基于单样本注册的人脸视频图像识别研究
|
摘要
近几十年来,人脸识别技术已经取得了突飞猛进的发展,现今的研究人员主要集中在多姿态、多样本的人脸识别方法研究中。对于多姿态、多样本的人脸识别,每个人各种变化的样本获取是很困难的,然而每个人的单张正面照片是很容易获取的,因此研究单样本的人脸识别技术具有重要意义。本文提出了基于单样本注册的人脸视频图像识别研究的方法,其主要的工作如下：
首先,对人脸识别技术的研究现状、发展历史、应用领域进行了论述,对常用的人脸识别方法做出了简要的总结。
其次,本论文提出了利用Haar特征的人脸分类器进行人脸检测,在检测到的人脸上用双眼分类器检测出双眼,为了更加确切的分割出人眼区域,本文综合运用Gabor滤波器和眼睛模板(左眼和右眼的模板)在双眼区域内进行匹配的方法精确定位出眼睛(左眼与右眼)坐标,根据眼睛坐标在人脸上的几何分布对人脸图像进行几何归一化,然后再对几何归一化后的人脸图像进行灰度归一化,以便提高下一步的人脸识别的准确率。
最后,采用基于小波特征融合的单样本注册人脸特征提取,进行分类识别。利用小波变换图像融合的方法,对注册的样本进行小波分解并把其低频信息存入库中,然后对测试人脸图像进行小波分解并提取其高频信息,把库中图像的低频信息与测试图像的高频信息进行融合,得到融合后的人脸图像,把融合图像与测试人脸图像之间的相容性(欧氏距离)作为分类特征,用SVM进行分类得出识别结果。
用本文所提出的方法在部分FERET彩色人脸图库和自己用普通摄像头采集的十一组视频图像进行实验,达到良好的分类识别效果。大量的实验结果和统计分析表明本论文所提出的方法在光照、姿态、表情、饰物等影响下有很强的鲁棒性,并且具有较高的正确识别率和正确的拒绝率。
Face recognition technology have improved by leaps and bounds in recent years. Now researchers focus on the study of multi-pose and multi-sample face recognition, but the method to obtain these images is very different. And the single front face image per person is easy to access. So it is very significant to study the face recognition with single training sample. This paper introduces the face recognition of video image based on single sample registered. The main works were as follows:
First, it is discusses that the developing history, research situation and application fields on face image detection and recognition. This paper gives a survey of existing face recognition methods in recent years.
Second, this paper introduces face detection using face classifier based on Haar features in the detecting images. When the face has been detected, using the trained binocular classifier extract the binocular region in face region. In order to more exact divide the eye areas, combine with Gabor filter and eye-template (left eye template and right eye template) matching accurately finds the eyes coordinates. According to the eyes geometric distribution in face, geometric normalize and grayscale normalize the face image to improve the face recognition rate.
Last, recognize and classify the face images by single-sample registered feature extraction with wavelet transform. Using the methods of wavelet transformation and image fusion, obtain the low frequency information of registered image and deposit it to library and obtain the high frequency information of test image, then fuse the low frequency information in library and the high frequency to a fusion image. The compatibility (Euclidean distance) between fused image and test image as classification feature, classify the test image by SVM.
To experiment on partial color face image in FERET and the 11 groups video image acquired by ordinary camera with the methods discussed above, achieved a good classify recognition. Extensive experiments and statistical analysis illustrates that these methods have strong robustness, high correct recognition rate and correct the rejection rate influencing by the illumination, posture, facial expression, decorations.
首先,对人脸识别技术的研究现状、发展历史、应用领域进行了论述,对常用的人脸识别方法做出了简要的总结。
其次,本论文提出了利用Haar特征的人脸分类器进行人脸检测,在检测到的人脸上用双眼分类器检测出双眼,为了更加确切的分割出人眼区域,本文综合运用Gabor滤波器和眼睛模板(左眼和右眼的模板)在双眼区域内进行匹配的方法精确定位出眼睛(左眼与右眼)坐标,根据眼睛坐标在人脸上的几何分布对人脸图像进行几何归一化,然后再对几何归一化后的人脸图像进行灰度归一化,以便提高下一步的人脸识别的准确率。
最后,采用基于小波特征融合的单样本注册人脸特征提取,进行分类识别。利用小波变换图像融合的方法,对注册的样本进行小波分解并把其低频信息存入库中,然后对测试人脸图像进行小波分解并提取其高频信息,把库中图像的低频信息与测试图像的高频信息进行融合,得到融合后的人脸图像,把融合图像与测试人脸图像之间的相容性(欧氏距离)作为分类特征,用SVM进行分类得出识别结果。
用本文所提出的方法在部分FERET彩色人脸图库和自己用普通摄像头采集的十一组视频图像进行实验,达到良好的分类识别效果。大量的实验结果和统计分析表明本论文所提出的方法在光照、姿态、表情、饰物等影响下有很强的鲁棒性,并且具有较高的正确识别率和正确的拒绝率。
Face recognition technology have improved by leaps and bounds in recent years. Now researchers focus on the study of multi-pose and multi-sample face recognition, but the method to obtain these images is very different. And the single front face image per person is easy to access. So it is very significant to study the face recognition with single training sample. This paper introduces the face recognition of video image based on single sample registered. The main works were as follows:
First, it is discusses that the developing history, research situation and application fields on face image detection and recognition. This paper gives a survey of existing face recognition methods in recent years.
Second, this paper introduces face detection using face classifier based on Haar features in the detecting images. When the face has been detected, using the trained binocular classifier extract the binocular region in face region. In order to more exact divide the eye areas, combine with Gabor filter and eye-template (left eye template and right eye template) matching accurately finds the eyes coordinates. According to the eyes geometric distribution in face, geometric normalize and grayscale normalize the face image to improve the face recognition rate.
Last, recognize and classify the face images by single-sample registered feature extraction with wavelet transform. Using the methods of wavelet transformation and image fusion, obtain the low frequency information of registered image and deposit it to library and obtain the high frequency information of test image, then fuse the low frequency information in library and the high frequency to a fusion image. The compatibility (Euclidean distance) between fused image and test image as classification feature, classify the test image by SVM.
To experiment on partial color face image in FERET and the 11 groups video image acquired by ordinary camera with the methods discussed above, achieved a good classify recognition. Extensive experiments and statistical analysis illustrates that these methods have strong robustness, high correct recognition rate and correct the rejection rate influencing by the illumination, posture, facial expression, decorations.
引文
[1]刘丽华.自动人脸识别方法研究与展望.内江科技,2005(5)：57-58.
[2]文沁,汪增福.基于三维数据的人脸表情识别.计算机仿真2005,22(7)：99-103.
[3]Beymer David, Poggio Tomaso. Face recognition from one example view. Proceedings of the IEEE International Conference on Computer Vision,1995:500-507.
[4]Martinez Aleix M. Recognizing imprecisely localized, partially occluded, and expression variant faces from a single sample per class. IEEE Transactions on Pattern Analysis and Machine Intelligence,2002,24(6):748-763.
[5]陈华杰,韦巍.非线性流形上多姿态人脸检测与识别(博士学位论文),浙江大学,2006.
[6]Wang Jie, Plataniotis K.N, Venetsanopolous, A.N. Selecting kernel eigenfaces for face recognition with one training sample per subject,2006 IEEE International Conference on Multimedia and Expo, ICME 2006-Proceedings,2006:1637-1640.
[7]李崇,刘党辉,沈兰蓀.基于Gabor变换的人眼定位方法.测控技术,2006,25(5)：27-32.
[8]Yuille A, Cohen D, Hallinan P. Face Extraction from Faces using Deformable Templates. In Proceedings IEEE Conference on Computer Vision and Pattern Recognition.1989:104-109.
[9]Zhang Jun, Yan Yong, Martin Lades. Face Recognition:Eigen face, Elastic Matching, and Neural Nets. Proceedings of the IEEE,1998,85(9):1422-1435.
[10]Mariano Alvira and Ryan Rifkin, An Empirical Comparison of SNOW and SVMs for Face Detection, A. I. Memo No.2001-004 January,2001, CBCL Memo 193.
[11]Osuna E, Freund R, Girosi F, etal. Training Support Vector Machines:an Application to Face Detection. Proceedings of Computer Vision and Pattern Recognition,1997:130-136.
[12]Mazanec J, Melisek M, Oravec M. etal. Support vector machines, PCA and LDA in face recognition. Journal of Electrical Engineering,2008,59(4):203-209.
[13]唐述敏,方景龙.基于支持向量机的人脸识别.计算机与数字工程,2005,33(7)：75-78.
[14]章毓晋.图像分割.北京：科学出版社,2001.
[15]李桂清,李陶深.扫描线种子填充算法的问题及改进.广西大学学报,1998,23(3)：289-293.
[16]Zhang Xiaozheng, Gao Yongsheng, Leung Maylor K. H.. Automatic texture synthesis for face recognition from single views, Proceedings 18th International Conference on Pattern Recognition, 2006:1151-1154.
[17]顾国庆,许彦冰.数字图像区域标定的方法.上海理工大学学报,2001,23(4)：295-299.
[18]汪世义,陶亮,工华彬.支持向量机和遗传算法的人脸识别方法.计算机工程与应用,2009,45(12)：164-165.
[19]Wu Bo, Ai Haizhou, Huang Chang. Fast Rotation Invariant Multi-view Face Detection Based on Real Adaboost, Proc. of IEEE Conference on Automatic Face and Gesture Recognition, 2004:79-84.
[20]Shcaperi R. E. A brief introduction to boosting, Proceedings of the sixteenth International Joint Conference of Artificial intelligence, San Francise:Publishers Inc,1999:1401-1306.
[21]Viola P, Jones M. Robust Real-time Object Detection, International Journal of Computer Vision, 2004,57(2):137-154,2004.
[22]钱艺,童卫青.嵌入式系统上基于近红外图像的人脸识别研究(硕士学位论文),华东师范大学,2009.
[23]Rainer L., Alexander K, Pisarevsky V. Empirical analysis of detection cascades of boosted classifiers for rapid object detection. In:DAGM'03 25th Pattern-Recognition Symposium,2003: 297-304.
[24]王慧泽,袭声荣.基于AdaBoost算法和Fisher线性准则的人脸识别研究(硕士学位论文),苏州大学,2008.
[25]褚力,张世永.基于两级Adaboost的LBP快速人脸识别的实现与应用(硕士学位论文),复旦大学,2008.
[26]胡慧.视频检索中人脸识别技术的研究与实现(硕士学位论文),北京邮电大学,2008.
[27]Keating C. F, Keating E. G.. Monkeys and mug shots cues used by mesus monkeys (Macaca mulatta) to recognize a human face, Comparative Psychology,1993,107 (2):131-139.
[28]D. Theory of communication. Journal of the Institute of Electrical Engineers,1946, 93(26):429-457.
[29]恽鸿峰.基于Gabor小波变换的人脸识别研究(硕士学位论文),吉林大学,2005.
[30]Lades M, Vorbriiggen J. C, Buhmann J, et al. Distortion invariant object recognition in the dynamic link architecture, IEEE Transactions on Computers,993,42(3):300-311.
[31]李云峰.基于Gabor小波变换的人脸识别(博士学位论文),大连理工大学,2005.
[32]黄艳,柯丽.基于小波变换的单样本人脸识别方法研究(硕士学位论文),沈阳工业大学,2009.
[33]Ke Li, Huang Yan. Eyes Location based on Dual-orientation Gabor Filters and Templates. Proceedings of the IEEE International Conference on Image and Signal, CISP 2009:3337-3340.
[34]Yang Guosheng, Wang Ting, Zhang Huanlong. Eye location method based on Gabor wavelet and topographic feature extraction. Proceedings of the 7th International Conference on Machine Learning and Cybernetics,2008:2857-2862.
[35]李崇,刘党辉,沈兰蓀.基于Gabor变换的人眼定位方法.测控技术,2006,25(5)：27-32.
[36]张艳君,宋京伟.人脸识别中利用眼球进行人脸归一化.北京联合大学学报(自然科学版),2006,20(2)：35-39.
[37]Gao Yufang, Jing Xiaojun. The key technology for face recognition in video search, Beijing university of posts and telecommunications graduate student degree thesis,2002:28-30.
[38]曹怀信,赵建伟.小波分析发展综述.咸阳师范学院学报,2002,17(6)：5-7.
[39]何保全.基于小波变换与核方法的人脸识别(硕士学位论文),四川大学,2006.
[40]杨福生.小波变换的工程分析与应用.北京：科学出版社,1999.
[41]张学帅,潘泉,赵永强.基于静态多小波变换的图像融合.光电子·激光,2005,16(5)：605-609.
[42]刘振慧,高晶敏,崔天横.一种基于小波变换的图像融合新算法.计算机工程与应用,2007,43(12)：74-76.
[43]李涛,田岩,柳健.一种基于模糊集和小波变换的图像融合方法.华中科技大学学报(自然科学版),2005,33(3)：24-26.
[44]Du shan, Ward R. Wavelet-based illumination normalization for face recognition, IEEE International Conference On Image Processing 2005.Genoa, Italy:IEEE,2005:954-957.
[45]Su Hongtao, Zhao Rongchun. Face recognition based on subspace of variant lighting direction [J]. Computer Engineering,2003,29(8):51-58.
[46]Neo Han Foon, Andrew Teoh Beng Jin, David Ngo Chek Ling. Face recognition Using Wavelet transform and non-negative matrix factorization. A I 2004, Berlin:Springer-Verlag Heidelberg, 2004:192-202.
[47]许高风,丁士圻.基于小波重构的人脸识别.数据采集与处理,2009,24(5)：626-631.
[48]闫荣华,彭进业,李岩.基于小波域奇异值分解的人脸识别方法.计算机工程,2007,33(4)：211-214.
[2]文沁,汪增福.基于三维数据的人脸表情识别.计算机仿真2005,22(7)：99-103.
[3]Beymer David, Poggio Tomaso. Face recognition from one example view. Proceedings of the IEEE International Conference on Computer Vision,1995:500-507.
[4]Martinez Aleix M. Recognizing imprecisely localized, partially occluded, and expression variant faces from a single sample per class. IEEE Transactions on Pattern Analysis and Machine Intelligence,2002,24(6):748-763.
[5]陈华杰,韦巍.非线性流形上多姿态人脸检测与识别(博士学位论文),浙江大学,2006.
[6]Wang Jie, Plataniotis K.N, Venetsanopolous, A.N. Selecting kernel eigenfaces for face recognition with one training sample per subject,2006 IEEE International Conference on Multimedia and Expo, ICME 2006-Proceedings,2006:1637-1640.
[7]李崇,刘党辉,沈兰蓀.基于Gabor变换的人眼定位方法.测控技术,2006,25(5)：27-32.
[8]Yuille A, Cohen D, Hallinan P. Face Extraction from Faces using Deformable Templates. In Proceedings IEEE Conference on Computer Vision and Pattern Recognition.1989:104-109.
[9]Zhang Jun, Yan Yong, Martin Lades. Face Recognition:Eigen face, Elastic Matching, and Neural Nets. Proceedings of the IEEE,1998,85(9):1422-1435.
[10]Mariano Alvira and Ryan Rifkin, An Empirical Comparison of SNOW and SVMs for Face Detection, A. I. Memo No.2001-004 January,2001, CBCL Memo 193.
[11]Osuna E, Freund R, Girosi F, etal. Training Support Vector Machines:an Application to Face Detection. Proceedings of Computer Vision and Pattern Recognition,1997:130-136.
[12]Mazanec J, Melisek M, Oravec M. etal. Support vector machines, PCA and LDA in face recognition. Journal of Electrical Engineering,2008,59(4):203-209.
[13]唐述敏,方景龙.基于支持向量机的人脸识别.计算机与数字工程,2005,33(7)：75-78.
[14]章毓晋.图像分割.北京：科学出版社,2001.
[15]李桂清,李陶深.扫描线种子填充算法的问题及改进.广西大学学报,1998,23(3)：289-293.
[16]Zhang Xiaozheng, Gao Yongsheng, Leung Maylor K. H.. Automatic texture synthesis for face recognition from single views, Proceedings 18th International Conference on Pattern Recognition, 2006:1151-1154.
[17]顾国庆,许彦冰.数字图像区域标定的方法.上海理工大学学报,2001,23(4)：295-299.
[18]汪世义,陶亮,工华彬.支持向量机和遗传算法的人脸识别方法.计算机工程与应用,2009,45(12)：164-165.
[19]Wu Bo, Ai Haizhou, Huang Chang. Fast Rotation Invariant Multi-view Face Detection Based on Real Adaboost, Proc. of IEEE Conference on Automatic Face and Gesture Recognition, 2004:79-84.
[20]Shcaperi R. E. A brief introduction to boosting, Proceedings of the sixteenth International Joint Conference of Artificial intelligence, San Francise:Publishers Inc,1999:1401-1306.
[21]Viola P, Jones M. Robust Real-time Object Detection, International Journal of Computer Vision, 2004,57(2):137-154,2004.
[22]钱艺,童卫青.嵌入式系统上基于近红外图像的人脸识别研究(硕士学位论文),华东师范大学,2009.
[23]Rainer L., Alexander K, Pisarevsky V. Empirical analysis of detection cascades of boosted classifiers for rapid object detection. In:DAGM'03 25th Pattern-Recognition Symposium,2003: 297-304.
[24]王慧泽,袭声荣.基于AdaBoost算法和Fisher线性准则的人脸识别研究(硕士学位论文),苏州大学,2008.
[25]褚力,张世永.基于两级Adaboost的LBP快速人脸识别的实现与应用(硕士学位论文),复旦大学,2008.
[26]胡慧.视频检索中人脸识别技术的研究与实现(硕士学位论文),北京邮电大学,2008.
[27]Keating C. F, Keating E. G.. Monkeys and mug shots cues used by mesus monkeys (Macaca mulatta) to recognize a human face, Comparative Psychology,1993,107 (2):131-139.
[28]D. Theory of communication. Journal of the Institute of Electrical Engineers,1946, 93(26):429-457.
[29]恽鸿峰.基于Gabor小波变换的人脸识别研究(硕士学位论文),吉林大学,2005.
[30]Lades M, Vorbriiggen J. C, Buhmann J, et al. Distortion invariant object recognition in the dynamic link architecture, IEEE Transactions on Computers,993,42(3):300-311.
[31]李云峰.基于Gabor小波变换的人脸识别(博士学位论文),大连理工大学,2005.
[32]黄艳,柯丽.基于小波变换的单样本人脸识别方法研究(硕士学位论文),沈阳工业大学,2009.
[33]Ke Li, Huang Yan. Eyes Location based on Dual-orientation Gabor Filters and Templates. Proceedings of the IEEE International Conference on Image and Signal, CISP 2009:3337-3340.
[34]Yang Guosheng, Wang Ting, Zhang Huanlong. Eye location method based on Gabor wavelet and topographic feature extraction. Proceedings of the 7th International Conference on Machine Learning and Cybernetics,2008:2857-2862.
[35]李崇,刘党辉,沈兰蓀.基于Gabor变换的人眼定位方法.测控技术,2006,25(5)：27-32.
[36]张艳君,宋京伟.人脸识别中利用眼球进行人脸归一化.北京联合大学学报(自然科学版),2006,20(2)：35-39.
[37]Gao Yufang, Jing Xiaojun. The key technology for face recognition in video search, Beijing university of posts and telecommunications graduate student degree thesis,2002:28-30.
[38]曹怀信,赵建伟.小波分析发展综述.咸阳师范学院学报,2002,17(6)：5-7.
[39]何保全.基于小波变换与核方法的人脸识别(硕士学位论文),四川大学,2006.
[40]杨福生.小波变换的工程分析与应用.北京：科学出版社,1999.
[41]张学帅,潘泉,赵永强.基于静态多小波变换的图像融合.光电子·激光,2005,16(5)：605-609.
[42]刘振慧,高晶敏,崔天横.一种基于小波变换的图像融合新算法.计算机工程与应用,2007,43(12)：74-76.
[43]李涛,田岩,柳健.一种基于模糊集和小波变换的图像融合方法.华中科技大学学报(自然科学版),2005,33(3)：24-26.
[44]Du shan, Ward R. Wavelet-based illumination normalization for face recognition, IEEE International Conference On Image Processing 2005.Genoa, Italy:IEEE,2005:954-957.
[45]Su Hongtao, Zhao Rongchun. Face recognition based on subspace of variant lighting direction [J]. Computer Engineering,2003,29(8):51-58.
[46]Neo Han Foon, Andrew Teoh Beng Jin, David Ngo Chek Ling. Face recognition Using Wavelet transform and non-negative matrix factorization. A I 2004, Berlin:Springer-Verlag Heidelberg, 2004:192-202.
[47]许高风,丁士圻.基于小波重构的人脸识别.数据采集与处理,2009,24(5)：626-631.
[48]闫荣华,彭进业,李岩.基于小波域奇异值分解的人脸识别方法.计算机工程,2007,33(4)：211-214.