Western China is rich in oil and gas resources, and many oil and gas pipelines are under construction or have been completed. However, many water-related natural hazards, such as landslides, collapses, rockfalls, and debris flows, have developed in the areas passed through by oil and gas pipelines and seriously threaten the operational safety of these pipelines. Therefore, it is urgent to carry out large-scale identification and assessment of pipeline geological hazards. At present, conventional on-site investigation, evaluation, monitoring, and early warning methods are difficult to apply for rapid identification and evaluation of pipeline geological hazards across large-scale areas. Based on this, this study takes the pipeline of Sinopec Marketing South China Branch in Yunnan Province as the research area. In this research, unmanned aerial vehicle (UAV) and photogrammetry technology were used to quickly and accurately obtain multi-phase images of an oil pipeline passing through the study area, and the images were post-processed to obtain multi-phase high-resolution, high-precision digital orthophoto maps and digital terrain models (DTMs) to identify landform changes and deformation. The focus of this research is to propose a set of technical methods for UAV point cloud filtering. The DTMs obtained based on this method can effectively identify unstable areas of oil pipelines. In addition, we have carried out numerical simulations under different motion scenarios in unstable regions, providing scientific support for future geological hazard prevention and mitigation and engineering practices in oil pipeline areas.