Intelligent Transportation System (ITS), is the development trend ofthe world traffic in the21century. In ITS, vehicle recognition emerges asthe fast development of the traffic. Vehicle recognition system not onlyfinds wide application in ITS, but also it is a hot point of research incomputer vision, image processing and pattern recognition.In this paper, the solution of vehicles recognition includes threeprinciple parts: object detection, feature extraction and objectclassification:Firstly, vehicles are detected as they move through the field of view,and the stopping information will be given.Secondly, features such as size, width, length, contour of vehicles, etc.can be extracted from the images at the stopping time of vehicles.The recognition schemes, SVM-based statistical recognition andPCA-based methods, and our improved solution that integrated featureextraction and PCA is developed. Comparison has been made betweenthem, and the integrated solution gets higher recognition rate. |