Real-time UAV Localization and Tracking in Multi-Weather Conditions using Multispectral Image Analysis*

With the increasing availability of unmanned aerial vehicles (UAV), their potential misuse has become a serious concern, posing a threat to public security. Existing tracking methods have limitations in detecting UAV effectively due to their small size, high speed, complex flight patterns, and complicated flying background. To address this problem, we propose a UAV localization and tracking method that uses multispectral images captured by specific hardware, which enhances the detection process by allowing for greater visibility in challenging weather conditions. The proposed method combines the YOLOv5 detection algorithm with the KCF tracking algorithm to provide a reliable solution for preventing potential misuse of UAV and enhancing public safety.Experimental results demonstrate that the proposed method provides a reliable solution for UAV localization and tracking in various weather conditions. The method was found to improve the inference speed compared to a single YOLOv5 model, demonstrating its potential for real-time UAV tracking and control. By combining multispectral image analysis, detection algorithms, and tracking algorithms, the proposed method provides an effective solution for preventing the potential misuse of UAV and enhancing public safety. This research presents a promising direction for future studies on UAV tracking and control.