)点云数据的不完整性问题,提出一种车载LiDAR点云数据分割以及基于分割后点云数据的半自动化建模方法。首先对点云数据进行标准格式转换及稀化;然后以不同地物的属性和几何特征为分割条件,分别建立道路、建筑物、树和路灯等附属设施的三维模型,并利用车载以及航空图像的纹理信息辅助建筑物的立面和顶面三维建模;最后以真实街景为实验区,基于拓普康IP-S2车载LiDAR点云数据,完成该街景的分割与建模。实验结果表明,该文提出的点云数据分割与街景地物重建方法比较简单,可实现道路和建筑物的半自动化分割;利用成熟的建模软件和方法,实现了建模的完整性和较强的可靠性。

A method of segmentation and semi - automated modeling for vehicle - borne light detection and ranging （LiDAR） point cloud data is presented in this paper. Firstly, the LiDAR point cloud data are converted into standard format and sampled sparsely. Then the geometric features of different objects are used to govern the data and model of 3D roads, buildings, trees, power poles and facilities. In view of the imperfection of vehicle - borne LiDAR point cloud data, the vehicle - borne and texture information of aerial image auxiliary is used to build the facade and the top surface of the three - dimensional modeling. Finally, a streetscape is reconstructed by using IP - $2 vehicle -home LiDAR point cloud data. The results show that the method proposed in this paper is simple and suitable for semi - automatic segmentation of roads and buildings.