Real-Time Vehicle Detection and Lane Tracking System using Advanced Computer Vision Techniques
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
This project introduces a Real-Time Vehicle Detection and Lane Tracking System aimed at improving traffic monitoring and road safety through advanced computer vision and deep learning. Using the YOLOv8 model, the system detects multiple vehicle classes in live video streams and assigns unique IDs through multi-object tracking. Lane boundaries are extracted using edge detection and region masking, enabling accurate lane classification for each vehicle. The system also identifies lane-crossing events and automatically logs them into a structured CSV file for further analysis. Additionally, the system estimates vehicle speed, tracks movement patterns, and generates activity logs using pixel displacement and frame-based analysis. A real-time Streamlit dashboard visualizes detections, lane overlays, and tracking information to provide a clear overview of road activity. With its modular and efficient design, the solution supports applications in smart traffic management, violation detection, and intelligent transportation systems, offering a scalable and accurate approach for real-time roadway surveillance.
Fields of Study & Subjects
Autonomous Vehicle Technology and Safety
Advanced Neural Network Applications
Video Surveillance and Tracking Methods
Publication details
- Journal / source
- International Journal for Research in Applied Science and Engineering Technology
- Volume
- 14
- Issue
- 3
- Pages
- 3835–3837
- Publication date
- 28 March 2026
- Language
- EN