Empirical experiment on origin destination estimation using automatic licence plate matching on a rural road of Thailand

Abstract

The growth of economic influences city authorities to plan and build infrastructure. The growth prediction mostly uses modern mathematical models. The accuracy of the model is highly depended on reliable parameters. Two crucial parameters for most models are traffic capacity and average travel speed along road networks and obtained from an origin-destination (OD) estimation. Thus, an OD estimation is necessary to obtain reliable parameters for the predictive models. Traditionally, the OD estimation is performed by human observation; however, an automatic license plate recognition and matching algorithm show potential of obtaining such data. Additionally, the observation is able to perform long term observation which is impractical for human. Therefore, this paper reports an empirical experiment on OD estimation using automatic license plate matching along with a real-world human observation survey in Mukdahan province. The results from both types of surveys are reported and compared. The different results between the both raises an opportunity for researching an optimal method of OD estimation in Thailand.