Research on the Detection Method of Precision Approach Path Indicator Lights Based on UAV Perspective

To address the position offset issue for small targets caused by the jitter effect during the flight validation of PAPI light angles using UAV, a precise real-time detection method based on YOLOv5 is proposed. First, by isolating the relative positions of the auxiliary lighting and the PAPI lights, background images without targets are extracted and added to the training dataset to improve the model's ability to distinguish between targets and background. Second, utilizing the image grayscale histogram, Gaussian filtering techniques are applied for data augmentation to improve the model's robustness. Finally, a small object detection layer is incorporated into the YOLOv5s model to enable accurate real-time detection of PAPI lights. The interference caused by auxiliary navigation lights to the detection of PAPI lights was effectively addressed by this methodology, enabling accurate and real-time detection of PAPI lights during flight verification processes.