This project aims to develop a lightweight YOLO-powered digital twin framework for real-time air traffic monitoring using airport surveillance cameras. The research investigates how computer vision techniques can be integrated into a digital twin capable of representing aircraft locations, movement trajectories and traffic flow in near real time.
The central research question is how a lightweight object detection and tracking framework can achieve reliable aircraft monitoring while balancing detection accuracy, tracking reliability, inference latency and computational cost. The project will also investigate the feasibility of adapting lightweight computer vision models to airport environments, where aircraft vary significantly in size, appearance and operating conditions.
The expected outcome is a practical and reproducible prototype together with an empirical evaluation of its performance under different deployment conditions, providing design recommendations for efficient AI-enabled digital twins for intelligent air traffic monitoring.