Integrating YOLOv5 and EasyOCR for Robust Thai License Plate Recognition

License Plate Detection and Recognition (LPDR) is a critical component of intelligent transportation systems (ITS), facilitating automated vehicle identification for applications such as traffic monitoring, law enforcement, and access control. In Thailand, the complexity of license plates—characterized by diverse formats, the use of Thai script, and varying environmental conditions—poses significant challenges for LPDR systems. This study proposes an enhanced LPDR framework that leverages deep learning models, specifically You Only Look Once version 5 (YOLOv5) for license plate detection and Easy Optical Character Recognition (EasyOCR) for Thai character recognition. A comprehensive dataset of over 7,000 images was collected and preprocessed using techniques such as image scaling, normalization, and data augmentation. The model achieved 94.3% detection accuracy, 97.0% segmentation accuracy, and 95.5% character recognition accuracy, demonstrating its robustness under real-world conditions. The proposed system outperforms conventional methods by effectively addressing challenges such as low-light conditions, motion blur, and complex backgrounds, making it a viable solution for Thai LPDR applications.