Massive and detailed data collected by the license plate recognition (LPR) systems installed at intersections can offer opportunities and support for a wide range of research directions in the transportation field. This paper summarizes the literatures on LPR-based research and corresponding methods from 2006 to the present, demonstrating the significant role of LPR data across multiple transportation research domains. The study surveys the literature and highlights how LPR data serves as a foundation for estimating and predicting traffic state, analyzing travel demand, optimizing traffic control, and enhancing traffic safety management, owing to its unique data advantages. By reviewing data modelling methods and models employed across different research directions, this study proposes new opportunities for future innovations in applications of LPR data. Specifically, studies on LPR-based traffic control strategies, proactive traffic management, intelligent vehicle infrastructure cooperative systems, and other related areas remains relatively underdeveloped. The application of more advanced techniques, such as reinforcement learning, hybrid learning, and large models (i.e. large language models and pre-trained foundation models), as well as the development of LPR-oriented simulation environment, are needed for further breakthroughs in LPR data modelling.
Ng et al. (Fri,) studied this question.