Intelligent Transportation System(ITS) has gained much attention recently. As an important issue with applications of ITS, the vehicle identification system has been one of the hotpots.A vehicle identification system applied to the vehicle surveillance is studied and implemented in this thesis. The process of this system includes four phases: the prospects for segmentation, target detection, feature extraction and classification models. Firstly the background difference method is used to extract vehicle goals in a complex background, and then by using the detection of region connectivity and the morphological image processing on the differential results the candidate region is identified, and further to identify the face image. Secondly, by applying the eigenface method and the method based on the texture feature, the car's faces can be extracted. Finally, the minimum distance method has been used for vehicle classification and identification. Experimental results show that the proposed vehicle identification method is simple, fast and effective. |