三维激光扫描点云的岩体产状智能识别研究
Research on Intelligent Recognition of Rock Mass Occurrence by Using 3D Laser Scanning Point Cloud
赵军 1陈宝龙1
作者信息
- 1. 安徽理工大学土木建筑学院,安徽 淮南 232001
- 折叠
摘要
目的 为了降低传统的人工测量方式带来的工作量大、效率低、工作环境恶劣等影响.方法 提出在深部工程或高陡边坡的地质工程测量中,采用三维激光扫描技术获取岩体结构面点云数据,结合一个特制三角形标靶将三维激光点云的局部坐标系下的法向量转化为大地坐标系下的法向量,再利用区域生长算法对结构面产状进行智能识别和输出.结果 试验结果表明,利用Delaunay三角剖分三维点云表面的重建,在改进区域生长算法对点云数据的处理下,能够对大批量数据快速处理并获得结果,其独特的空圆特性、剖分中产生较少的畸形三角形和耗时极少的运行时间非常适合海量的点云数据的计算.结论 和传统方法相比,该方法具有快速、自动、高精度和远距离测量的优点,对无GPS信号或者GPS信号弱的岩爆频发地区的地质测量具有重要意义.
Abstract
Objective In order to solve the problems of the traditional manual measurement methods,such as large workload,low efficiency and poor working environment.Methods It was proposed to use 3D laser scanning tech-nology to obtain the point cloud data of rock structure surface in the geological engineering measurement of deep engineering or high steep slope.With a special triangular target used to transform the normal vector under the lo-cal coordinate system of the 3D laser point cloud into the normal vector under the geodetic coordinate system,the area growth algorithm was utilized to identify intelligently the structural surface yield and output.Results The ex-perimental results showed that the reconstruction of 3D point cloud surfaces by using Delaunay triangular profi-ling,with the improved region growing algorithm for point cloud data,was able to process large amounts of data quickly and obtain the results.Its unique null-circle characteristics,production of fewer aberrant triangles in the profiling and low running time were well suited to the computation of large amounts of point cloud data.Conclu-sion Compared with the traditional method,this method has the advantages of fast,automatic,high-precision and long-distance measurement,which is of great significance for the geological measurement in the rock burst-prone areas without GPS signals or with weak GPS signals.
关键词
三维激光扫描/特制三角形标靶/深部工程/结构面产状Key words
3D laser scanning/special triangular target/deep engineering/structural plane occurrence引用本文复制引用
基金项目
国家自然科学基金资助项目(41102198)
国家自然科学基金资助项目(51204168)
出版年
2024
NETL
NSTL
万方数据