Real time drone surveillance system –for detection, distance estimation and automatic target lock using YOLO

This task presents a practical way to monitor drones and protect aircraft. The system detects and tracks drones in real-time. It uses computer vision and deep learning, employing a Yolov8 model to identify drones reliably. This works well even in poor lighting, background noise, or partial visibility. Besides detection, the system estimates the drone's distance from the monitoring station and checks it against a safety limit. If the limit is exceeded, the system quickly issues a warning and locks onto the target. This design prioritizes efficiency and works well on devices with limited computing power, making it suitable for laptops and larger applications. Possible uses include civil surveillance, industrial security, and defence operations. The proposed YOLOv8 model achieved MAP@0.5 of 0.94, precision of 0.91 and recall as 0.90, with the average inference speed as 3.8 ms per frame, confirming it is suitable for real time operations. Future upgrades may include support for multi-camera setups, swarm identification, and trajectory prediction, enhancing its role in protecting restricted airspace.