摘要
在隧道建设中,开挖面围岩蕴含着指导施工的珍贵地质信息,围岩结构面提取作为围岩精细化分析判别的关键步骤至关重要。传统围岩三维数据获取多依赖激光扫描技术或多视角视觉三维重建技术,存在采集速度慢和无法恢复真实尺度的问题;同时,结构面提取方法在应对大范围且破碎的三维岩体数据时,存在通用性不足和准确性低的问题。针对隧道围岩的结构面快速提取,设计了一套快速三维采集的方法设备,提出了一种基于分域聚类算法的结构面提取方法。首先,快速采集设备集成视觉成像技术和激光测点技术,将图像数据和激光点距数据标定融合,实现了快速准确采集;其次,提取方法将复杂分布的点云数据空间划分为多个子域独立提取结构面;最终,融合子域间结果,实现了围岩峒壁和掌子面区域主要结构面的自动提取。所设计的设备和方法在一个实际施工隧道上进行了示范应用,单个掌子面数据采集时间小于10 min,减少了对施工时间的影响和对采集人员的技术依赖;经过后处理恢复的数据尺度,精度误差小于1 cm;所设计的结构面提取的准确率可达到60%以上,可协助施工工程师掌握施工现场情况。
In tunnel construction,important geological information is contained in surrounding rocks of the excavation face.Extracting rock discontinuities is a key step in detailed analysis of surrounding rocks.Traditional methods for surrounding rock 3D data acquisition mostly depend on laser scanner and multi-view stereo 3D reconstruction technology,which cause problems of low acquisition speed and lack of real rock size.Meanwhile,extracting rock discontinuities have problems of insufficient versatility and low accuracy when dealing with 3D data of fragmented rock masses in a large area.Concerning the fast extracting of rock discontinuities in the tunnel,a set of acquisition equipment was designed and a rock discontinuity extraction method based on domain clustering was proposed.Firstly,the fast acquisition equipment integrated the vision imaging technology and laser point-measurement technology was used to calibrate and fuse the image data and laser point distance data to achieve fast and accurate acquisition.Then,the complex distributed point cloud data space were divided into multiple sub-domains to extract rock discontinuities independently.Finally,the results between sub-domains were merged to realize the automatic extraction of main discontinuities in the surrounding rock wall and tunnel face area.The designed equipment and the method were applied in an actual construction tunnel.The acquisition time of a single tunnel face is less than 10 min,which reduces the impact on construction duration and the technical dependence on acquisition staff.After post-processing,the accuracy error of the restored data scale is less than 1 cm and the accuracy of rock discontinuities extraction can reach more than 60%,which can help the construction engineers master the construction site conditions.
作者
苗长龙
李世东
陈宝林
张继锋
谢炜婕
全珍玲
郑晓卿
MIAO Changlong;LI Shidong;CHEN Baolin;ZHANG Jifeng;XIE Weijie;QUAN Zhenling;ZHENG Xiaoqing(Yangtze Delta Region Institute of Tsinghua University,Zhejiang,Jiaxing Zhejiang 314006,China;Zhejiang Provincial Institute of Communications Planning,Design&Research Company Limited,Hangzhou Zhejiang 310006,China)
出处
《计算机应用》
CSCD
北大核心
2021年第S01期336-340,共5页
journal of Computer Applications
关键词
围岩结构面提取
点云聚类
三维数据分析
三维数据采集
多视角成像
surrouding rock discontinuities extraction
point cloud clustering
3D data analysis
3D data acquisition
multi-view imaging
