In this study, the possibility of measuring rock surface roughness by means of open-source photogrammetric methods was investigated. With simple commercially available digital cameras, samples of varying roughness were measured. As a reference, the samples were also measured with a high-resolution white light strip scanner and additionally manually with a contour gauge. By comparing the digital datasets it became clear that the open-source structure-from-motion (SFM) algorithms were able to capture the overall topography but with clearly less accuracy than the white light scanner. This had a marked influence on the determined roughness parameter JRC (joint roughness coefficient). For smooth surfaces the JRC was higher than the reference value, for rough surfaces it was lower. In general, with the introduced method it was possible to reproduce the surface of the rocks but care has to be exercised when roughness parameters are to be deduced from the datasets.
The shear strength of rock joints is primarily dependent on their surface roughness. In order to achieve a save and cost-efficient design in rock engineering it is desirable to measure joint roughness as accurate as possible. In the past, tools like contour gauges (Barton & Choubey 1977) and mechanical profilometers (e.g. Weissbach 1978) were used to measure unevenness of rock surfaces. Nowadays, high-resolution laser scanners (e.g. Fardin et al. 2001, Milne et al. 2009) or white light strip scanners (Tatone & Grasselli 2013) are applied to investigate rock surfaces. But, as these methods being cost-intensive, alternatives are needed. Therefore, in this study, open-source photogrammetric software was utilized to measure surface roughness. With a simple off-the-shelf digital camera in connection with well-established structure from motion algorithms five surfaces varying in roughness were investigated.