Estimation of Origin-Destination Trip Patterns on Expressway using Automatic License Plate Recognition Data

Abstract

Origin-Destination (OD) matrix is one of the key transportation data used in the traffic analysis and planning. Traditional methods for collecting the OD information including individual interviews are limited by the number of people involved during the interview effort. Traffic data collection equipment such as loop detectors and microwave sensors are not able to identify individual vehicles. In addition, GPS devices can be used to identify individual vehicles, but it is required to be installed in every vehicle traversing the road network. With the advancement of computer vision technology, it is possible to use the automatic license plate recognition (ALPR) system in reidentifying the same vehicles traveling a particular pair of origin and destination. In this paper, we explore the use of automatic license plate recognition in estimating the expressway network’s OD table. We proposed a matching technique for license plate data matching at the on and off ramps within the expressway network. The performance of the matching technique is evaluated using detection rate, ALPR accuracy, and matching accuracy. In sum, the proposed license plate matching strategy can improve the accuracy of the OD estimation.