Research On Image Acquisition And License Plate Location In Vehicle License Plate Recognition System

License plate recognition technology is an important research topic of the intelligent traffic management system,is widely used in the field like road traffic monitoring,automatic charging system,district vehicle management system,safety monitoring and so on.Usually,the license plate recognition system mainly includes four modules: image acquisition,license plate location,character segmentation,character recognition.Therein,the quality of the license plate that taked by the image acquisition model is very important to the license plate location,the results of license plate location directly affects the accuracy of the subsequent two operation process.Therefore,study on license plate image acquisition and localization algorithm is very valuable,this article mainly do the the following research on that two modules.License plate image acquisition module,the CCD camera was selected to gather the vehicle image after comparing to CMOS image sensor,the front-end trigger module for getting vehicle images of acquisition module is studied as well as the back-end communication module for transmitting images to PC.License plate locating module,a license plate locating method based on HSV color model combined with the properties of gray vehicle license edge feature and texture feature is given in this paper.This locating module is viewed as the combination of the preprocessing course,secondary license plate positioning and license plate tilt correction.Therein,the HSV color model is mainly used in the pretreatment process,the properties of gray vehicle license edge feature and texture feature are applied in the subsequent two modules.(1)In the license plate pretreatment process,a license plate preprocessing algorithm based on HSV color model is designed and implemented,the features that H,S,V of HSV color model has no influence to each other is used to brighten the vehicle images,then do gray,denoising and binarization processing;(2)In the secondary positioning process,a license plate location algorithm that based on edge detection and projection is proposed,which including a rough positioning method based on a new Canny operator as well as a precise positioning method based on projection method;(3)Inthe vehicle image tilt correction process,choose the better one Radon transformafter comparing to themethod of Hough transform.Finally,the license plate locating algorithm was realized in Matlab platform,300 vehicle image that captured by CCD camera are divided into two groupsto be used in the experiments,the first groupincluding 200 vehicle image with no obvious inclination,the results show that the accuracy of this algorithm reached 98%;the other groupis made upof 100 vehicle image that with horizontal tilt,the algorithm of location and horizontal tilt correction method that proposed in this paper obtained the accurate location rate of 96%.