Integrated Vehicle Detection, Tracking and Sign Recognition for Autonomous Vehicles using YOLOv8

Abstract

Autonomous vehicles require robust perception systems for safe navigation in complex environments. Thispaper presents an integrated approach for vehicle detection, tracking, andtraffic signrecognition usingYOLOv8, astate-of- the-art deep learning-based object detection algorithm. The system ensures real-time, highly accurate vehicle detection, complemented by a Kalman filter-based tracking module for continuous object tracking. A sign recognition module leveraging YOLOv8 enhances traffic rule compliance and decision-making. The system's effectiveness is validated through rigorous testing on benchmark datasets and real-world scenarios. The proposed system integrates deep learning with probabilistic tracking, enabling autonomous vehicles to function seamlessly in diverse conditions. YOLOv8 ensures high computational efficiency, making it suitable for embedded deployment. The Kalman filter enhances tracking accuracy by predicting object movements and mitigating occlusions and motion blur. Traffic sign recognition ensures regulatory adherence, optimizing decision-making. Experiments on benchmark datasets, including COCO and KITTI, demonstrate superior performance in precision, recall, and mean average precision (mAP). The system also exhibits robustness in challenging conditions such as low-light environments and dense traffic. By effectively adapting to real- world complexities, this research advances the development of intelligent and reliable autonomous navigation systems.