Résumé:
License plate detection and recognition systems play a crucial role in various applications such
as traffic surveillance, parking management, and law enforcement. In this paper, we propose
a deep learning-based license plate detection and recognition system. We leverage the power
of the YOLOv5 model for license plate detection, which provides efficient and accurate object
detection capabilities. For Algerian license plate recognition, we employ a CNN model trained
on a large dataset of labeled license plate images.
Through extensive experiments and evaluations, we achieve outstanding results. Our sys-
tem achieves an impressive precision of 87% in license plate detection, accurately identifying
license plates in diverse environmental conditions. Moreover, in the license plate recognition
phase, our CNN model achieves a remarkable accuracy of 93%, successfully recognizing and
extracting characters from the detected license plates.
The combination of YOLOv5 for efficient license plate detection and the CNN model for
accurate recognition results in a robust and effective license plate detection and recognition
system. The system’s high precision in detection and accuracy in recognition make it suitable
for real-world applications requiring reliable license plate analysis.
