A novel 3D geomorphic change detection framework for large-scale active dump slope using multi-temporal imagery

Slope stability of dumps in the mining industry is a critical issue due to frequent incidents involving loss of life, equipment damage, and operational disruptions. This study presents an innovative methodology integrating unmanned aerial vehicle (UAV)-based 3D photogrammetric reconstruction, point cloud analysis, change detection, and numerical modeling to assess slope stability. High-resolution data were acquired using a DJI drone. Geometric features were analyzed at multiple scales using advanced point cloud techniques. A key advancement was the implementation of a Hybrid registration approach that yielded the lowest RMS error (0.642), which ensures precise spatial alignment. Change detection analysis identified a maximum displacement of 31.13 m in the active dumping zone. Numerical modeling of the critical section with a Factor of Safety (FOS) of 1.107, confirming near-instability conditions. This multi-modal approach not only improves assessment precision but also promotes proactive slope management, offering broad applicability across various geotechnical contexts.