摘要
盾构法因安全高效在地铁隧道等交通设施建设中被广泛应用,为保障建成后隧道的平稳运行需进行定期稳定性检测。相对于全站仪等传统低效率作业方法,短时间内获取大量数据的三维激光扫描技术逐渐成为盾构法隧道检测的首选方法。但隧道内管线、轨道等设施同样被扫描并掺杂在获得的点云数据中,成为影响隧道断面参数计算的噪声点。本文提出一种基于三维最小二乘与RANSAC算法的隧道断面检测方法,该方法首先基于三维最小二乘(Three-Dimensional Least Square Method,3D-LSM)计算隧道点云中轴线整体方向向量来获取隧道断面,进而将获取的三维断面数据转换到二维平面上,随后基于随即抽样一致算法(Random Sample Consensus,RANSAC)建立拟合去噪模型,根据断面数据拟合计算隧道断面的半径和椭圆度。通过青岛某地铁隧道精密检测工作表明:该方法能够应用于三维激光扫描计算隧道断面的椭圆度和半径中,并有效地克服了噪声点对拟合精度的扰动,提高模型拟合计算精度和对粗差点抵抗性。
The shield tunneling method is widely used in the construction of transportation facilities such as subway tunnels due to its safety and efficiency.Regular stability testing is required to ensure the smooth operation of the tunnel after completion.Compared to traditional inefficient operation methods such as total stations,three-dimensional laser scanning technology that obtains a large amount of data in a short period of time has gradually become the preferred method for shield tunneling inspection.However,facilities such as pipelines and tracks inside the tunnel are also scanned and doped in the obtained point cloud data,becoming noise points that affect the calculation of tunnel cross-section parameters.This article proposes a tunnel section detection method based on the 3D least squares and RANSAC algorithm.The method first calculates the overall direction vector of the tunnel point cloud axis based on the 3D least squares(3D-LSM)to obtain the tunnel section,and then converts the obtained 3D section data to a 2D plane,subsequently,a fitting denoising model was established based on the Random Sample Consensus(RANSAC)algorithm,and the radius and ellipticity of the tunnel cross-section were calculated by fitting the cross-section data.The precision detection work of a subway tunnel in Qingdao shows that this method can be applied to 3D laser scanning to calculate the ellipticity and radius of the tunnel section,effectively overcoming the disturbance of noise points on the fitting accuracy,improving the model fitting calculation accuracy and resistance to gross errors.
作者
陈凯
邵成立
宫宁
刘建英
黄鹏
陈帅
CHEN Kai;SHAO Chengli;GONG Ning;LIU Jianying;HUANG Peng;CHEN Shuai(Qingdao Surveying&Mapping Institute,Qingdao 266032,China;Qingdao Key Laboratory of Marine and Land Geographic InformationIntegration and Application,Qingdao 266032,China;National Marine Data and Information Service,Tianjin 300012,China)
出处
《城市勘测》
2024年第5期155-159,163,共6页
Urban Geotechnical Investigation & Surveying
基金
国家自然科学基金(41474001)。
关键词
隧道检测
点云数据
三维最小二乘法
RANSAC算法
隧道断面半径和椭圆度
tunnel detection
point cloud data
three dimensional least squares method
RANSAC algorithm
the radius and ellipticity of tunnel cross-section
