Unmanned Aerial Vehicle (UAV) technique has been extensively applied in geohazards assessment. In this study, we illustrate the advantages of the UAV technique in the remote sensing and risk assessment of rockfalls, through a case study of the rockfalls located in Guangzhou, China; in which, the UAV images taken are adopted for identifying the geohazards bearing bodies, based upon deep learning image recognition techniques. On the basis of a visual inspection of the high-resolution aerial photographs, the potential rockfalls are identified. Then, the post-failure runout behavior of the identified rockfalls is analyzed utilizing the numerical software of Particle Flow Code (PFC). From which, the zones that might be impacted by the rockfalls and the destructive forces on the potential geohazards bearing bodies (i.e., buildings) in these zones could be derived. On the basis of the geohazards bearing bodies recognized and the post-failure runout behavior analysis of the rockfalls, a rapid risk assessment of the rockfalls in the studied region is conducted.