The influence of discontinuities on the mechanical behavior of rock masses, demands a detailed knowledge about the geometrical properties of the existing discontinuity network. Traditional measurement techniques provide a rough knowledge about a discontinuity network but are also prone to bias. To increase the reliability of discontinuity models, remote sensing techniques like Close-Range Terrestrial Digital Photogrammetry (CRDTP) are increasingly applied for rock mass characterization. This research contributes to the trend of automatic rock mass characterization and focuses on the analysis of a digital surface model. This paper proposes a method to identify discontinuity sets in a point cloud and calculate the spacing of the sets. The discontinuity sets are semi-automatically identified with the open-source software DSE (Discontinuity Set Extractor). The program analyzes the density distribution of the point normal vectors in combination with a co-planarity test. The subsequent calculations apply DBSCAN to cluster and assign the points in the point cloud to singe discontinuity sets. After identifying the different discontinuity sets, the sub-members of each discontinuity set are again searched with DBSCAN in Matlab® (The Mathworks Inc.) and exported as a structure map to calculate the normal spacing between the single members of each set with ShapeMetriX3D Analyst. The point cloud is generated with CRTDP, using the digital mapping tool ShapeMetriX3D (3GSM GmbH), which is also used to validate the results of the calculations.
There hardly exists a rock mass without three or more different discontinuity sets. Hence, it is of crucial importance in rock engineering to have a profound knowledge about the discontinuity network within a rock mass, since present discontinuities, like bedding planes, joints or cracks, represent the weak zones, which separate singular polyhedrons of competent rock blocks and control failure mechanisms. Furthermore, discontinuities influence the mechanical properties and the behavior of rock masses.
Geotechnical engineers traditionally acquire this data by geological and geotechnical surveys, which combine visual and manual technologies to characterize rock masses. The most common tool is a geological clino-compass to directly measure the dip and dip direction of discontinuities. This method can be extended by scan lines and mapping windows which reduce the inherent bias of a single measurement. But traditional geotechnical data acquisition is nonetheless limited in accessibility, geological or geotechnical knowledge, time and scale. The results are often subjective rather than objective and therefore not reproducible [1, 2].
