AI-BASED TRAFFIC DENSITY ESTIMATION AND CONGESTION PREDICTION USING REAL-TIME VIDEO ANALYTICS AND DEEP LEARNING

Authors

  • MD. SAIRA BANU Author

DOI:

https://doi.org/10.62643/ijerst.2026.v22.n3.4845

Abstract

The rapid growth in urban vehicle populations has resulted in increased traffic congestion, longer travel times, higher fuel consumption, and elevated environmental pollution. Conventional traffic monitoring systems often rely on manual observation, fixed sensors, or historical traffic information, which may not adequately capture dynamic traffic conditions or provide timely congestion forecasts. To address these limitations, this paper proposes an artificial intelligence (AI)-based framework for traffic density estimation and congestion prediction using video analytics. The proposed system processes realtime or recorded traffic video streams to detect and classify vehicles, estimate traffic density, and identify congestion conditions. A deep learning-based object detection model, such as You Only Look Once (YOLO), can be employed to identify vehicles including cars, buses, trucks, and motorcycles. Computer vision techniques are subsequently used to count detected vehicles, analyze their spatial distribution, and estimate traffic density within a predefined region of interest. Based on extracted traffic features, including vehicle count, occupancy, and temporal variations in traffic flow, a predictive model can estimate short-term congestion levels and classify road conditions into categories such as low, moderate, and high congestion. The proposed framework aims to improve traffic visibility, facilitate early identification of congestion, and support intelligent transportation system applications. Its effectiveness can be evaluated using vehicle detection precision, recall, mean average precision, density estimation error, and congestion prediction metrics. The framework provides a scalable foundation for integrating AI-powered video analytics into smart city traffic management systems.

Downloads

Published

23-09-2026

How to Cite

AI-BASED TRAFFIC DENSITY ESTIMATION AND CONGESTION PREDICTION USING REAL-TIME VIDEO ANALYTICS AND DEEP LEARNING. (2026). International Journal of Engineering Research and Science & Technology, 22(3), 2245-2258. https://doi.org/10.62643/ijerst.2026.v22.n3.4845