A New Method of Origin-Destination Survey Using License Plate Recognition by Recursive Traffic Zone Division

Abstract

This paper proposes a new method of using License Plate Recognition (LPR) for Origin-Destination (OD) survey. LPR has shown many advantages in OD estimations such as lower average acquisition costs than other OD surveys, more information extracted from vehicles passing by LPR devices etc. We recommend a recursive method of traffic zone division as to flexibly divide traffic zones by network conditions or practical needs. In order to improve the speed of license plate matching process, we adopt an index structure of LPR database by the coding of license plates. Two important special OD cases, having-O-and-no-D and no-O-and-having-D are handled to supplement OD information by expanding the searching space-time scopes in database. The methodology is perfectly validated by an example of OD survey for a road network in CBD of Beijing.