Authors:
J. Angelin Jeba, S. Rubin Bose, R. Regin, P. Banupriya, Andino Maseleno

Addresses:
Department of Electronics and Communication Engineering, S.A. Engineering College, Chennai, Tamil Nadu, India.  School of Computer Science and Engineering, SRM Institute of Science and Technology, Ramapuram, Chennai, Tamil Nadu, India. Department of Science and Humanities, Dhaanish Ahmed College of Engineering, Chennai, Tamil Nadu, India. Department of Information Systems, Bakti Nusantara Institute, Lampung, Indonesia.

Abstract:

Urban traffic monitoring is essential for effective traffic management and public safety. However, traditional traffic monitoring systems face significant challenges, including high implementation costs, computational complexity, and an inability to provide real-time analysis, leading to delayed responses to traffic incidents and inefficient resource allocation. This paper presents a simplified real-time traffic monitoring system that leverages pre-trained YOLOv8 for vehicle detection and a basic web dashboard for visualisation. The problem statement addresses the critical need for affordable, efficient, and real-time traffic monitoring solutions that can operate with minimal computational resources while maintaining high accuracy. The methodology involves processing live video streams from webcams or traffic cameras through a YOLOv8-based vehicle detection pipeline that classifies vehicles into four categories (cars, trucks, buses, and motorcycles) with high accuracy. The system then processes the detection results to generate real-time traffic analytics, including vehicle counts, classification breakdown, and performance metrics, which are displayed through an intuitive web interface. Experimental results demonstrate the system's ability to detect vehicles with high accuracy (90% mAP) and low latency (25-30 FPS), making it suitable for deployment in resource-constrained environments. The implementation focuses on simplicity and practicality, using minimal computational resources while providing valuable traffic insights for traffic management centres and urban planning applications.

Keywords: Vehicle Detection; Traffic Visualisation; Vehicle Counting; Real-Time Traffic Analysis; Surveillance Systems; Web Dashboard Analytics; Deep Convolution Neural Network.

Received on: 03/01/2025, Revised on: 29/03/2025, Accepted on: 24/05/2025, Published on: 09/12/2025

DOI: 10.69888/FTSCS.2025.000526

FMDB Transactions on Sustainable Computing Systems, 2025 Vol. 3 No. 4, Pages: 242-252

• Views : 54
• Downloads : 9

Download PDF