Automated License Plate and Over-Speed Detection System
• Designed and developed an automated system that detects vehicle speed and, if over-speeding occurs, extracts & stores the license plate number of the vehicle to generate fine.
• Employed YOLOv5 for object detection, WPOD-NET & OpenCV for object localization and character segmentation, and Django & MySQL for developing a responsive User interface.
• Trained MobileNet, Xception, and ResNet50 models for character recognition, observing the highest accuracy of 90% (MobileNet); tested on manual dataset achieving an absolute error rate of 0.98 km/hr
@inproceedings{singh2023adaptive,
title={An adaptive license plate recognition framework for moving vehicles},
author={Singh, Guruanjan and Mehta, Krutik and Mishra, Apoorva and Chawla, Harnish and Shekokar, Narendra},
booktitle={AIP Conference Proceedings},
volume={2794},
number={1},
year={2023},
organization={AIP Publishing}
}