利用综合特征的车牌定位方法  

Vehicle License Plate Location By Comprehensive Feature

作　　者：李文举[1] 韦丽华[1] 王洪东[1] 朱正强[1] 

机构地区：[1]辽宁师范大学计算机与信息技术学院,辽宁大连116081

出　　处：《小型微型计算机系统》2011年第8期1655-1659,共5页Journal of Chinese Computer Systems

基　　金：辽宁省博士科研启动基金项目(20061052)资助

摘　　要：车牌定位是车牌识别系统的关键技术之一.提出利用结构特征、纹理特征和颜色特征的车牌定位新方法.首先,利用顶帽变换抑制背景;其次,进行垂直边缘检测和形态滤波,通过结构特征进行车牌粗定位;再次,对候选区域的垂直投影应用一维小波分解滤噪,然后重构垂直投影,计算纹理统计量并构造纹理特征向量,应用BP神经网络识别车牌的字符纹理进行车牌的再定位;最后,对候选区域进行基于边缘颜色对的彩色边缘检测,根据其水平投影值进行车牌的精定位.对各种条件下拍摄的314幅含有车牌的车辆图像应用本算法,定位准确率达到98.7%.Vehicle license plate location is one of the key technologies in vehicle license plate recognition.This paper proposes a novel method of the vehicle license plate locating based on structure feature、texture feature and color feature of the plates.Firstly,it makes use of top-hat transformation to suppress the background of a vehicle-image.Secondly,it does the vertical edge detection and morphological filtering to the image,and fulfills the coarse location of the license plate according to the structure feature.Thirdly,it applies one-dimension wavelet decomposition to the vertical projection of candidate regions to filter the noise,then reconstructs the vertical projection,calculates texture statistics,constructs texture feature vector,and relocates the plate by using BP neural network to recognize the character feature.Finally,it detects the color edge of the candidate regions based on edge color pair,and then fulfills the fine location of the license plate according to the horizontal projection value of the edge images.This method has been applied to 314 vehicle-images which were taken in kinds of conditions for test,and the location accuracy is 98.7%.

关 键 词：车牌定位 结构特征 纹理特征 颜色特征 顶帽变换 小波变换 边缘颜色对 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]