With the advent of advanced computer vision techniques and the support of Deep learning models, it has become very feasible to do traffic analysis. This study explores the creation of analytical and computer vision algorithms for traffic analysis which can precisely detect, track, count and analyze traffic situations. An efficient traffic analysis would ensure safety and proper urban planning. This paper proposes a deep learning based intelligent system which would perform automatic vehicle detection, tracking and counting of three different types of vehicles namely, car, bus and truck. One of the prominent deep learning models - YOLO was selected and employed for this purpose. Competitive analysis of different YOLO models is performed for various parameters, where YOLOv8 emerges out to be the winner, which gives an accuracy more than of $\mathbf{9 7 \%}$ at a decent processing time.
Employing YOLO model for traffic monitoring on roadways
21.11.2024
2547415 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Engineering Index Backfile | 1931
|Probe vehicle based real-time traffic monitoring on urban roadways
Online Contents | 2014
|Automated Traffic Clogging Elusion in Roadways
IEEE | 2018
|Pavement Preservation on High-Traffic-Volume Roadways
Online Contents | 2011
|Traffic Control Help for Low-Volume Roadways
British Library Online Contents | 1997