Study On Recognition Method Of Road Vehicles

Along with the development of social economy, the management of traffic system becomes more and more complex. In ITS, the system of vehicle detection and identification which provides necessary data sources, is the core technology of realizing non-stopping toll in highway. So the research of vehicle identification techniques not only has important academic value, but also could probably increase social and economic benefit greatly.A method of vehicle identification based on the loop electromagnetic induction is presented. This method carries on the data acquisition of vehicle's electromagnetic induction characteristics by the loop induction detector of vehicle. The message of vehicle types is received by counting the frequency of the oscillator, and then the data is transmitted to the host computer. After the data were gathered, the data will be recognized systemically and pretreated firstly, such as filtering,noise reduction. Then a method based on the wave separation and waveform coding is presented to vehicle feature extraction, and a pattern database of vehicle types features'code is established, which can quickly carry on the recognition and classification automatically by minimum distance classifier to new vehicle sample. This system has on-line modified and pattern base extension function through manual intervention.The algorithm of fast recognition the road vehicles and its case of algorithm application have proposed, and the implementation process of algorithm is described in detail. Classifier has a simple structure. The probability of this case has been verified through practical examples, and a higher identification rate is gotten.