Towards semi-automatic discontinuity characterization in rock tunnel faces using 3D point clouds

Searching for an efficient and reliable method to reduce manual intervention and subjective parameter selection during the discontinuity characterization process of rock tunnel faces is an important task. This paper presents a new method for semi-automatic discontinuity characterization in rock tunnel faces using 3D point cloud data, which consists of the following five procedures: (1) regions of interest selection, (2) octree-based local curvature calculation, (3) automatic discontinuity sets classification, (4) cluster analysis for point cloud data, and (5) visualization of classified discontinuity sets. To evaluate the performance of the proposed method, the point cloud data of three rock tunnel faces are collected and established using a photogrammetry-based scheme. Experiments are carried out to identify the optimized parameter values (point cloud density, plane radius, and discontinuity set number) for discontinuity characterization in a typical rock tunnel face. Using the optimized parameter values, the proposed method showed excellent performance and the statistical discontinuity orientation data were in good agreement with those obtained through in situ measurements. In comparison with the other two state-of-the-art methods (the Discontinuity Set Extractor (DSE) method and the qfacet fast marching (qfacet FM) method), the proposed method demonstrates improved processing efficiency and reasonable accuracy for discontinuity characterization using 3D point clouds. Overall, this study provides a step forward towards automatic 3D discontinuity characterization and visualization in the field for rock tunnel engineers.