License Plate Location Method Based On Image Processing

Vehicle License Plate Recognition (VLPR) is one of the important part in the modern intelligent transportation system. It has wide range of application. It analysis each the vehicle images to obtain each car license plate numbers, thus completing recognition process. It was based on the technology of digital image processing, pattern recognition, computer vision etc. Through the license plate identification can realize parking charges administration, the traffic flow control measurement, vehicle positioning, automobile guard against theft, highway speeding automation supervision etc. has a realistic significance to maintain traffic safety and city security, prevent traffic jams, realize the traffic automation management. And the license plate localization is an important part of the license plate location. It's quality affects the system function directly.This paper studied the method of domestic and foreign representative license plate location. For the low running speed situation, we improved a method of VLP which is based on gray images, i.e. use the mathematical morphological operation to obtain many pieces of similar license area, coarse position. scanning the single line of the image after binary, use the rule among the black and white scanning hopping, determine the license plate area, this method is adopted the one-line scanning, reduced the computation time, improved the speed of the license plate location. On the basis of the coarse position of gray image, we improved a method of using the color information to locate the license area. i.e. segment the color of image, set the background's pixel for 1, and the others for 0. Fill the image by the method of mathematical morphology, we can conclude the license plate area according to the pixel size of license plate area. In this method the effect of the license background's connectivity is better than the number's. It's more suitable for mathematical morphology filling to gain better connected area, improve the positioning accuracy.