A recent development in the use of lidar remote sensing techniques is ground-based laser scanning. Laser scanning of rock faces yields the spatial relation between all scanned rock surface points, at a very high resolution, basically a dense "point cloud" in three-dimensional space. The subject of this research is to obtain discontinuity information from the point cloud data set, using an approach that can be automated. The first step in this methodology is to interpolate the point cloud data using 3D Delaunay triangulation in order to create a 3D surface. As a 3D triangulated surface, the scanned rock face is represented by a large number of triangles. The orientation of each triangle can subsequently be computed using basic geometrical rules, analysis of the kernel density stereo plots of the orientation of all triangles, reveal that specific discontinuity sets can be recognised. Obviously, if this approach can be further developed and fully automated, this would give the site engineer or geologist, in realtime, evidence on the internal structure of any discontinuous rock mass. Particularly in areas where access to rock outcrops is poor, application of this technique will be very promising.
The idea to obtain discontinuity information from an exposed rock mass through remote sensing is not new. Analogue stereo photogrammetric techniques already allowed the measurement of orientations of individual discontinuities (Rengers, 1967). More recently, applications have been developed that use digital imagery and data processing instead. Basic photogrammetric principles combined with pattern recognition routines allow the user to create 3D models of visually any object in a cost-effective way (Pollefeys et al. 2000). These applications do, however, require time-consuming data processing to arrive at the final 3D model and still require manual outlining of discontinuity surfaces in order to calculate orientations. Feng et al. (2001) demonstrated the use of a non-reflector total station to measure fracture orientations. Although good results were obtained, the amount of data points that can be acquired is limited and the manual operation of the total station still requires a large amount of effort on-site. Laser scanning as a remote sensing technique may provide an attractive alternative approach. It is likely to offer the desired data quality and quantity to arrive at a statistically sound analysis of rock mass discontinuities without elaborate and time consuming data acquisition and data processing. The main advantage over photogrammetry is that laser scanning provides direct co-ordinate measurement and does not rely upon matching approaches to solve the correspondence problem between conjugate feature points within the scene (Baltsavias, 1999). In other words, ground-based laser scanning provides an instant 3D model with very high accuracy and data density, which will allow the user to analyse the data in realtime. The second motivation behind this research is that, considering this large accuracy and high spatial density of laser measurements and the possibilities that modern (3D) data modelling techniques offer, more information can be obtained from the reflected laser signal than merely the geometrical aspects.
