Investigation of Landslide Kinematics Derived from UAV- and Ground-based Imagery - Precision Analysis of 3D Point Clouds

During an eight week field campaign at the Super-Sauze landslide (South French Alps), several UAV- and ground-based photographs have been acquired. Additionally, high-resolution terrestrial laser scanning (TLS) has been performed in order to capture a precise reference data set. The High-resolution and multi-temporal, but also disordered and oblique images were used to generate 3D point clouds from Structure-from-Motion and Multi-View Stereo (SfM-MVS) algorithms. These multi-temporal digital surface models (DSMs) were required for a precise ortho-rectification, as well as for morphometric and kinematic investigations. In order to determine the quality of the point clouds, all photogrammetric data were compared to the time corresponding TLS-based DSMs. The analysis of DSMs and orthophotos is one part of a field experiment that has the objective to characterize the relationship between fracture processes (slidequakes and local tremor events), slope movements, fissure developments and pore water pressure build-up. The main focus of this contribution is a comprehensive evaluation of the SfM-MVS-based point clouds in order to demonstrate the potential but also the limits of such approach.