Abstract: In recent years, unmanned aerial vehicles (UAVs) have been increasingly deployed for reservoir area inspections.However, the operations still rely on traditional manual flight modes, which lack automatic inspection.Both manual and UAV inspections face challenges in achieving high-frequency crack identification.Current research on UAV-enabled automated high-frequency crack detection remains limited.In this study, we employed an automated UAV inspection system, utilizing a highly integrated UAV hangar and networked control platform to execute remote, autonomous inspection missions.We developed a high-frequency crack identification methodology for slopes in the Lianghekou Hydropower Station reservoir area.Equipment selection and site deployment were analyzed based on mobility, recharging methods, UAV types, payload capacity, and inspection mission requirements.An automatic data processing workflow centered on the UAV hangar was established.Furthermore, a multi-dimensional interpretation and AI-driven algorithm for surface crack identification approach was proposed.Validation was performed using nearly three years of satellite data and inspection-monitoring results.The results indicated that the UAV automatic inspection technology identified 392 cracks in the study area, with a manual field-verified accuracy rate of 94.8%.Satellite and ground monitoring data corroborated the method′s accuracy and feasibility.The approach demonstrates minimal susceptibility to environmental, spatiotemporal factors, offering a novel approach for reservoir area inspections.