Study on a Real-Time Vehicle Speed Measuring Method at Highway Toll Station

Abstract

Real-time vehicle speed detection is a key technology in intelligent transportation system. At present, coil speed measurement, laser speed measurement and radar speed measurement are common methods with high precision but high price. This paper proposes a method of vehicle realtime speed detection at the highway toll station based on binocular vision technology. Firstly, the method is presented to extract the feature points from the time-series images of moving vehicle. Secondly, the correlation of these feature points is calculated in time and space to obtain the homonymous points. Finally, the vehicle speed is calculated according to the homonymous points. Experiments on vehicle speed measurement based on binocular vision were carried out at highway toll stations, and the results showed that when the vehicle speed was less than 30 km/h, the algorithm was accurate in speed measurement and the error percentage was under 10%. This vehicle speed measurement method can satisfy the requirements of automatic vehicle speed measurement at highway toll stations within a lower hardware cost.