基于提升小波的矩不变量车牌字符识别方法

J4 ›› 2016, Vol. 33 ›› Issue (6): 662-670.

基于提升小波的矩不变量车牌字符识别方法

高强¹,刘斌²

湖北大学计算机与信息工程学院湖北武汉 430062

收稿日期:2015-09-28 修回日期:2015-12-14 出版日期:2016-11-28 发布日期:2016-11-28
通讯作者: 刘斌(1963-)，湖北人，博士，教授，博士生导师，从事模式识别、图像处理及小波理论的研究。 E-mail:liubin3318@163.com
作者简介:高强（1988-）湖北人，硕士研究生，从事模式识别与图像处理的研究.E-mail：shengqiang8814@qq.com
基金资助:
国家自然科学基金面上项目（No.61471160），湖北省自然科学基金重点项目（No.2012FFA053） The research foundation: The National Natural Science Foundation of China（No.61471160）, The Key Project of the National Natural Science of Hubei Province（No.2012FFA053）

Vehicle license plate character recognition based on lifting wavelet transform and invariant moment

GAO Qiang, LIU Bin

School of Computer and Information Engineering,Hubei University,Wuhan 430062,China

Received:2015-09-28 Revised:2015-12-14 Published:2016-11-28 Online:2016-11-28

摘要/Abstract

摘要： 车牌识别系统的关键在于字符的识别，而字符识别的核心是字符特征的提取。小波变换可以获取字符的细节结构特征，而不变矩能很好地对其进行描述，故将两者结合起来提取字符的特征；同时利用张量积小波分解的高频子图具有方向性的特点，提取字符的笔画特征，最终得到反映字符结构和统计特征的联和特征向量，并且字符图像的分解采用第二代提升小波算法进一步降低了计算复杂性。实验结果表明，此方法提取得到的联合特征向量能够取得98%的字符综合识别率，可以满足实际应用的要求。

关键词: 模式识别；提升小波变换；不变矩；统计特征；结构特征；笔画特征

Abstract: The key of the license plate recognition system is the character recognition and the core of the character recognition is to extract character features. The wavelet transformation can obtain the details and the structural features for characters and the invariant moment can describe it well. So we combine the two to extract the features of the characters. Meanwhile the directional high frequency sub-images which are decomposed by the tensor product wavelet can be extracted the character strokes feature. Finally we get the alliance feature vector that reflect the structural and statistical features of characters using the second generation lifting wavelet algorithm to further reduce the computational complexity. The experiments results show that the proposed method can achieve 98% recognition rate and can meet the requirements of practical application.

Key words: pattern recognition; lifting wavelet transform; invariant moment; statistical features; structural features; character strokes

高强刘斌. 基于提升小波的矩不变量车牌字符识别方法[J]. J4, 2016, 33(6): 662-670.

GAO Qiang, LIU Bin. Vehicle license plate character recognition based on lifting wavelet transform and invariant moment[J]. J4, 2016, 33(6): 662-670.

参考文献

[1] Wang Tao. Research of License Plate Recotion Arithmetic Based on Wavelet Transformation and Neural Network, 2007:44-45 (in Chinese).
[2] Sun Hua, Zhang Hang. Survey on Chinese Character Recognition Method [J]. Computer Engineering(计算机工程), 2010 ,36(20) (in Chinese).
[3] Trier O D, Jain A K , Taxt T. Feature extraction methods for character recognition-A survey [J]. Pattern Recognition, 1996, 9(4): 641-662.
[4] Song Jiatao, Liu Jilin. Analysis and Extraction of Structural Features of Alphabetic and Digital Characters on Vehicle License Plate [J]. Journal of Image and Graphics(中国图象图形学报),2002,7A(9) (in Chinese).
[5] Luo Huiwu, Tang Yuanyan, Wang Yi, Lan Lijun. License Plate Character Recognition Based on Structural Features and Grayscale Pixel Features Algorithm [J].Computer Science(计算机科学) ,2011, 38(11) (in Chinese).
[6] Shin J, Sakoe H. Optimal Stroke-correspondence Search Method for On-line Character Recognition [J]. Pattern Recognition Letters, 2002, 23(6): 601-608 (in Chinese).
[7] Ding Xiaoqing. Chinese Charaeter Reeognition:A Review [J]. ACTA ELECTRONICA SINICA(电子学报),2002,9(9) (in Chinese).
[8] Wang Xiaohong. Moment Technique and its Applications in Image Processing and Recognition Department of Computer Seieneeand Engineering Northwestem Polytechnical University 2001.4 (in Chinese).
[9] Guan Shuan, Zhou Hong,Lin Guangrong. Fuzzy classification of rice kernel shape based on Zernike moments [J]. Journal of Henan University of Technology(河南工业大学学报),2005,26(1) (in Chinese).
[10] Qu Xinliang, Wei Shou-shui, Du Yang. A recognition algorithm for vehicle-license-plate characters based on wavelet transform [J]. Journal of Shandong University of Architecture and Engineerin(山东建筑工程学院学报),2004,19(3) (in Chinese).
[11] Shuang Xiaochuan, Zhang Ke. Research of handwritten numeral recognition based on statistical and structure characteristics [J]. Computer engineering and design(计算机工程与设计),2012,33(4) (in Chinese).
[12] Mallat S G. A Theory for Multi-resolution Signal Decomposition:The Wavelet Representation
[J]. IEEE Trans. on Pattern Analysis and Machine Intelligence,1989,1(7)：674-693.
[13] Mallat S G. Multiresolution Approximations and Wavelet Orthonormal Bases of L2(R)[J]. Ransactions of the American Mathematical Society, 1989,315(1) :69-87.
[14] I. Daubechies. Orthonormal bases of compactly supported wavelets [J]. Comm. Pure Appl. Math,1988,41：909-996.
[15] Hu M K. Visual pattern recognition by moment invariants IRE trans [J]. Information Theory，1962,8: 179-182.
[16] Jin Qi, Dai Ruwei. Wavelet invariants based on momentre paesentation [J]. PR&AI(模式识别与人工智能), 1995, 8(3): 179-187 (in Chinese).
[17] Li Leida, Guo Baolong, Sun Lei. RST Invariant Watermark Based on Local Zernike Moments [J]. Journal of Optoelectronics Laser(光电子?激光) ,2007,18 ( 1 ) : 117-120 (in Chinese).
[18] Teague M.R. Image analysis via the general theory of moments [J]. Journal of the Optical Society of America ,1980 , 70 ( 8 ) : 920-930 (in Chinese).
[19] Zhang Qin, Yin Junxun. Application Wavelet Trnasformation in Chinees Charcater Recognition [J]. Journal of Circuits and systems(电路与系统学报),1996,1(3):63-67(in Chinese).
[20] Jin Lianwen, Peng Xiulan, Yin Junxun. A New Feature Extraction Approach to Handwritten Chinese Character Based on Wavelet Transform and Elastic Meshing Techniques [J]. Journal of Image and Graphics(中国图象图形学报) ,1998, 3(7) (in Chinese).
[21] Sweldens W., The lifting scheme: A construction of second generation wavelets [J].
SIAM J. Math. Anal, 1997, 29(2):511-546.
[22] Daubechies I.,Sweldens W. Factoring wavelet transforms into lifting steps[J]. J. Four. Anal. Appl.,1998, 4:247-269.
[23] J. F .Boyce, W. J . Hossack . Recognition Moment Invariants for Pattern Recognition [J] . Pattern Recognition , 1983 , 10 (1) : 451- 456.
[24] Wang Pu, Dong Huiying. Extracting Method of Mixed Features for Car License-Plate Characters Recognition[J] Journal of Shenyang Institute of Technolog(沈阳工业学院学报),2003,22(1) (in Chinese).