Automated License Plate Detection and Recognition using YOLOv8 and OCR With Tello Drone Camera
Hanif Fakhrurroja, Dita Pramesti, Abdul Rofi Hidayatullah, Ahda Arif Fashihullisan, Harry Bangkit, Nanang Ismail
Abstract
This paper proposes an Automated License Plate Recognition (ALPR) system using YOLOv8 and Optical Character Recognition (OCR). This system is developed with the Python programming language and uses the OpenCV library for image processing and Pytesseract library as an OCR engine. The camera used to detect vehicle license plates is a DJI Tello Drone camera. The purpose of this research is to implement an electronic ticketing system for vehicles that violate traffic rules. The ALPR system consists of several stages. First, YOLOv8 is used to detect the vehicle license plate in the image captured by the DJI Tello Drone camera. YOLOv8 is an object detection method that is efficient and accurate in recognizing various objects, including vehicle license plates. Once the vehicle license plate is detected, the license plate image will be processed using the OpenCV library to improve image quality and facilitate the character recognition process. Furthermore, the system will use OCR, specifically the Pytesseract library, to recognize the characters on the vehicle number plate. From 50 photo samples, the success rate for license plate detection is 100%, and character recognition is 66%. The proposed ALPR system not only achieves high accuracy rates, but it is also efficient, versatile, cost-effective, and scalable. Because of these benefits, it is a useful contribution to the field of ALPR and traffic management systems, with the potential for future developments and broader applications.