<p>Discontinuity features and local geometric characteristics of rock mass are critical in analyzing the evolution and failure mechanisms of potential rockfalls in high and steep hazardous rock masses. Due to the inaccessibility of terrain and the height of these rock masses in the Three Gorges Reservoir area, this paper presents a workflow based on unmanned aerial vehicle (UAV) photogrammetry for the rapid and precise identification of potential rockfalls. This method combines semi-automatic discontinuity recognition technology with kinematic analysis of rock bodies considering local slope directions, allowing for an intuitive and quantitative interpretation of potential rockfall distributions from point cloud data. The results indicate that the Diaozui rock mass in the Three Gorges Reservoir area has four typical structural planes, with J1 and J2 being the most predominant, forming parts of the rock body’s vertical surfaces in different directions. J4 has the lowest proportion but still represents an important set of structural planes, forming vertical faces in two gullies. Moreover, the Diaozui rock body is at risk of failure under all four modes of failure (planar sliding, wedge sliding, flexural toppling, and direct toppling), with planar sliding being the most likely. However, in the overhanging parts of the rock body, the flexural toppling mode occupies a significant proportion. This study proposes a workflow that quantitatively interprets the discontinuities and potential rockfalls in high-steep rock mass, which provides a theoretical basis for the stability monitoring and rockfall mitigation of high-steep hazardous rock mass in the Three Gorges reservoir area.</p>