Earth-rock dam leakage inspection and identification based on an unmanned aerial vehicle equipped with infrared thermal imagers

The manual detection of leakage in dams can be time consuming and labour intensive. A timely detection of leakage hazards is key to ensuring the structural safety of dams. This paper uses an unmanned aerial vehicle (UAV) equipped with infrared thermal imagers, proposes a dual-loop proportional-integral-derivative (PID) intelligent control inspection method based on the boustrophedon method and develops a thermal infrared remote sensing technology suitable for dam leakage inspection. The approach is designed to overcome the existing challenges and address the complexity and safety risks associated with conventional equipment deployment. Combined with the adaptive-area-growth threshold-based segmentation technology, an intelligent leakage-area-growth identification method based on thermal infrared remote sensing images was constructed. The feasibility of this method was verified by inspecting and identifying leakages in a reservoir located in Heihe City, China, as an experimental site for actual engineering. The results showed that the accuracy of the temperature measured by thermal infrared remote sensing was negatively correlated with the detection distance and linearly and positively correlated with the reflectance. The influence of the relative humidity on the temperature measurement accuracy was relatively small. The proposed method can automatically and accurately identify the leakage area, significantly improve the efficiency and adaptability of leakage detection and provide a feasible technical method for dam leakage inspection and identification problems.