基于点云的岩体间距与粗糙度的自动化提取

Automatic Extract of Rock Mass Spacing and Roughness Based on Point Clouds

基于单相机双目三维重构技术,提出一种获得隧道掌子面三维点云模型的规范化拍摄流程和自动化提取岩体间距与粗糙度特征参数的方法。间距信息自动化提取流程为采用基于张量投票理论及其优化处理技术识别结构面迹线,通过分组、产生虚拟测线与迹线相交,从而得到各组的平均间距。粗糙度信息自动化提取流程为对岩体分别沿着水平和垂直方向均匀切割获得二维粗糙度轮廓线,采用均方根与粗糙度系数(JRC,joint roughness coefficient)的关系计算不同位置不同方向的JRC。将该方法应用于安徽明堂山隧道在建隧道掌子面三维点云模型的获取以及岩体间距和粗糙度特征参数的提取中,并给出了计算结果。基于单相机双目三维重构技术优势在于可在环境恶劣的在建隧道中获得稳定的、精度较高的掌子面三维点云模型;对岩体间距和粗糙度特征参数进行自动化提取,作为自动化的测试计算方法,可为工程应用提供一定参考。

A normalized images production method for the generation of 3D point clouds of tunnel face based on the binocular three-dimensional reconstruction theory with single camera is provided. An effective method for automatic rock mass spacing and roughness detection is also proposed. The main flow of automatic spacing extraction is first detecting traces using the Normal Tensor Voting Theory and post-processing technique,then grouping traces,generating a virtual line to intersect with traces and finally calculating average space of each group. The main flow of automatic roughness calculation is first generating vertical and horizontal cuttings at equal intervals to get the corresponding surface roughness profiles,then calculating the JRC values in different directions and different positions according to the relationship between JRC and the root mean square. The method was applied to obtain 3D point clouds of tunnel faces of Mingtang tunnel in Anhui,and spacing and roughness information were extracted from these3 D models. The results show stable 3D point clouds model of tunnel face with high precision can be generated based on the binocular three-dimensional reconstruction theory in the conditions of tunnel under construction. The automatic spacing and roughness detection method could be used as a supplement to traditional direct measurement method.