Real-time detection and recognition of license plate using YOLO11 object detection model

Abstract

This research explores the need for accurate and efficient license plate detection and recognition using advanced computer vision and deep learning techniques. Traditional approaches to license plate identification are often manual, time-consuming, and prone to error, driving the need for automated solutions. This study aims to develop a real-time system for license plate detection and recognition by utilizing the You Only Look Once (YOLO) 11 object detection model in combination with Google Cloud Vision. The project involves gathering a dataset of high-resolution license plate images, annotating these images with bounding boxes, and performing data augmentation to enhance variability and standardize the data. The dataset is divided into training, validation, and testing sets to optimize model performance. The YOLO11 model is trained on this annotated dataset, followed by a detailed performance evaluation. The research also includes a review of existing methods in license plate detection and character recognition, highlighting the role of deep learning techniques in intelligent traffic solutions. Experimental results indicate that the YOLO11 model achieves impressive precision and recall rates, validating its effectiveness for this task. The final model is deployed on a real-time platform, integrating YOLO11 for detecting license plates and Google Cloud Vision for character recognition, thereby demonstrating its practical utility in traffic monitoring systems. This study provides a scalable and efficient framework for license plate recognition, contributing to the development of smart transportation solutions that enhance vehicle tracking and traffic management.