The Structure from Motion (SfM) photogrammetric technique has emerge as an efficient alternative for remote 3D rock mass characterization, compared to laser scanner (LiDAR) or stereoscopic photogrammetry, due to its economy and ease of use. In a similar way, the recent development of the drone-based technology has turn UAVs ("Unmanned Aerial Vehicles") in a more accessible device for field applications in geotechnical engineering; allowing the acquisition of high quality images from a safe distance and without the need to stablish direct contact with the rock mass. However, the close distance applicability of UAV-SfM photogrammetry has not yet been investigated in detail to characterize joint roughness at close range (<10 m). In this work we employ the SfM technique for the generation of 3D models of the joint surfaces from aerial images taken at a relatively short distance from the slope (10, 7.5, 5, and 2.5 m). Roughness profiles are extracted from the 3D data, and their Z2 statistical parameter is used to estimate the Joint roughness coefficient (JRC). Finally, the JRC value of those profiles -obtained with the UAV-SfM approach- have been compared with those obtained with traditional measurements based on manual methods. The proposed methodology is applied to a real case in an ancient open-cast mine in Northern Spain. The results obtained at different distances are compared to analyze the potential of UAV-SfM photogrammetry to develop accurate close-distance models. Results show that it is not necessary to get too close to the slope in order to get the best results, as this may cause overestimation of the JRC value.
Joint roughness is one of the main factors considered in the joint strength criteria [1, 2] and the JRC (Joint Roughness Coefficient) is the most commonly used methodology to quantify rock surface roughness on field applications. However, since its determination is often performed by manual methods (that need direct contact to the rock joint) and visual comparison (that can be very subjective), many authors aim to correlate this coefficient with other statistical parameters that can be objectively quantified [3, 4]. In that sense, Structure from Motion (SfM) technology has recently emerged as an efficient alternative for remote 3D rock mass characterization, becoming even more versatile when combined with Unmanned Aerial Vehicles (UAVs). For instance, Tomás et al. [5] have recent applied this technology using a heavyweight professional drone (DJI Matrice 600) for semiautomatic identification of discontinuities sets of a rocky railway cutting. Moreover, Erharter et al. [6] have used a compact portable drone (DJI Mavic Pro) for digital mapping and kinematic analysis of landslides in Alpine terrain, comparing field measurements made with compass to digitalized discontinuity measurements using manual, semi-automated and automated techniques. Similarly, Salvini et al. [7] have studied the applicability of 3D digital point clouds, employing medium weight professional drone (Aibot X6), to determine discontinuity surface roughness characteristic in a quarry slope using images taken from a distance higher than 10 m. However, the close distance (<10 m) applicability of UAV-SfM approach has not been yet investigated in detail to characterize joint roughness.