A framework for origin-destination estimation using license plate recognition for Thai rural traffic

Abstract

Origin-Destination (OD) estimation is a crucial parameter for planning city transportation network. Classical methods for collecting the OD information including individual interviewing or using pressurized tube counting equipment; such example methods are practically limited by the number of people involved during observation and interview, traffic delay, budget, and short observation period. Accuracy from human observation and short observation period are doubtfully accurate. Using image processing technology, the OD estimation can be performed for longer period by analyzing vehicle movement from recorded video or online video feeds. Moreover, an individual average speed for each vehicle or type of vehicle are obtainable from identify the license plate number. This paper proposes of a framework for applying a license plate recognition technology for OD estimation for Thai rural traffic. The proposed framework obtains trustable estimation for each individual vehicle. The results can be verified later by human or other methods from recorded video. Therefore, higher accurate and trustable estimation is obtained and make available for later usage.