基于车载LiDAR技术的公路三维建模与应用

发布时间：2018-05-27 23:29

本文选题：车载激光测量系统 + 点云　；参考：《首都师范大学》2013年硕士论文

【摘要】：随着公路交通行业的快速发展,新增路网大量需求,又由于早期建成的公路设计标准较低和超大服役等因素,使得道路新建、改建、扩建、维护等工程对专业测量的高效、精确、快速、安全等需求日益迫切,车载LiDAR (Light Detection And Ranging)技术应运而生。车载激光测量系统可以快速采集公路三维信息,包含点云坐标、属性信息、影像数据等,点云具有精度高、信息丰富等特性。车载LiDAR技术应用于公路测量相关任务,对提高效率和节约成本具有重大意义。本文简要介绍了利用车载激光测量系统进行公路数据采集的工作方法以及技术要点,然后详细阐述了基于车载点云的公路三维建模方法,并对关键算法进行了设计与实现,主要包括地面点滤波、数据修补与抽稀、点云构网以及特征地物分类提取等内容。原始数据包含路面、护栏、车辆、附属设施、树木和草丛等信息,首先采用基于扫描线的滤波方法将地面点与非地面点分离,接着对因遮挡造成的路面数据损失进行修补,并可以有选择性地进行数据抽稀,然后基于扫描线间三角剖分的方法构建三角网；另外,本文特征地物分类提取的研究对象是树和路灯,主要根据其各自空间形态特点进行。目前,车载激光测量系统的应用日渐增多,车载点云处理技术也亟待完善,针对不同专业应用领域需要有相应的处理技术和方法。本研究即对车载LiDAR技术应用于公路三维信息获取与建模提出了解决方案和关键技术方法,并选取了一段具有代表性的公路点云数据,对提出的方法进行了验证分析,得到了较理想的结果。结果证明,车载点云的精度满足公路测量精度需求,基于车载点云生成的公路三维模型可以实现公路勘测相关应用,如生成纵、横切面剖面图,计算面积和土方量,查询任意点位信息,动画展示和水淹没分析等,该技术可应用于公路勘测设计和改扩建、公路基础设施普查、公路景观可视化等工程。
[Abstract]:With the rapid development of highway traffic industry, a large number of new road network demand, and due to the early built highway design standards and oversized service and other factors, such as road construction, reconstruction, expansion, maintenance and other projects to professional survey efficiency. The demand of precision, speed and security is more and more urgent. The LiDAR Light Detection And Ranging) technology emerges as the times require. The vehicle laser measurement system can quickly collect 3D highway information, including point cloud coordinates, attribute information, image data, etc. The point cloud has the characteristics of high accuracy and rich information. It is of great significance to improve the efficiency and save the cost by applying the LiDAR technology to highway survey related tasks. This paper briefly introduces the working method and technical points of highway data acquisition using vehicle-mounted laser measurement system, and then expounds the road 3D modeling method based on vehicle point cloud in detail, and designs and implements the key algorithm. It mainly includes ground point filtering, data repairing and thinning, point cloud structure network and feature classification and extraction. The original data includes road surface, guardrail, vehicle, ancillary facilities, trees and grass. Firstly, the ground point is separated from the non-ground point by the method of scanning line filtering, and then the road surface data loss caused by occlusion is repaired. The data can be thinned selectively, and then triangulation can be constructed based on the method of triangulation between scanning lines. In addition, tree and street lamp are the research objects of feature extraction in this paper, which are mainly based on their spatial morphological characteristics. At present, the application of vehicle laser measurement system is increasing day by day, and the point cloud processing technology of vehicle needs to be improved urgently, and the corresponding processing techniques and methods are needed for different professional application fields. In this study, we put forward the solution and key technology methods for the application of vehicle-mounted LiDAR technology in highway 3D information acquisition and modeling, and selected a representative section of road point cloud data, and analyzed the proposed method. A better result is obtained. The results show that the accuracy of vehicle point cloud can meet the needs of highway measurement accuracy, and the highway 3D model based on vehicle point cloud can be used for highway survey, such as the generation of longitudinal and cross-sectional profiles, the calculation of area and earthwork. The technology can be used in highway survey, design, reconstruction and extension, highway infrastructure survey, highway landscape visualization and so on.
【学位授予单位】：首都师范大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P225.2;U412.2

【参考文献】