The Research On License Plate Recognition System Based On Artificial Neural Network

As one of the focal points and hot problems of studying in the field of modern traffic engineering, the license plate recognition(LPR) has attracted more and more attention.Through the research of four key links of the license plate location, image preprocess, character segmentation, character recognition on the license plate recognition system, the paper has designed an intact license plate recognition system and has carried on simulation in the environment of MATLAB.In the location part of license plate, the paper has adopted the license plate location method which is based on color character and vein character. At first, colored picture is changed from RGB space into HIS space. The paper has realized the thick location of the license plate by using the character of blue colour degree is about and S value of saturation degree is bigger in the blue bottom and white word license plate. Then, after making thick location, the image is measured marginally by using Canny operator. Finally, according to the character of license plate image's high-frequency of changing black-and-whitely, the paper has realized the accurate location of the license plate. 240°In the part of image preprocess, the paper has handled the license plate location image with greyscale, furthermore, used Otsu method to change greyscale image into two-value image. In fact, as license plate frame and up-low rivet will interfere with the recognition work of license plate, this part has removed the license plate frame and up-low rivet.In the part of character segmentation, according to the current license plate design principle, paper has segmented the license plate image into 7 characters waited to recognise by using the improved horizontal projection, and normalized the character after segmenting. Practice has proved this method is effective on solving the problems which are Chinese character unconnected, cement of the character, interference of the noise and the dot existing between the first 2 characters and the following 5 characters.In the part of character recognition, aiming at four kinds of different recognition problems of Chinese character, the letter, letter or figure, figure, the paper has designed four kinds of different classification devices by using the improved BP neural network. Utilizing 13 features extraction, the paper regards its result as the input of the network. Finally, the license plate number is composed of different recognition results. The paper has choosed 30 picture samples to test the designed artificial system, and the experimental result shows the integral recognition rate is 70% .