Automated Vehicle Detection and Real-time Number Plate Recognition
Loading...
Date
2024-07-04
Journal Title
Journal ISSN
Volume Title
Publisher
University of Msila
Abstract
EN
In this project, we focused our work on the Algerian market and developed a system for automatic license plate recognition (ANPR) in real time. The system was trained using deep learning algorithms on a dataset of Algerian license plates with different photographing angles.
It integrated neural processing units (NPU) to improve image processing performance and recognition speed. We compared the performance of different AI models, including Faster RCNN and YOLO. It showed that YOLOv5 model achieved a high detection accuracy of 99% in real time.
This embedded system can be used in various areas of the Algerian market, such as parking control, vehicle tracking and access point control.
Description
Keywords
Automatic Number Plate Recognition, Neural Processing Unit, Real-Time, Algerian License Plates, ANPR, LPR, ALPR, NPU, Faster RCNN, YOLO.