车牌字符识别技术研究
详细信息    本馆镜像全文|  推荐本文 |  |   获取CNKI官网全文
摘要
随着国民经济的日益发展,各种车辆迅速增加,在改变人们生活便利的同时,也引发了许多交通管理和监控问题,智能交通管理系统能提高交通管理的效率,使之科学化、规范化。车牌自动识别技术是智能交通管理系统的核心部分。基于图像的车辆牌照自动识别系统一般包括车牌定位、字符检测、字符识别三个主要模块。车辆牌照自动识别系统的关键指标之一是字符识别的准确率和识别速度。
     随着车牌识别系统应用范围的逐渐扩大,比如夜间车牌识别,高速路车牌识别等,对车牌识别系统识别速度和识别准确率都提出了更高的要求,因此就有必要研究如何在实时条件下提高系统识别准确率。
     本文首先根据车牌汉字分辨率低,很难有效二值化的特点,使用Gabor变换直接提取灰度车牌汉字的图像特征。提出了自己的一些参数选取方法,解决了车牌汉字识别率低的问题。
     车牌识别系统受环境影响很大,切分得到的车牌字符在特征空间中是以多聚类形式存在的,因此我们把多模板聚类方法用到了车牌字符识别中。提出了根据不同字符之间的关系的思想来定义新的类内类间距离公式,并提出了一种改进的准则函数。多模板匹配进一步提高了车牌汉字识别率。
     多分类器集成是现今流行的一种模式分类方法。本文在比较现今流行的一些多分类器集成方法的基础上,结合现有系统中已有分类器的特点,提出了一种结合投票的置信度加权多分类器集成方法。根据多分类器集成需要,引入了最近邻分类器置信度估计方法,并提出了一种置信度映射方法,用来增强组成多分类器系统的各个基本分类器之间的置信度可比性。实验证明多分类器集成
With the development of national economy, the amount kinds of vehicles have increased rapidly. Although makes in deed our life convenient, it cause a lot of traffic surveillance problems, Intelligent traffic surveillance (ITS) can improve the efficiency of the traffic surveillance, makes it scientificalness, standardization. License plate recognition (LPR) technique is the kernel of ITS. License plate recognition consists of three modules in general: plate location, character detection and character recognition. Two of the key indices of this kind of technique are the character recognition accuracy and speed.
    With application area of LPR enlarge, for example, nighttime LPR, free way LPR, there has a higher demands with the recognize speed and accuracy of LPR. So need to research how to raise the performance of LPR in real time condition.
    Because of the low-resolution rate of Chinese character recognition, we use Gabor transform to extract monochrome Chinese character features. Analyzed Gabor parameters and adopt some new parameter selection methods, solved low-resolution of Chinese character recognize.
    Because of the complex environment, the characters what we got are in a multiple cluster style in feature space, so we use multiple template to recognize Chinese characters. Defined new in cluster and between cluster distance, and new rule functions, experiments show multiple templates can efficiently improve the
引文
[1] 张世辉,孔令富.汉字识别及现状分析[J].燕山大学学报,2003,27(4):367—369.
    [2] 潘翔等.基于小波的车牌汉字特征提取[J].中国图像图形学报,2003,8(10):1218-1222.
    [3] John Daugman. High Confidence Recognition of Persons by Rapid Video Analysis of Iris Texture[R]. European Convention on Security and Detection, 1995, 244-251.
    [4] Turner M R. Texture Discrimination by Gabor Functions[J]. Biol Cybem, 1986, 55: 71-82.
    [5] 宋刚,徐光佑.基于特征的人睑验证[D].清华大学毕业设计论文,2001.
    [6] 王学文,丁晓青,刘长松..基于Gabor变换的高鲁棒汉字识别新方法[J].电子学报,2002,9:1317-1321.
    [7] Lee Seong Whan, Kim Yong Joon. Direct extraction of topographic feature for gray scale character recongnition[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1995, 17(7): 724-729.
    [8] 施泽生等.一种基于多模板匹配的在线手写签名认证方法[J].计算机应用,2004,24:165-167.
    [9] J. G. Daugman. Complet Discrete 2-D Gabor Transforms by Neural Networks for image analysis and compression[J]. IEEE Trans on ASSP, 1988, 36(7): 1169-1179.
    [10] 左文明,黎绍发,曾宪贵.BP算法在手写体汉字识别中的应用[J].计算机工程与设计,2003,24(10):71-72.
    [11] 王晓光,王晓华.一种基于SVM的车牌汉字的有效识别方法[J].计算机工程与应用,2004,24:208-209.
    [12] 童学锋.一种简单有效的多分类器综合方祛[J].计算机工程,2003,29(17):110-111.
    [13] 施泽生,蔡洪滨,饶海潮等.一种基于多模板匹配的在线手写签名认证方法许[J].计算机应用,2004,24:165-167.
    [14] 袁超,张长水.基于多模板匹配的自动人脸检测[J].电子学报,2000,3:95-98.
    [15] 韦燕凤,彭思龙.基于模扳聚类与综合的多模板快速定位算法[J].中国图像图形学报,2004.9(3):314-317.
    [16] Suen C Y(Eds). Frontiers in Handwriting Recognition[C]. International Workshop on Frontiers in Handwriting Recognition, 1990: 131-143.
    [17] Selfridge. a paradigm for learning, held at the National Physical Lab, 1958, 513-526.
    [18] Kanal, L. Patterns in pattern recognition. IEEE Trans. Information Theory, 1974, 20: 697-722.
    [19] Hansen, L. K., Salamon. Neural network ensembles. IEEE Transactions onPattern Analysis and Machine Intelligence, 1990, 12: 993-1000.
    [20] Turner, K., Ghosh, J. Analysis of decision boundaries in linearly combined neural classi, ers. Pattern Recognition, 1996, 29: 341-348.
    [21] R. Barandela, J. S. Sanchez, R. M. Yaldovinos. New Applications of Ensembles of Classifiers, Pattern Anal Applic, 2003, 6: 245-256.
    [22] 董火明等.多分类器融合的指纹全局特征协同识别[J].电路与系统学报,2005,10(3):58-63.
    [23] 朱虹等.基于多特征的多分类器融合决策印鉴识别[J].计算机工程与应用,2004,24:215-217.
    [24] Giorgio Valentini, Francesco Masulli. Ensembles of learning machines[J]. Computer Science, 2002, 2486: 3-22.
    [25] Dietterich, T. G. Ensemble methods in mchine learning[J]. Multiple Classifier Systems, 2001, 1857.
    [26] 寇忠宝等.基于Multi2Agent的分类器融合[J].计算机学报,2003,26(2);174-179.
    [27] R. P. W. Duin. The combining classifier: to train or not totrain?[C]. In Proc. of the 16th Intl. Conf. on Pattern Recognition, 2002, 765-770.
    [28] Peng-Tao Jia, Hua-Can He, Wei Lin. Decision by maximum of posterior probability average with weights: a method of multiple classifiers combination[J]. Machine Learning and Cybernetics, 2005, 4: 18-21.
    [29] 景志宏等.基于多神经网络分类器的目标识别仿真实验研究[J].系统仿真学报,2003,15(3):441-443.
    [30] 崔国辉等.基于子带处理多分类器融合的说话人识别[J].山东大学学报(工学版),2004,34(4):108-111.
    [31] 龚振雄等.特征描绘法[M],清华大学出版社,1998.
    [32] 林晓帆,丁亮青等.最近邻分类器置信度估计的理论分析[J].科学通报,1998,43(3):322-325.
    [33] 林晓帆,丁晓青等.字符识别的置信度分析[J].清华大学学报(自然科学版),1998,38(9):47-50.
    [34] 王年等.基于神经网络的汽车牌照自动识别[J],安徽大学学报(自然科学版),2000,24(1):45-49.
    [35] 吴炜等.一种基于Hausdorff距离的车牌字符识别算法[J].计算机应用研究,2004,2:258-260.
    [36] Fumera G, Roli F. A Theoretical and Experimental Analysis of Linear Combinersfor Multiple Classifier Systems[J]. Pattern Analysis and Machine Intelligence, IEEE Transactions, 2005, 27(6): 942-956.
    [37] Karahan, M, Hakkani-Tur, D, Riccardi, G, rur, G. Combining classifiers for spoken language understanding[J]. Automatic Speech Recognition and Understanding, 2003, 3: 589-594.

© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号

地址:北京市海淀区学院路29号 邮编:100083

电话:办公室:(+86 10)66554848;文献借阅、咨询服务、科技查新:66554700