Enhancing UAV Detection and Tracking Range Utilizing KNN and Canny Edge Detection

Object detection and tracking play an important role in autonomous vehicle systems. In the past few years, object detection using deep learning from images has become a center of attention. This paper presents a novel approach for the detection and tracking of fixed-wing Unmanned Aerial Vehicles (UAVs) using the YOLO version 8 deep learning algorithm in BGR color space vs grey color space using Canny edge detection and KNN to see the outline. Leveraging the state-of-the-art capabilities of YOLO v8, real-time and accurate detection of UAVs from ground perspective is achieved. In this study, object detection and tracking application is performed by training the YOLO v8 algorithm, annotation tool used was CVAT, with the help of Python software language with a dataset consisting of photographs of fixed wing UAV taken from the ground. This research contributes to advancing the field of UAV detection and tracking, particularly focusing on the capabilities and advancements of the YOLO v8 algorithm in real-time object detection and tracking applications.