Automated rock mass characterisation using 3-D terrestrial laser scanning

Abstract

The research investigates the possibility of using point cloud data from 3-D terrestrial laser scanning as a basis to characterise discontinuities in exposed rock massed in an automated way. Examples of discontinuities in rock are bedding planes, joints, fractures and schistocity. The characterisation of discontinuities is of importance, since they determine to a large extend the geotechnical behaviour of the entire rock mass. The conventional way of characterising discontinuities is by manual geological survey using geological compass and measuring tape. A logical alternative to the conventional methods for surveying rock faces is the use of 3-D terrestrial laser scanning. A 3-D terrestrial laser scanning survey yield a 3-D point cloud but this data does not yet provide the information on the character of the discontinuities that can be seen in the rock exposure. In this research two different approaches are followed: the first approach uses surface reconstruction through interpolation of the point cloud and the second approach makes use of direct segmentation of the original point cloud. The main conclusion of this research is that it is possible to automate the derivation of discontinuity orientation and spacing information with both methods. Point cloud segmentation is however, the most preferred approach, since it does not require prior surface reconstruction, is therefore faster, and is not strongly influenced by vegetation and other noise in the data. Point cloud segmentation uses the original point cloud, so there is no data loss, which is unavoidable with a surface reconstruction approach.