Research Of Vehicle Identification Based On DSP And AR Parameter Model

Intelligent transportation system ITS (intelligent Transportation System) is the world'domain research topic of transportation. the research of ITS based on the exi- sting transportation condition fully carries on the reasonable transportation assignm- ent and the management using the modern high-tech.At the meanwhile, through the satellite navigational system, the automobile automatic direction system,the tra- nsportation information correspondence and the computer .management and so on. The entire road network traveling capability can be rapidly enhanced.Therefor, the realization of a safe, fast and the convenient transportation program become real.Computation of vehicles information , system of automatic monitoring and entire automatic charge system are the key for intelligence transportation system.At the meanwhile ,vehicle type recognition is regarded as the precondition of the road automation,therefore the development of vehicle type recognition system price which is inexpensive and the suitable for our national condition has fairy important practical significance.The present paper based on gathering of sound signal and the seismic motion signal production and the foundation of scene gathering tentative data and the method of the parameter model analysis and the platform of the DSP technology deeply research vehicle type recognition technology based on vehicles sound signal and seismic motion signal.This article first introduced the pattern recognition and the development of current vehicle type recognition technology, then it has summarized the theory of spectrum estimated AR model analysis, because this algorithmic had been tested on MATLAB through the simulation confirmation, therefore we only introduce the use of Yule-Walker equation. In the realization process of the AR parameter model, based on TI Corporation's TMS320C5402 platform, we bring forward a proposal and discuss the issues and solutions we encountered.Finally,selecting the same step, we analyze situations of the same speed and different speed in the analysis of the automobile signaling characteristic, the result finally indicates that the same speed vehicles and vehicles with different speed can be recognized and also do the different sizes.At the meanwhile, the withdraw of respective vehicle type (or vehicle speed) is perfect,it showed that parameter model method is suitable for distinguishing automobile the vehicle type though the automobile tonic train signaling in the highway. Therefore,analyse the automobile vehicle type with the parameter model method is a feasible and effective new way.