Light-Enhanced Deep Learning Approach for Aircraft Visual Inspection Using Cooperative UAVs

Aircraft inspection is a critical aspect of ensuring aviation safety. However, traditional inspection methods are time-consuming, labor-intensive, and lack accuracy, especially under poor lighting conditions. In this paper, we propose a novel light-enhanced deep learning approach for automating aircraft visual inspection using two drones. One drone (camera drone) captures images and footage perpendicular to the inspection surface, while the other drone (lighting drone) provides illumination from a variable inclined angle to enhance visual inspection and crack detection. We developed advanced deep learning techniques to detect surface anomalies on aircraft effectively. Experimental results on a large-scale dataset demonstrate improved performance compared to existing methods, achieving a high accuracy rate. The proposed method has the potential to significantly improve inspection efficiency, reduce costs, save time, and ultimately enhance aviation safety.