Study on Validity of Floating Car Data Based on Vehicle License Plate Recognition

摘要

As the economic development, urbanization and motorization ownership of vehicles are all moving in the fast track, traffic congestion continues to worsen in big cities. Although a great deal of investment in transport infrastructure has been conducted by the government, no obvious alleviation on traffic congestion has been seen. So researchers seek to traffic information to improve traffic management and extend the capacity of transportation system. Real time and accurate traffic information becomes the focus of traffic engineering. As more and more vehicles equipped with GPS, real time traffic data can be achieved by floating car (or prove vehicle) (e.g.(1)). We can estimate link travel speed through floating car data (FCD), but how precise of the estimation we can not know. For the difficulty of acquiring real travel speed data to validate the accuracy, this paper selects expressway as example to acquire link travel speed time through vehicle license plate recognition (VLPR). We can see that as a floating car data system with large number of samples. By sampling method, some conclusions can be achieved such as the running difference between taxi and common cars,the accuracy in different volume and different sample number condition.