An Intelligent Architecture for Automatic Number Plate Detection Using OCR and YoloV8 Model

Abstract

Automatic number plate recognition (ANPR) equipment in security, protection, and commercial has improved over the previous few years. NPR using computer vision (CV) is utilized to give rapid and precise recognition and classification. Several electronic methods have been implemented to classify vehicle process particulars based on vehicle license NP images using Yolov8 object detection methodologies. In the research architecture it utilized OCR and Yolov8 detection model for ANPRS. A yolov8 model trains the model to detect the vehicle number plate.  The vehicle number plate registration is cropped, resized, and ROI extracted from the image; a yolov8 model uses OCR to verify the number plate and letters. They used the Kaggle ANPRS database and attained a maximum accuracy of 98.1 %. The implemented model could detect ANPRS on real-world car number plate images. The proposed model can be used for safety regions, army fields, and government agencies.