Reconhecimento Automático de Placas de Automóveis no MERCOSUL Usando YOLOv8 e Tesseract OCR

Automatic license plate recognition is a technology that uses camera systems, image processing and character recognition algorithms to identify and register license plates. It has various applications, such as access control in parking lots, automatic tolls, traffic monitoring, security in restricted areas and law enforcement in the case of stolen cars or those with irregular documentation. In this work, the YOLOv8 object detector was used to perform automatic license plate recognition, thus generating a set of images of the license plate object. The image set undergoes a preprocessing composed of two stages: image super-resolution technique and filtering operations with the OpenCV library. Tesseract OCR was used to perform optical character recognition of the image. The automatic license plate recognition system was evaluated in a set of 21 cases, managing to correctly identify the license plate in all cases and recognize the characters in 19 cases.