The Moving Vehicle License Image Restoration And The Vehicle License Recognition System Design

With the development of economy, the rapidly increasing number of vehicles, the traffic problems have become more and more serious.It makes the intelligent transportation system become a hot research field and get wide attention as well. LPR(license plate recognition) is one of the important task in the filed of intelligent transportation system of computer vision and pattern recognition technology. It is also one of the important part in intelligent transportation system.The theme of this paper is to identify the motive vehicle license plate in intelligent traffic management system. It has designed a physical construction and logical structure in the restoration of the image of moving license plate and identification system which are consist of all basic functions. Meanwhile, it has carried out a deep research in the aspect of these key technologies: the restoration of image of moving license plate and recognition.In the process of the restoration of image of moving license plate, it has posed a moving license plate restoration method based on Kalman movement model, which is aimed at dealing with the situation that mainly on the current monitoring system with overspeeding, and caused the ambiguous image. This method has established Kalman movement model, and analyze the process of the motion-blurred image degradation, based on the features of the speeding cars. This method approaches the actual car model. Practice proves that the effect of restoration is obvious. Meanwhile, it is laid on a well foundation for positioning and recognition.In the positioning and identification of the plates, the combination of the connected site search and the projection examine of the optional site of license plate has been used. Then, the method of BP neural network for recognition of license plate has been used in it as well.As the experiment shows, the method this thesis adopted can accuratly position the segmentation licence plate and give recognition to the exceeding vehicles.The algorithm of this paper designed will provide the value of the reference to some extent to the design of the practical lience plates recognition system.In this thesis, the related experiments and applications has been done by using C + + and MATLAB programming language. Meanwhile, the related function in OPENCV (Intel (R) Open Source Computer Vision Library) which is subsidized by Intel Corporation has been used for recognition and dealing with motive target image.