Efficient Vehicle Tracking and Authentication with YOLOv8: Integrating Deep Learning and OCR for Real-Time License Plate Detection

Computer vision plays an important role in this technology-driven world. Using computer vision, the images can be acquired and processed to achieve the high dimensional data of the images. In late 1960s computer vision was start emerging which is considered as the colonist for the latest technology Artificial intelligence. The machine learning and deep learning comes under the artificial intelligence as the subsets. The object detection falls under the category of computer vision. Tracking of each license plate is considered as a most difficult task as the population increases massively day by day. Also, to identify whether a vehicle is authorized or unauthorized in distinct environment is a challenging one. When the population becomes higher and higher leads to many complications such as vehicle theft, misusing vehicles, and unauthorized accessing of vehicles. To overcome these issues many algorithms are handled and the best method is done with the deep learning algorithm. The project aims at designing a system by training the model using deep learning which allows to detect the number plate accurately using open cv module and then the text of number plate is extracted using optical character recognition (OCR) and thus fixed database is compared with the code written in python and thus the vehicle can be detected whether it is authorized or unauthorized. By implementing this project, the complications associated with the detection of vehicles can be overcome. This project aims at designing a most efficient and conceded way for detecting the license plates. For this purpose, the deep learning method is handled which follows the basic algorithm CNN (convolution neural networks) and the model created using the deep learning method follows the YOLOv8, the latest version used for the object detection. In this process the YOLOv8 (you only look once) is handled to make the detection easier since it follows the single neural network to identify the boundary boxes and the class of the trained model. CNN is the considered a for the mathematical calculation as it is the best method for the detection of license plate in real time through openCV camera.