基于点云的管道三维自动建模方法研究

Study on Pipeline 3D Automatic Modeling Method Based on Point Cloud

摘要:

随着我国城市地下空间的不断建设，各种类型的管廊已经成为社会和人们生产生活中不可缺少的物质条件。本文针对市政基础设施、地下空间中错综复杂的管道难以观测和三维建模的问题，采用三维激光扫描技术获取管廊的点云数据，并对点云数据中存在的噪点采用拉普拉斯算法进行去除，以提高管道中心线的提取效率。在此基础上提出一种基于RANSAC算法的管道中心线提取方法，并对提取的中心线进行等平面处理，基于OpenGL和MFC框架开源库结合RANSAC算法研发了利用点云进行管道三维自动建模模块，并实现了复杂管道的可视化。

Abstract:

Accompanying with continuous construction of urban underground space in China, various types of pipe corridors have become indispensable material conditions in society and production and life of people. In this paper, pointing to difficult observation and 3D modeling of complex pipelines in municipal infrastructure and underground spaces, by using three-dimensional laser scanning technology, the point cloud data of the corridor has been gained. By using Laplacian algorithm, the remove noise in the point cloud data has been removed to improve the extraction efficiency of the pipeline centerline. A method of pipeline center line extraction based on RANSAC algorithm has been proposed, and the center line in the same plane has been made. Through the OpenGL and MFC open source library framework, combining with the RANSAC algorithm, the gallery 3D modeling module by using the point cloud visualization of petrochemical pipeline has been set up.