Automatic License Plate Recognition

Automatic license plate recognition locates plates in images or video and converts their characters into text for applications such as traffic monitoring and vehicle access control. Common pipelines combine detectors such as YOLO with optical character recognition, while other approaches use character-level models, transformers, tracking, image restoration, or vision-language models. Evaluation spans regional plate datasets and real-world settings with varied scripts, lighting, viewpoints, occlusion, and low resolution. A central challenge is balancing recognition accuracy and robustness with real-time, low-power deployment, while improving performance on unfamiliar conditions and resisting physical-world attacks.