Experimental investigation of the optimal image acquisition conditions for Unmanned Aerial Vehicles (UAVs) used in dam safety inspections

In order to enhance the effectiveness of Unmanned Aerial Vehicle (UAV) image acquisition for dam safety inspections, this study focuses on three crucial factors: shooting distance, wind speed, and illumination conditions. Using the DJI Phantom 4 Pro UAV as the experimental platform, a series of UAV experiments were conducted under varying shooting distances, wind speeds, and illumination conditions. The image information extraction effectiveness was evaluated through a combination of image adaptive binarization, UAV displacement analysis, and deep learning techniques. The objective was to determine the optimal conditions for acquiring high-quality images during UAV-based dam inspections. The results indicate that the UAV demonstrates excellent hovering capabilities and avoids collision with the target surface under specific conditions, including a shooting distance of 3 meters, a unidirectional wind speed ranging from 2.5 to 4.0 m/s, and favorable sunny backlit conditions. This study optimizes the image acquisition scheme for UAV inspections and provides valuable insights for dam safety inspections utilizing UAV technology. Furthermore, future advancements in automatic inspection systems, sensor technology, and intelligent flight control algorithms can further enhance the comprehensive application of UAVs in dam safety inspections.