UAV DETECTION AND TRACKING IN CHALLENGING AERIAL ENVIRONMENTS USING YOLOV11

Abstract

Security, surveillance, and public safety concerns have been greatly increased by the fast growth of Unmanned Aerial Vehicle (UAV) utilization in both commercial and civilian fields. This study assesses the most recent developments in visual tracking and object detection based on deep learning to address the increasing demand for precise and efficient UAV detection and tracking systems. In particular, we examine the performance of state of the art You Only Look Once (YOLO) object detectors—YOLOv11-L, YOLOv11-M, YOLOv10-M, and YOLOv9-M—on the Anti-UAV-300 dataset, a demanding benchmark that includes RGB and infrared drone footage captured under a variety of environmental conditions. All models were fine-tuned and evaluated across standard object detection metrics, including precision, recall, mean Average Precision (mAP), and inference speed, using a consistent training pipeline with tailored augmentation strategies. While YOLOv11-M provided an optimal balance between accuracy and real-time efficiency, YOLOv11-L demonstrated the maximum performance, with a [email protected]:0.95 of 0.810 and nearperfect precision and recall. Furthermore, the Center Location Error (CLE), Intersection over Union (IoU), and tracking precision were evaluated for two baseline single-object tracking algorithms: SiamRPN and CSRT. The results indicated that both trackers were substantially below the thresholds necessary for reliable UAV tracking in practical applications. The findings emphasize the limitations of conventional tracking algorithms in dynamic aerial surveillance scenarios and demonstrate the superiority of recent YOLO architectures for UAV detection. This research emphasizes the significance of detectiontracking pipelines that are learning-based and integrated, and it establishes a foundation for future research on real-time, robust UAV threat monitoring systems.

Date of Award	2025
Original language	American English
Awarding Institution	HBKU College of Science and Engineering