基于激光扫描的隧道管片结构变形数据处理及可视化分析方法

An Automatic Data Processing and Visualization Analysis Method for Deformations of Tunnel Segment Structure Using Laser Scanning Data

基于激光扫描技术(简称LST),提出了一种数据处理方法用于隧道管片结构变形的可视化分析。通过LST获得的三维点云数据通过投影及极坐标转换,可计算得出隧道管片结构内弧面闭合轮廓的二维横断面。根据二维横断面中各轮廓点与设计坐标的偏差,可得到隧道管片结构内弧面上各位置处的变形值,理论精度在1 mm以内。为降低计算误差并实现结构变形的可视化分析,文章采用了一种"聚合"的点云处理算法,对点云数据进行降噪,并提出结构内弧面轮廓上的非正常点。采用传统的数据处理方法仅能识别特定规则的隧道横截面轮廓,而通过该方法则可结合LST得到的三维点云数据对内部轮廓无遮挡关系的闭合隧道断面进行轮廓识别及变形分析。结果表明,该方法能够提供准确的变形数据,具有较高的数据处理效率,可用于实验室隧道管片结构变形监测,并能够满足工程应用的要求。

Based on Laser Scanning Technology(LST),a data processing method is proposed to automatically visualize the tunnel cross sections and collect deformation information.In this method,Cartesian coordinates of each data point are transferred to polar coordinates,and two-dimensional tunnel cross sections can be extracted from three-dimensional point cloud data obtained by LST.The tunnel deformation is calculated by subtracting radius vectors of deformed cross section from designed cross section and visualized through adjustable deformation factors.The deformation data at arbitrary azimuthal angles could be gained with an accuracy of 1 mm.To achieve the ideal calculation accuracy and ensure visualization performance,a novel algorithm adopting point discretization and regeneration technique is developed to automatically detect and treat the noise and discontinuity points from the point cloud data.While the traditional data processing methods could only handle a specific type of tunnel cross section,the proposed method is able to deal with arbitrary shapes of tunnel sections as long as three-dimensional cloud data are available.The results show that the proposed method could provide accurate deformation data with high data processing efficiency for monitoring the deformation of tunnel cross sections in laboratory,and potentially for real-life tunnels.