UAV-based rockfall hazard detection via roughness analysis in Karaköprü, Şanlıurfa using photogrammetric point clouds

Abstract

This study focuses on rockfall hazard detection in urban terrain, particularly the Karaköprü district of Şanlıurfa, utilizing UAV-based photogrammetric dense point cloud analysis. The research introduces an innovative approach that combines advanced geospatial technology and roughness analysis to identify and assess rocks situated above ground surfaces, thereby mitigating potential risks to transportation routes and buildings. The methodology involves quantifying surface irregularity using roughness analysis, where the distance between points and their best-fitting planes is computed based on a carefully selected kernel size. The results demonstrate that this approach effectively marks rocks across the study area, albeit with considerations for potential misclassifications. The dataset is derived from a photogrammetric UAV flight, yielding over 77 million three-dimensional points, while manual examination informs the choice of a 30 cm kernel size, later applied to the entire dataset. In summary, this research showcases the potential of UAV-based photogrammetric point cloud analysis to enhance urban safety and infrastructure resilience in sloping terrains, emphasizing the significance of prior knowledge, kernel size selection, and point density for achieving accurate and reliable results. This approach holds promise for safeguarding urban populations and critical infrastructure in similar urban and geological contexts.