基于实景三维建模和点云投影分析的岩体裂隙识别应用  

Application of 3D real scene modeling and point cloud projection analysis in rock mass crack identification

作　　者：刘大鹏 李志胜 LIU Dapeng;LI Zhisheng(Guangdong Surveying and Mapping Engineering Company Limited,Guangzhou,Guangdong 510700,China)

机构地区：[1]广东省测绘工程有限公司,广东广州510700

出　　处：《北京测绘》2024年第10期1431-1436,共6页Beijing Surveying and Mapping

基　　金：广东省科技计划(2021B1111610001)。

摘　　要：本研究旨在探讨无人机(UAV)实景三维建模技术和点云投影分析在岩体裂隙识别中的应用。首先,利用无人机搭载高清摄像镜头,对岩体进行高精度的三维实景扫描,获取大量的岩体表面数据;然后,通过图像处理和三维重建算法,对获取的数据进行处理和分析,实现对岩体裂隙的准确识别和测量;最后,结合实地验证和地质勘探实际需求,验证该技术在岩体裂隙识别中的可行性和有效性。研究结果表明,无人机实景三维建模和点云投影分析能够实现对岩体裂隙的高精度识别和三维重建,裂缝识别准确率由83.4%提升至93.1%,漏提率由30.3%降低至15.4%。该技术不仅在提高识别精度和效率方面具有显著优势,同时也能够降低传统岩体勘探过程中的人力和物力成本。This study explored the application of unmanned aerial vehicle(UAV)-based three-dimensional(3D)real scene modeling technology and point cloud projection analysis in rock mass crack identification.Firstly,a UAV equipped with a high-definition camera lens was used to conduct high-precision 3D real scene scanning of the rock mass and obtain a large amount of surface data of the rock mass.Secondly,through image processing and 3D reconstruction algorithms,the obtained data was processed and analyzed to achieve accurate identification and measurement of rock mass cracks.Finally,according to on-site verification and actual geological exploration needs,the feasibility and effectiveness of this technology in identifying rock mass cracks were verified.The research results indicate that UAV-based 3D real scene modeling and point cloud projection analysis can achieve high-precision identification and 3D reconstruction of rock mass cracks,and the accuracy of crack identification has increased from 83.4%to 93.1%,while the omission rate has decreased from 30.3%to 15.4%.This technology not only has significant advantages in improving identification accuracy and efficiency but also can reduce the labor and material costs in traditional rock mass exploration.

关 键 词：无人机(UAV) 实景三维建模 岩体裂隙识别 点云投影 

分 类 号：P231[天文地球—摄影测量与遥感]