License Plate Recognition System Based On Digital Image Processing Technology Research

With the continuous development of social economy and science and technology, information technology, communication technology, sensing technology and control technology in one of the intelligent transportation system(Traffic System Intelligent) has been more and more mature. As an important part of the vehicle license plate recognition technology has become the current research hot spot, many experts and scholars both at home and abroad have carried out in-depth research on it. The license plate recognition based on image processing and pattern recognition technology, is widely used in the field of parking lot management system, city crossroads e-police and save a lot of manpower, high-speed toll management system and material resources, the development of the semantic security of the state's major.In this paper, the four key steps of license plate recognition system, digital image preprocessing, license plate location, license plate character segmentation and character recognition, are studied. A complete license plate recognition system from image acquisition to image recognition is built, and the process and display of the whole license plate recognition is completed by using GUI MATLAB interface. In the aspect of image preprocessing, this paper firstly transforms the color image into black and white image, and then it is processed in the aspects of drawing, uniformity, median filtering, etc., and carries out the two value. In the aspect of license plate location is studied in this paper according to the previous method based on the license plate location color feature and texture feature, combining the advantages of the two methods, to improve the accuracy of positioning, and reduce the time required for the position. In character segmentation, it is still on the basis of the previous algorithms, which add the standard normalization, the tilt geometry correction and the removal of the frame, and so on. In the license plate character recognition, in depth study of the method of character recognition based on several commonly used, I introduces the character recognition method based on BP neural network is used, and randomly selected 60 pieces of photos for license plate recognition, the correct recognition rate reaches 90%.