Primary Research On The Vehicle License Plate Location

As a comprehensive real-time computer vision system, the vehicles license plate recognition (LPR) mainly includes the division of license plate localization and recognition of character. Because of the complexity of the scene, the unpredictability of license position and picture quality , the license plate localization division system is still something unsatisfied. therefore the license plate localization division algorithm is always a hot spot of research in this field.By the study of the image texture ,we can draw the conclusion that the density part of texture is the headlight, radiator and the car license. The texture of the headlight and the radiator is almost all horizontal, but the texture of the car license filled with characters is vertical and evener.There are many informations of the car license edge image obtained by the classical edge detect operators method is useless.However the high frequency sub-image which is decomposed by the wavelet is able to describe the car license position better. On the basis of predecessor's study, this article proposes a kind of car plate recognition algorithm, to withdraw the car license vertical texture by the means of wavelet and themathematics morphology. This algorithm is regarded the license plate area as a whole but not several characters while locating the license plate. By this means , we can get the genuine car license from several candidate car licenses regions, considering the car plate definite information (such as width, length and breadth ratio of the plate). Finally, with the method of the vertical projection and the horizontal projection ,we can determines accurate position of the car in the primitive image. By the decomposition of wavelet,the sub-image is the quarter of the original image,and the the texture of the license is more clear than the traditional edge detection. There are certain superiorith about the precision of localization and the speed.The confirmation of the algorithm mentioned in this article, the image and wavelet toolbox that provided by Matlab6.5 are fully used.The experimental simulation results indicates that there is a good car license localization effect.