Vehicle Tracking and License Plate Recognition using Deep Learning

In this paper, we present a vehicle monitoring system designed to track the entry and exit of vehicles in a factory. This system employs deep learning techniques to monitor vehicles entering and exiting the factory premises and to detect license plates at specific time intervals. The monitoring system utilizes two models: YOLOv8 for tracking vehicle movements through the gate and Optical Character Recognition (OCR) for license plate recognition. The dataset for this system was constructed using proprietary data collected from a factory in Ho Chi Minh City, Vietnam. The implementation was performed on a Tesla T4 GPU (via Google Colab). Experimental results demonstrate that our object detection model achieved a mean Average Precision (mAP) of 0.86 on its test split from our dataset, and the real-time system achieved an accuracy of 97-98% in monitoring vehicles within the factory. These results show that the system is highly effective and reliable for real-time vehicle monitoring and license plate recognition in an industrial setting.