Automatic Number Plate and Speed Detection using YOLO and CNN

This research paper showcases a leading-edge Automatic Vehicle Number Plate Recognition (ANPR) system that capitalizes the potential of YOLOv8 (You Only Look Once) and Convolutional Neural Networks (CNN). The central objective of this system is to streamline the precise extraction of license plate information, with a prominent emphasis on precision, automation, and versatility. Moreover, it offers a rugged solution for thorough traffic monitoring and enforcement by integrating vehicle speed calculations. The combination of YOLOv8 and CNN significantly augments the image-processing capabilities of the system. To implement this ANPR system successfully, a series of critical steps have been followed, consisting of system setup, data collection, YOLOv8-based license plate detection with the help of CNN for licence plate character recognition, speed calculation through OpenCV, integration of components, rigorous testing, and finally, deployment. The fusion of these advanced technologies and methodologies, as elucidated in this paper, pledges to overhaul license plate recognition in the context of automated vehicle surveillance and traffic management. With an accuracy of 98.5% in the detection of number plates accompanied by an accuracy of 96% in speed recognition along with automation, and adaptability, this ANPR system holds tremendous potential for improving law enforcement, security, and transportation efficiency.