License Plate Recognition using YOLOv9 and EasyOCR

• This repository contains code and resources for a project aimed at recognizing vehicle license plates using the YOLOv9 object detection model and EasyOCR for optical character recognition (OCR). The project involves detecting license plates in images and extracting the text from the detected plates.

• The dataset is downloaded from Kaggle: https://www.kaggle.com/datasets/andrewmvd/car-plate-detection/data. It contains 433 images with bounding box annotations of the license plates within the images.

• The YOLOv9 model has been customized for this project. For example, the number of classes is set to 1, representing the license plate label. Additionally, some changes made to the hyp.scratch-high.yaml file include lr0: 0.001, lrf: 0.001, iou_t: 0.5, and shear: 0.3.

• EasyOCR has been used to detect the text on the license plates based on the bounding box predictions of the YOLOv9 model. The languages used in EasyOCR are English and French.

• The test image are download from https://www.topreg.co.uk/assets/media/44-golf-reg-1090x670.jpg.