隧道断面自动提取方法研究  

Research on the Automated Extraction of Tunnel Cross Section

作　　者：许磊[1] 王长进[2] 

机构地区：[1]武汉大学,武汉430079 [2]铁道第三勘察设计院集团有限公司,天津300142

出　　处：《铁道工程学报》2016年第8期94-99,共6页Journal of Railway Engineering Society

摘　　要：研究目的:在隧道施工和运营期间,需要经常测量隧道结构断面,传统基于全站仪或断面仪的方法测量效率较低。采用地面三维激光扫描仪,对隧道进行分站式扫描,将各站扫描点云进行拼接融合,可以快速获得整条隧道激光点云。目前基于激光点云的隧道断面提取都是针对圆形和矩形的隧道形状,不适用多种结构类型的隧道。本文提出一种基于射线相交的隧道断面自动测量方法,满足多种隧道结构断面测量要求,并结合工程实际对该方法的精度和可靠性进行验证。研究结论:(1)利用三维激光点云数据进行隧道断面自动测量,不仅效率高,而且精度高;(2)基于八叉树算法对离散点云进行组织管理,采用随机采样一致性算法对数据进行滤波,实现断面点云噪声点剔除;(3)采用矩形分割算法实现隧道断面点云快速分割,基于射线相交的方法可实现多种设计形状的结构断面测量;(4)该研究结果可应用于施工或运营期间多种设计形状的隧道结构断面自动测量。Research purposes： During tunnel construction and operation, tunnel cross section must be measured often. The traditional method using total station or profile measure instrument is time - consuming. The current methods are based on cylinder and rectangletype of tunnel, which can not meet the demand of variety type of tunnel structural section measurement. This paper took substation scanning of the tunnel via ground - based laser scanner, then matched and fused the point cloud data of each scanning station. Prosing a new automatic structural section measuring algorithm can realize all types of structures＇tunnel contour measuring task. Considering actual engineering project, we have tested the proposed method and accuracy. Research conclusions：（ 1 ）Realizing the automatic measurement of tunnel cross -section with three- dimensional laser point cloud data, this method can achieve high efficiency and high precision at the same time. （2）To organize the raw point cloud data based on octree algorithm and filter data noise using random sample consensus （RANSAC） algorithm. （3） Section cloud can be segmented using rectangle algorithm. Variety tunnel section can be automatically measured based on ray line intersection. （4）The research results can be used in tunnel structural section measuring, which is fitted the demand of variety tunnel section design type during construction and operation.

关 键 词：隧道断面 三维激光扫描 点云分割 随机采样一致性 

分 类 号：P225[天文地球—大地测量学与测量工程]