Comparative Analysis of YOLOv5, YOLOv4, and SSD for Aircraft Detection in Airport Environments Using Remote Sensing Data

Accurate and efficient detection of aircraft at airports is crucial for enhancing airport security, ensuring safe operations, and improving overall air traffic management. As airports grow in complexity and traffic density increases, automated object detection systems powered by advanced algorithms have become essential to support real-time monitoring and decision-making processes. This study presents a comprehensive evaluation of the YOLOv5 object detection algorithm for identifying aircraft at airports, comparing its performance against YOLOv4 and SSD. The goal was to determine the most effective algorithm for real-time detection in complex airport environments. YOLOv5 was trained and tested on a dataset of annotated images of aircraft, and its performance was assessed using precision, recall, mean average precision (mAP), F1 score, and inference speed as key metrics. Results show that YOLOv5 outperforms both YOLOv4 and SSD, achieving the highest precision (0.759), recall (0.772), mAP (0.766), and F1 score (0.765), while maintaining a competitive inference speed of 54 ms. In comparison, SSD demonstrated faster inference speed at 48 ms but lower detection accuracy, while YOLOv4 exhibited the lowest performance across all metrics. These findings indicate that YOLOv5 is the most suitable algorithm for real-time aircraft detection at airports, balancing high detection accuracy with efficient processing speed. This makes it a valuable tool for applications such as airport security and aerial monitoring.