Real-Time Traffic Management Using Deep Learning and Object Detection using YOLOv8

Abstract

As urban traffic management challenges continue to grow, there is an urgent demand for advanced technologies to improve real-time monitoring and control systems. This study introduces an innovative method for realtime traffic management by combining deep learning with YOLOv8-based object detection. By leveraging the speed and accuracy of YOLOv8, our system achieves precise identification of vehicles and pedestrians in dynamic urban settings. The model seamlessly integrates with existing surveillance infrastructure, allowing for intelligent analysis of live video streams to monitor traffic flow and congestion. The system’s adaptability is showcased through training on diverse datasets, and its impact extends to optimizing traffic signal control for enhanced transportation efficiency. Results suggest the potential of our approach to transform urban traffic control systems with improved real-time decision-making capabilities.