The Research Of Vehicle Model Recognition Based On Wavelet Transform And Moment Features

In the Intelligent Transportation Systems, to count and get the information of automobile is the pivotal guidance for the foundation of the vehicle auto--inspect and road auto--tolling system. And the recognition technology of vehicle model is important for the automatization of road traffic, the further research of which can enhance the automatization degree of road traffic system and has the gravely actual meaning for promoting the development of Intelligent Transportation Systems. The intelligent recognition of vehicle model is an important part of Intelligent Transportation Systems. In the Intelligent Transportation Systems, the recognition of vehicle model plays a very important role in the vehicle auto--inspect and road auto--tolling system, and is also one of the hotspots and difficulties in the Intelligent Transportation Systems areas.The present methods about the recognition of vehicle model are reviewed in this dissertation, and also the extraction and recognition algorithm of vehicle model features based on the wavelet analysis and moment character are proposed. There are many kinds of vehicle model at present, and the classification of vehicle model is mostly based on the external features, so the extraction and choice of the representative external features can ensure the correct classification. The wavelet analysis, which is an analysis method of multi-scale, has the good ability to describe time-frequency information, and can reflect the whole approach information and local detail information of the signal in each scale. The invariable moment, which is invariable to the translation, scaling and rotation of image, is widely used in object recognition, image classification, image condensation, scene matching and so on. This paper studies on two vehicle model recognition methods——based on wavelet moment and on linear moment. The wavelet analysis and object moment feature have the respective virtues. The wavelet moment is formed by joining the both ones for object recognition, which is invariable to the translation, scaling and rotation as the moment, and also can grasp the detail features of images to improve...