Vehicular License Plate Detection Data Fusion and Algorithms for Dynamic Traffic Signal Control

The license plate recognition (LPR) data have become widely available at urban intersections in China. The LPR data is featured with measured headways with time stamps at all approaches of the intersections such that traffic flow rate and operational conditions can be further estimated in replacement of traditional loop data for actuated traffic signal control. This paper presents a study of developing innovative algorithms to dynamically determine the signal control parameters against varying traffic conditions measured by the LPR data sets. The performance of the proposed control algorithm was evaluated via simulation by comparing the LPR-based actuated signal control scheme and the existing pre-timed control, with the use of the data obtained at intersections in City of Jinan, Shandong Province, China. The results indicate an obvious improvement of the traffic control performance and initially prove the developed algorithms adaptive to the LPR-based actuated traffic signal control schemes.