In real-time applications such as heavy traffic surveillance and gate pass automation in different mining industries, vehicle number plate recognition is a crucial task due to an increase in traffic, dirt, and bad weather. So manual tracking of all the vehicle number plates at the entry and exit gates in mining industries as well as in different traffic surveillance areas is not possible without a proper number plate detection and recognition model. So, to avoid such scenario, a smart traffic system is needed to recognize different vehicle number plates efficiently. Due to the presence of complexity in Indian heavy vehicle number plates, we have proposed an Automatic Vehicle Number Plate Recognition (AVNPR) model. We have trained and validated our proposed model in a single dataset of 20293 different vehicle images. As the YOLOv8n algorithm has not been tested in real-time, so we have used YOLOv8n for the detection purpose and got a mean Average Precision (mAP) of 98.8%. Further, the recognition is done by using the EasyOCR engine with an average recognition accuracy of 92.32%. Our proposed AVNPR model is found to be better than that of the other existing models in terms of average detection and recognition accuracy.
Automatic Vehicle Number Plate Detection and Recognition System using YOLOv8n and EasyOCR
27.09.2024
985716 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Improved Highway Vehicle Detection Algorithm for YOLOv8n
IEEE | 2023
|Edge Based Intelligent Secured Vehicle Filtering and Tracking System Using YOLO and EasyOCR
Springer Verlag | 2025
|Edge Based Intelligent Secured Vehicle Filtering and Tracking System Using YOLO and EasyOCR
Springer Verlag | 2025
|Automatic Vehicle Number Plate Recognition System (AVNPR) Using OpenCV Python
Springer Verlag | 2021
|