[1]张 梦,曾毓敏,李鹏程.基于加权小波变换和2D-PCA的人脸识别改进算法[J].南京师范大学学报(工程技术版),2015,15(02):055.
 Zhang Meng,Zeng Yumin,Li Pengcheng.Improved Algorithm of Face Recognition Based on Weighted Wavelet Transform and 2D-PCA[J].Journal of Nanjing Normal University(Engineering and Technology),2015,15(02):055.
点击复制

基于加权小波变换和2D-PCA的人脸识别改进算法
分享到:

南京师范大学学报(工程技术版)[ISSN:1006-6977/CN:61-1281/TN]

卷:
15卷
期数:
2015年02期
页码:
055
栏目:
计算机与信息工程
出版日期:
2015-06-20

文章信息/Info

Title:
Improved Algorithm of Face Recognition Based on Weighted Wavelet Transform and 2D-PCA
作者:
张 梦曾毓敏李鹏程
南京师范大学物理科学与技术学院,江苏 南京 210023
Author(s):
Zhang MengZeng YuminLi Pengcheng
School of Physics and Technology,Nanjing Normal University,Nanjing 210023,China
关键词:
人脸识别加权小波变换2D-PCA算法最近邻分类器
Keywords:
face recognitionweighted wavelet transform2D-PCA algorithmnearest neighbor classifier
分类号:
TP391.4
文献标志码:
A
摘要:
基于小波变换的人脸识别方法通常将图像变换成低频和高频信息,传统的人脸识别算法大多数都是基于小波变换后的低频信息,没有充分利用高频信息,造成了高频信息中对识别有利信息的丢失. 本文提出了一种基于加权小波变换和2D-PCA的人脸识别改进算法. 首先基于二维离散小波(2D-DWT)对图像进行二层小波变换,将所得的低频信息和水平、垂直和对角高频信息进行加权融合. 在此基础上,采用二维主成分分析(2D-PCA)方法进行特征提取; 最后采用最近邻分类器进行分类识别. 基于ORL标准人脸数据库的实验结果表明,本文提出的方法比传统的2D-PCD识别算法和2D-DWT+2D-PCA识别算法有更好的识别效果,且人脸受光照等因素的影响表现出良好的鲁棒性.
Abstract:
Typically the image is converted into low and high frequency information for face recognition method based on wavelet transform,and the traditional face recognition algorithm based on the low-frequency information of wavelet transform,without making full use of high-frequency information,resulting in the loss of high frequency information. This paper presents an improved approach based on the weighted wavelet transform and 2D-PCA face recognition algorithm. Firstly,the second layer wavelet transform of image based on the two-dimensional discrete wavelet(2D-DWT),which contains low-frequency information and horizontal,vertical and diagonal high-frequency information with different weights. Then,using the two-dimensional principal component analysis(2D-PCA)method for feature extraction; finally,using the nearest neighbor classifier for classification. Experimental results based on ORL face database show the proposed approach is superior to the traditional 2D-PCD recognition algorithms,as well as 2D-DWT+2D-PCA recognition algorithms,and it performs good robustness affected by light.

参考文献/References:

[1] Moghaddam B,Pentland A. Face Recognition:From Theory to Applications[M]. Berlin:Springer Verlag,1998.
[2]Li B,Liu Y. When eigenfaces are combined with wavelets[J]. Knowledge-Based Systems,2002,15(5):343-347.
[3]曹智鑫,蔡茂国,陈剑勇. 结合小波局部特征与LDA的人脸识别[J]. 电视技术,2014,38(19):184-187.
Cao Zhixin,Cai Maoguo,Chen Jianyong. Face recognition combining wavelet transform local features and LDA[J]. Television Technology,2014,38(19):184-187.
[4]何旭东. 基于小波分析与NMF人脸识别的研究与实现[D]. 南京:南京理工大学计算机科学与工程学院,2013.
He Xudong. Research and implementation of face recognition based on wavelet analysis and NMF[D]. Nanjing:College of Computer Science and Engineering,Nanjing University of Science and Technology,2013.
[5]苏士美,王燕,王明霞. 基于加权小波分解的人脸识别算法研究[J]. 郑州大学学报:工学版,2014,35(1):5-9.
Su Shimei,Wang Yan,Wang Mingxia. Research of face recognition based on weighted wavelet decomposition[J]. Journal of Zhengzhou University:Engineering Science Edition,2014,35(1):5-9.
[6]刘秀丽,彭复员. 基于小波变换的加权特征脸识别算法[J]. 计算机应用研究,2007,24(10):163-165.
Liu Xiuli,Peng Fuyuan. The weighted feature face recognition algorithm based on wavelet transform[J]. Application Research Of Computers,2007,24(10):163-165.
[7]Kirby M,Sirovich L. Application of the Karhunen-Loeve procedure for the characterization of human faces[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,1990,12(1):103-108.
[8]M Turk A P. Eigenfaces for recognition[J]. Journal of Cognitive Neuroscience,1991,3(1):71-86.
[9]Yang M. Kernel eigenfaces vs. Kernel fisherfaces:face recognition using kernel methods[C]//Fifth IEEE International Conference on Automatic Face and Gesture Recognition. IEEE,2002:215-220.
[10]Gottumukkal R,Asari V. An improved face recognition technique based on modular PCA approach[J]. Pattern Recognition Letters,2004,25(4):429-436.
[11]J Y,D Z,AF F,et al. Two-dimensional PCA:a new approach to appearance-based face representation and recognition[J]. IEEE Trans Pattern Anal Mach Intell,2004,26(1):131-137.
[12]侯建华,牟海军,罗艳,等. 基于MMP-2DPCA的人脸识别方法[J]. 中南民族大学学报:自然科学版,2013,32(1):70-74.
Hou Jianhua,Mou Haijun,Luo Yan,et al. Face recognition method based on MMP-2D-PCA[J]. Journal of Central South University for Nationalities:Natural Science Edition,2013,32(1):70-74.
[13]翟俊海,翟梦尧,王华超. 基于小波变换和2DPCA的人脸识别[J]. 河北大学学报:自然科学版,2010,30(5):574-579.
Zhai Junhai,Zhai Mengyao,Wang Huachao. Face recognition based on wavelet transform and 2D-PCA[J]. Journal of Hebei University:Natural Science Edition,2010,30(5):574-579.
[14]周国民,陈勇,李国军. 人脸识别中应用小波变换的两个关键问题[J]. 浙江大学学报:理学版,2005,32(1):34-38.
Zhou Guomin,Chen Yong,Li Guojun. Two key problems in face recognition using wavelet transform[J]. Journal of Zhejiang University:Sciences Edition,2005,32(1):34-38.
[15]林克正,李艳芳,辛晨. 基于2D-DWT与FLD的人脸识别算法[J]. 计算机工程,2011,37(11):195-196.
Lin Kezheng,Li Yanfang,Xin Chen. Face recognition algorithm based on 2D-DWT and FLD[J]. Computer Engineering,2011,37(11):195-196.
[16]吴献超,刘莎,闫琳,等. 基于小波子带能量加权的人脸识别方法[J]. 宁波大学学报:理工版,2009,22(2):207-211.
Wu Xianchao,Liu Sha,Yan Lin,et al. Face recognition method based on the wavelet subband energy weighted[J]. Journal of Ningbo University:Natural Science & Engineering Edition,2009,22(2):207-211.

相似文献/References:

[1]汪 璠,沈世斌,章 悦,等.基于HOG-SIFT融合优化的多人脸特征提取方法[J].南京师范大学学报(工程技术版),2020,20(03):043.[doi:10.3969/j.issn.1672-1292.2020.03.008]
 Wang Fan,Shen Shibin,Zhang Yue,et al.A Multi-Face Feature Extraction Method Based onHOG-SIFT Feature Fusion Optimization[J].Journal of Nanjing Normal University(Engineering and Technology),2020,20(02):043.[doi:10.3969/j.issn.1672-1292.2020.03.008]

备注/Memo

备注/Memo:
收稿日期:2014-11-06.
通讯联系人:曾毓敏,教授,研究方向:语音信号与图像处理. E-mail:zengyumin@njnu.edu.cn
更新日期/Last Update: 2015-06-20