纹理特征和两级分类器相结合的人脸识别
|
摘要
为了提高人脸识别率和识别效率,提出一种纹理特征和两级分类器相结合的人脸识别方法。采用灰度共生矩阵表示人脸图像的纹理特征,计算待识别人脸图像与模板间欧式距离,采用拒识阈值进行评判,如果人脸图像归属类别清楚,则采用欧式距离分类器进行识别,否则将待识人脸图像送入SVM分类器进行识别,采用ORL人脸数据库和Yale人脸数据库进行仿真实验。仿真结果表明,相对于单一人脸识别器,两级分类器不仅提高了人脸识别效率,而且提高了人脸识别率,具有更好的人脸识别性能。
In order to improve the face recognition rate and recognition efficiency, this paper proposes a new face recognition model based on texture feature and two class classifier combination. Texture features are extracted by gray level co-occurrence matrix, and the Euclidean distance between face image and template, and then the rejection criteria is used for evaluation. If the face image category is clearly, Euclidean distance classifier is used to identify the face, otherwise face image is recognized by SVM classifier. The simulation experiment is carried out on the ORL face database and Yale face database. The simulation results show that, compared with the single classifier, the proposed classifier not only has improved the recognition efficiency, but also improved the rate of face recognition. It has better face recognition performance.
In order to improve the face recognition rate and recognition efficiency, this paper proposes a new face recognition model based on texture feature and two class classifier combination. Texture features are extracted by gray level co-occurrence matrix, and the Euclidean distance between face image and template, and then the rejection criteria is used for evaluation. If the face image category is clearly, Euclidean distance classifier is used to identify the face, otherwise face image is recognized by SVM classifier. The simulation experiment is carried out on the ORL face database and Yale face database. The simulation results show that, compared with the single classifier, the proposed classifier not only has improved the recognition efficiency, but also improved the rate of face recognition. It has better face recognition performance.
引文
[1]Xu Yong,Zhong Aini,Yang Jian,et al.LPP solution schemes for use with face recognition[J].Pattern Recognition,2010,43(12):4165-4176.
[2]Hsieh P C,Tung P C.A novel hybrid approach based on subpattern technique and whitened PCA for face recognition[J].Pattern Recognition,2009,42(5):978-984.
[3]王惠明,史萍.图像纹理特征的提取方法[J].中国传媒大学学报:自然科学版,2006,13(1):49-52.
[4]刘广起,郑晓势,张晓波.基于图像纹理特征提取的车牌定位算法[J].中国图象图形学报,2005,10(11).
[5]Wang Jianzhong,Ma Zhiqiang,Zhang Baoxue,et al.A structure-preserved local matching approach for face recognition[J].Pattern Recognition Letters,2011,32(3):494-504.
[6]甘俊英,张有为.基于BP神经网络的人脸识别[J].系统工程与电子技术,2003,25(1):113-115.
[7]曾阳艳,叶柏龙.基于支持向量机的人脸识别系统的研究[J].计算机工程与应用,2008,44(15):182-184.
[8]白雪飞,李茹.基于2DPCA和RBF神经网络的人脸识别方法[J].计算机工程与应用,2007,43(34):200-203.
[9]谢赛琴,沈福明,邱雪娜.基于支持向量机的人脸识别方法[J].计算机工程,2005,35(6):186-188.
[10]Guo G,Li S Z,Chan K L.Support vector machine for face recognition[J].Image and Vision Computing,2011,19(9/10):631-638.
[11]张健.基于机器视觉的强化木地板表面质量检测方法研究[D].北京:北京林业大学,2009.
[12]陈惠明.图像欧式距离在人脸识别中的应用研究[J].计算机工程与设计,2008,29(14):3735-3737.
[13]马永军,方凯,方廷健.基于支持向量机和距离度量的纹理分类[J].中国图象图形学报,2002,7(11):1151-1155.
[14]叶剑华,刘正光.基于LBP和Fisher faces的多模态人脸识别[J].计算机工程,2009,35(11):193-195.
[15]Ahonen H A,Pietikainen M.Face description with local binary patterns:application to face recognition[J].IEEE Trans on Pattern Analysis and Machine Intelligence,2006,28(12):2037-2041.
[2]Hsieh P C,Tung P C.A novel hybrid approach based on subpattern technique and whitened PCA for face recognition[J].Pattern Recognition,2009,42(5):978-984.
[3]王惠明,史萍.图像纹理特征的提取方法[J].中国传媒大学学报:自然科学版,2006,13(1):49-52.
[4]刘广起,郑晓势,张晓波.基于图像纹理特征提取的车牌定位算法[J].中国图象图形学报,2005,10(11).
[5]Wang Jianzhong,Ma Zhiqiang,Zhang Baoxue,et al.A structure-preserved local matching approach for face recognition[J].Pattern Recognition Letters,2011,32(3):494-504.
[6]甘俊英,张有为.基于BP神经网络的人脸识别[J].系统工程与电子技术,2003,25(1):113-115.
[7]曾阳艳,叶柏龙.基于支持向量机的人脸识别系统的研究[J].计算机工程与应用,2008,44(15):182-184.
[8]白雪飞,李茹.基于2DPCA和RBF神经网络的人脸识别方法[J].计算机工程与应用,2007,43(34):200-203.
[9]谢赛琴,沈福明,邱雪娜.基于支持向量机的人脸识别方法[J].计算机工程,2005,35(6):186-188.
[10]Guo G,Li S Z,Chan K L.Support vector machine for face recognition[J].Image and Vision Computing,2011,19(9/10):631-638.
[11]张健.基于机器视觉的强化木地板表面质量检测方法研究[D].北京:北京林业大学,2009.
[12]陈惠明.图像欧式距离在人脸识别中的应用研究[J].计算机工程与设计,2008,29(14):3735-3737.
[13]马永军,方凯,方廷健.基于支持向量机和距离度量的纹理分类[J].中国图象图形学报,2002,7(11):1151-1155.
[14]叶剑华,刘正光.基于LBP和Fisher faces的多模态人脸识别[J].计算机工程,2009,35(11):193-195.
[15]Ahonen H A,Pietikainen M.Face description with local binary patterns:application to face recognition[J].IEEE Trans on Pattern Analysis and Machine Intelligence,2006,28(12):2037-2041.