A Novel Approach to Identify and Interpret License Plates with YOLOv5 and Tesseract

Modern intelligent transportation systems rely on License Plate Recognition (LPR) to automate tasks such as monitoring traffic, checking toll fees and handling law enforcement. This paper presents using YOLOv5 for object detection and Tesseract OCR for character recognition. The purpose is to accurately capture and interpret license plates when conditions are tough, like reduced lighting, blocked views, slanted angles and cluttered surroundings. The dataset uses publicly accessible images and labeled data which makes the dataset applicable for many cases. A well define sequence consisting of preprocessing images, locating objects and extracting text using OCR technology is established. The systems are assessed with the help of precision, recall, accuracy and F1-score metrics. The results of our approach are more successful than Haar cascades, SSD+CRNN and other common OCR technologies. A confusion matrix proves model accuracy, but a graphical bar chart gives a more transparent view of model performance. Our findings indicate the model is more accurate and stable, making it a good fit for law enforcement and traffic management systems.