Identifying and mapping potentially adverse discontinuities in underground excavations using thermal and multispectral UAV imagery
Identifying and mapping potentially adverse discontinuities in underground excavations using thermal and multispectral UAV imagery
复制标题
DOI：
10.1016/j.enggeo.2019.105470
复制
发表时间：
2020-03
影响因子：
7.4
通讯作者：
R. M. Turner;M. Maclaughlin;S. Iverson
中科院分区：
文献类型：
--
作者：
R. M. Turner;M. Maclaughlin;S. Iverson
岩土制图、岩体特征和地下开挖检查是确保开挖几何形状和地面支撑设计适合稳定条件的关键。缺乏通道、时机不佳或危险条件都是阻碍工程人员在现场收集高质量数据的因素。无人机(UAV)和便携式传感器技术的进步现在允许工程人员从地下工作场所远程捕获数据，这些数据可用于开发设计、生成分析输入和做出更明智的决策。地面坠落是地下环境中许多事故的源头，也是许多岩土工程勘察和分析的重点。在这项研究中，作者调查了如何使用便携式无人机安装的热成像仪和多光谱成像仪来探测和量化受构造控制的坚硬岩石中的不利地质不连续性。多个热、多光谱、RGB(红、绿、蓝)和LiDAR(光探测和测距)数据集是从美国西部蒙大拿州巴里克黄金阳光矿的5个地下露天采矿场的5个飞行中获取的。使用现成的软件，对热图像、RGB图像和多光谱图像进行处理，以创建单独的三维点云和网格，并使用LiDAR数据进行地理参考。使用免费软件CloudCompare中的工具绘制了在三维点云和网格中识别的不连续面。其中一个热模型确定了露天采场顶柱中发育的楔形体，而四个多光谱模型包含了足够详细的信息，可以绘制单个不连续面。调查结果表明，这些便携式成像仪是一种可行的工具，可用于帮助工程人员识别和绘制不利的地质不连续面和独特的岩体成分。本文描述了在地下挖掘中捕获、处理和解释热图像和多光谱图像的技术，并可作为今后对个别地下场地进行调查的基础。
Geotechnical mapping, rock mass characterization, and inspections of underground excavations are critical to ensuring that excavation geometry and ground support design are appropriate for stable conditions. A lack of access, poor timing, or hazardous conditions are all factors that can prohibit engineering personnel from collecting high-quality data in the field. Advances in unmanned aerial vehicles (UAV) and portable sensor technology now allow engineering personnel to remotely capture data from underground workings that can be used for developing designs, generating inputs for analyses, and making more informed decisions. Ground falls are a source of many accidents in the underground environment, and they are the focus of many geotechnical investigations and analyses. In this study, the authors investigated how portable UAV-mounted thermal and multispectral imagers could be used to detect and quantify adverse geological discontinuities in hard rock masses that are structurally controlled. Multiple thermal, multispectral, RGB (red, green, and blue), and LiDAR (light detection and ranging) data sets were captured from 5 flights in sub-level open stopes at the Barrick Golden Sunlight Mine in the state of Montana in the western United States of America. Using off-the-shelf software, the thermal, RGB, and multispectral images were processed to create individual three-dimensional point clouds and meshes, which were georeferenced using the LiDAR data. Discontinuities identified in the three-dimensional point clouds and meshes were mapped using tools found in the freeware CloudCompare. One of the thermal models identified wedges that had developed in the crown pillar of an open stope, and 4 of the multispectral models contained enough detail for mapping individual discontinuities. The results of this investigation indicate that these portable imagers are viable tools that can be used to aid engineering personnel in identifying and mapping adverse geological discontinuities and unique rock mass composition. The techniques for capturing, processing, and interpreting thermal and multispectral imagery captured in underground excavations are described in this paper and can be used as the basis for future investigations at individual underground sites.