不良图像中的人脸检测方法研究
|
摘要
随着互联网的发展,网络信息已经成为一种人们熟知的便捷信息来源,但网络上大量的色情淫秽等不良信息已经严重干扰了人们正常的网络生活。由于图像比文本具有更丰富的信息,相比之下不良图像具有更大的危害性,因而对不良图像识别过滤是建立网络色情过滤系统的关键。对不良图像的识别,首先要确定人体的存在,检测人脸就是首要工作。人脸的准确定位还能为后续的表情识别、人体姿态检测等任务提供重要信息。因此,自动化的人脸检测就成为不良图像分析的关键,具有深远的社会意义。
为了研究不良图像中的人脸定位问题,本文在分析了现有人脸检测方法的基础上,针对不良图像中人脸具有的复杂模式,设计了一个层次化人脸检测框架,利用肤色、Haar-Like特征、边缘形状、灰度等信息检测定位人脸。本文主要工作如下:
(1)利用肤色检测作为人脸检测的预处理过程,确定了高斯肤色模型参数,建立了YCbCr空间的高斯肤色模型。本文设计了基于K-均值聚类方法的肤色分割算法,完成了图像的二值化。通过对孔洞数和面积的计算,选择出可能包含人脸的候选肤色块,减小了下一步人脸检测的搜索空间。
(2)本文实现了基于Haar-Like特征和Adaboost的人脸检测算法,并将其运用到不良图像的人脸检测中。实验结果表明,对于正面和略带表情及遮挡的人脸有较好的检测效果。在算法实现过程中,本文针对Adaboost训练时间过长的缺陷,在训练弱分类器时,推导出计算左右错误率的递推公式,运用该递推公式加快了弱分类器的训练速度。
(3)针对前面方法漏检的情况,本文运用Hough变换检测椭圆初步定位到人脸的大致位置,从不良图像中截取多个人脸样本,构造平均人脸模板,用该模板与检测到的椭圆区域匹配判断是否是人脸,提高了检测率。
With the development of Internet, netnews has become a familiar and convenient information source. But much pornographic and other undesirable information has seriously disturbed people's normal life on the net. Because images contain more information than texts, compared to texts, pornographic images are even more harmful. Thus recognition of pornographic images is the key of the system with filtering pornography online. In order to recognise the pornographic images, the first step is to determine the existence of the human body. So detecting human face is the primary task. In addition location face exactly can also provide important information for later tasks such as identification of face expression and detection of human body posture. Therefore, automated detection of face has become the pivotal step to analyse pornographic images, and it has far-reaching social significance.
In order to study the problem of face detection in the pornographic images, this thesis initially discusses the the existing face detection methods, and then aiming at the complex model about human face in pornographic images, a hierarchical framework for face detection is designed. The framework has used several kinds features to detect face, such as color, Haar-Like features, edge shape, gray information. The main work of this paper is as follows:
(1) skin segmentation is the pretreatment step for face detection. The parameters of skin Gaussian model is determined and the model based on the color space of YCbCr is established. In this paper, a color segmentation algorithm based on the K-means clustering method is proposed. This algorithm can change the original image into skin binary image. Through calculating the number of holes and the area, a candidate block is selected which generally contain face. The process of skin segmentation reduces the space which the later steps of face detection would search for.
(2) In this paper, a face detection method based on Haar-Like features and Adaboost is implemented and applied into the pornographic images. The experimental results show that this method is feasible and effetive for positive and slight sheltery face. Considering original adaboost arithmetic costs too much time for training weak classifiers, a recurrence formula is deduced to calculate the false positive rate. This formula can accelerate the training speed obviously.
(3) Aiming at undetected images, this paper build an average face template through clipping some face samples from pornographic images. And then hough transform and template matching method are used to locate the position where face would be and verify whether the probable position is face indeed. The final detection result will be obtained and the detection rate is improved.
为了研究不良图像中的人脸定位问题,本文在分析了现有人脸检测方法的基础上,针对不良图像中人脸具有的复杂模式,设计了一个层次化人脸检测框架,利用肤色、Haar-Like特征、边缘形状、灰度等信息检测定位人脸。本文主要工作如下:
(1)利用肤色检测作为人脸检测的预处理过程,确定了高斯肤色模型参数,建立了YCbCr空间的高斯肤色模型。本文设计了基于K-均值聚类方法的肤色分割算法,完成了图像的二值化。通过对孔洞数和面积的计算,选择出可能包含人脸的候选肤色块,减小了下一步人脸检测的搜索空间。
(2)本文实现了基于Haar-Like特征和Adaboost的人脸检测算法,并将其运用到不良图像的人脸检测中。实验结果表明,对于正面和略带表情及遮挡的人脸有较好的检测效果。在算法实现过程中,本文针对Adaboost训练时间过长的缺陷,在训练弱分类器时,推导出计算左右错误率的递推公式,运用该递推公式加快了弱分类器的训练速度。
(3)针对前面方法漏检的情况,本文运用Hough变换检测椭圆初步定位到人脸的大致位置,从不良图像中截取多个人脸样本,构造平均人脸模板,用该模板与检测到的椭圆区域匹配判断是否是人脸,提高了检测率。
With the development of Internet, netnews has become a familiar and convenient information source. But much pornographic and other undesirable information has seriously disturbed people's normal life on the net. Because images contain more information than texts, compared to texts, pornographic images are even more harmful. Thus recognition of pornographic images is the key of the system with filtering pornography online. In order to recognise the pornographic images, the first step is to determine the existence of the human body. So detecting human face is the primary task. In addition location face exactly can also provide important information for later tasks such as identification of face expression and detection of human body posture. Therefore, automated detection of face has become the pivotal step to analyse pornographic images, and it has far-reaching social significance.
In order to study the problem of face detection in the pornographic images, this thesis initially discusses the the existing face detection methods, and then aiming at the complex model about human face in pornographic images, a hierarchical framework for face detection is designed. The framework has used several kinds features to detect face, such as color, Haar-Like features, edge shape, gray information. The main work of this paper is as follows:
(1) skin segmentation is the pretreatment step for face detection. The parameters of skin Gaussian model is determined and the model based on the color space of YCbCr is established. In this paper, a color segmentation algorithm based on the K-means clustering method is proposed. This algorithm can change the original image into skin binary image. Through calculating the number of holes and the area, a candidate block is selected which generally contain face. The process of skin segmentation reduces the space which the later steps of face detection would search for.
(2) In this paper, a face detection method based on Haar-Like features and Adaboost is implemented and applied into the pornographic images. The experimental results show that this method is feasible and effetive for positive and slight sheltery face. Considering original adaboost arithmetic costs too much time for training weak classifiers, a recurrence formula is deduced to calculate the false positive rate. This formula can accelerate the training speed obviously.
(3) Aiming at undetected images, this paper build an average face template through clipping some face samples from pornographic images. And then hough transform and template matching method are used to locate the position where face would be and verify whether the probable position is face indeed. The final detection result will be obtained and the detection rate is improved.
引文
[1].Chellappa R,Wilson C L,Sirohey S.Human and machine recognition of face:A Survey Proceedings of the IEEE,1995,83(5):705-740.
[2].T.Sakai,M.Nagao and I.Kanade.Computer Analysis and Classification of Photographs of Human Faces.Proc.First USA-Japan Computer Conference,1972,2-7.
[3].Belhumeur P,Hespanha J,Kriegman D.Eigenfaces vs.Fisher faces:Recognition Using Class Specific Linear Projection.IEEE Trans.Pattern Analysis and Machine Intelligence,1997,vol.19,no.7,pp.711-720.
[4].M.H.Yang,D.J.Kriegman and N.Ahuja.Detecting Faces in Images:A Survey.IEEE Trans.Patern Analysis and Machine Intelligence.Jan 2002,24(1):34-58.
[5].孙宁 邹采荣 赵力.人脸检测综述[J].电路与系统学报,2006.12,Vol.11,No.6.
[6].E.Hjelm and B.K.Low.Face Detection:A survey.Computer Vision and Image Understanding,Sept.2001,vol.83,110.3,pp.236-274.
[7].G Yang,T S Huang.Human Face Detection in Complex Background Pattern Recognition[J].1994,27(1):53-63.
[8].卢春雨,张长水,闻方等.基于区域特征的快速人脸检测法[J].清华大学学报(自然科学版),1999,39(1):101-105.
[9].C.Kotropoulos and I.Pitas.Rule-Based Face Detection in Frontal Views Proc[C].Int'l Conf.Acoustics,Speech and Signal Processing,1997,4:2537-2540
[10].冯建强,刘文波,于盛林.基于灰度积分投影的人眼定位[J].计算机仿真,2005.4,第22卷,第4期.
[11].H P Graf,et al.Locating Faces and Facial Parts[A].Proc.1st Workshop Automatic Face and Gesture Recognition[C].1995,41-46.
[12].Kin Chong Yow,Roberto Cipolla.Feature-Based Human Face Detection.CUED/FINFENG/I'R,August 1996,249.
[13].T.K.Leung,M.C.Burl,and P.Perona.Finding Faces in Clutered Scenes Using Random Labeled Graph Matching.Proc.Fifth IEEE Conf.Computer Vision,1995,5:637-644.
[14].S.A.Sirohey.Human Face Segmentation and Identification.Technical Report CS-TR-3176.Univ.of Maryland.1993.
[15].Wang J G,Tan T N.A new face detection method based on shape information.Pattern Recognition Letters,2000,21(6-7):463-471.
[16].M.F.Augusteijn and T.L.Skujca,Identification of Human Faces through Texture-Based Feature Recognition and Neural Network Techonology,Proc.IEEE Conf.Neural Networks,1993,pp.392-398.
[17].Y.Dai and Y.Nakano,Face-Texture Model Based on SGLD and Its Application in Face Detection in a Color Scene,Pattern Recognition,1996,vol.29,no.6,pp.1007-1017.
[18].J.C.Terrillon,M.N.Shirazi,H.Fukamach,S.Akamatsu.Comparative Performance of different skin chrominance models and chrominance spaces for the automatic detection of human faces in color images[J],IEEE Proceeding of the International Conference on Face and Gesture Recognition,2000,54-61.
[19].Akamatsu:The IEEE Third International Conference on Automatic Face and Gesture Recognition[J], Nara, Japan, April 1998,14-16.
[20].Rein-Lien Hsu, M.Addel-Motaleb, and A.K.Jain.Face Detection in Color Images.Transactions on Patern Analysis and Machine Intelligence.2002,24(5): 696-706.
[21].T.Sakai, M.Nagao, and S.Fujibayashi.Line Extraction and Pattern Detection in a Photograph[J],Pattern Recognition, 1969,1:233-248.
[22].Craw , Elljs H, Lishman J.Automatic extraction of face features.Pattern Recognition Letters,1987,5(2): 183-187.
[23].Yuille A,Hallinan P,Cohen D.Feature Extraction from Faces Using Deformable Templates[J].International Journal of Computer Vision.1992,8.2:99-111.
[24].M Turk.A Pentland.Eigenfaces for Recognition [J].J.cognitive Neuroscience, 1991, 3(1),71-86.
[25].B Moghaddam, A Pentland.Probabilistic visual learning for object representation [J].IEEE Trans.Pattern Analysis and Machine Intelligence, 1997,19(7): 696-710.
[26].Kah-Kay Sung, Tomaso Poggio.Example-Based Learning for View-Based Human Face Detection.IEEE Translation on Patem Analysis and Machine Intelligence, Jan 1998, 20(1): 39-50.
[27].H M El-Bakry, H Stoyan.Fast neural networks for sub-matrix (object/face) detection, Circuits and Systems [A].ISCAS '04.Proceedings of the 2004 International Symposium [C], 2004,764-767.
[28].Edgar Osuna, Robert Freund and Federico Girosi.Training Support Vector Machines: an Application to Face Detection.Proceedings of CVPR'97,1997,130-136.
[29].J C Platt.Sequential minimal optimization: A fast algorithm for training support vector machines [R].MSR: Technical Report MSR-TR-98-14,1998.
[30].Nefian A V.A Hidden Markov Model-Based Approach for Face Detection and Recognition[J].school of Electrical and Computer Engineering, Georigia Institute of Technology, 1998.
[31].A.V.Nefian, M.H.Hayes Ⅲ.An embedded HMM based approach for face detection and recognition [C], IEEE Inter.Conf.on Acoustics, Speech and Signal Processing,Phoenix AZ.1999.
[32].R Meir, G Raetsch.An introduction to Boosting and Leveraging [A].Advanced Lectures on Machine Learning, LNCS [C].2003.119-184.
[33].P Viola, M Jones.Robust real time object detection [A].8th IEEE International Conference on Computer Vision (ICCV) [C].Vancouver, British Columbia, 2001.
[34].Elli Angelopoulou.Understanding the Color of Human Skin.In Proceedings of the SPIE Conference on Human version and Electronic Imaging VI (SPIE) 2001,SPIE Vol.4229,SPIE Press,2001,243-251
[35].S.L.Phung, D.Chai, A, Bouzerdounm.A novel skin color medel in YcrCb color space and its application to human face detection.IEEE International Conference on Image Processing(ICIP/2002), Rochester, USA, Sep.2002,22-25.
[36].J.S.Tang, S.Acton Locating human faces in a complex background including non-face skin color Journal of Electronic Imaging 2003,12(3), 423-430.
[37].N.Tsapatsoulis, S.Kollias Face Detection in color images and video sequence Electrotechnical Conf., 10th Mediterranean, 2002,Vol.2 pp:498-502.
[38].Henry Chang and Ulises Robles.Face Detection.Internet Publication, available at http://www-cs- students.stanford.edu/robles/ee368/main.html.
[39].MENSER, B., AND WIEN, M.2000.Segmentation and Tracking of Facial Regions in Color Image Sequences.In Proceeding SPIE Visual Communications and Image Processing, 2000, 731-740.
[40].lausi D A.K-means Iterative Fisher(KIF) unsupervised clustering algorithm applied to image texture segmentation[J].Pattern Recognition,2002,35:1959-1972.
[41].于殿泓.图像检测与处理技术[M].西安电子科技大学出版社,2006,p182-183.
[42].Y.Freund and R.E.Schapire.Experiments with a New Boosting Algorithm.In Proc.13th International Conference on Machine Learning,1996,pp.148-156.
[43].Freund,Y.and Schapire,R.E.A decision theoretic generalization of on-line learning and an application to boosting,Journal of Computer and System Science,1997,55(1):pages:119-139.
[44].Viola Paul,Jones Michael.Rapid Object Detection using a Boosted Cascade of Simple Features.In:Proc.IEEE Conf.on Computer Vision and Pattern Recognition,Kauai,Hawaii,USA,2001.
[45].赵楠.基于AdaBoost算法的人脸检测[D].北京大学,2005.6.
[46].罗明刚.基于adaboost算法的人脸检测系统的设计与实现[D].昆明理工大学,2006.
[47].王爱玲 叶明生 邓秋香.MATLAB R2007图像处理技术与应用[M].电子工业出版社p166-171.
[48].赵瑞.基于肤色和模板的人脸检测算法研究[D].西北工业大学,2004.3
[49].Krishnan Nallaperumal,Senior Member,IEEE,Ravi Subban Member,IEEE,Krishnaveni3.K,Member,IEEE,Lenin Fred4,Member,IEEE,R.K.Selvakumar5 Member,IEEE.Human Face Detection in Color Images Using Skin Color and Template Matching Models for Multimedia on the Web.
[50].David Rademacher.Face Detection.Colorado School of Mines.Introduction to Computer Vision and Image.Processing.EGGN 498C December 6,2001.
[2].T.Sakai,M.Nagao and I.Kanade.Computer Analysis and Classification of Photographs of Human Faces.Proc.First USA-Japan Computer Conference,1972,2-7.
[3].Belhumeur P,Hespanha J,Kriegman D.Eigenfaces vs.Fisher faces:Recognition Using Class Specific Linear Projection.IEEE Trans.Pattern Analysis and Machine Intelligence,1997,vol.19,no.7,pp.711-720.
[4].M.H.Yang,D.J.Kriegman and N.Ahuja.Detecting Faces in Images:A Survey.IEEE Trans.Patern Analysis and Machine Intelligence.Jan 2002,24(1):34-58.
[5].孙宁 邹采荣 赵力.人脸检测综述[J].电路与系统学报,2006.12,Vol.11,No.6.
[6].E.Hjelm and B.K.Low.Face Detection:A survey.Computer Vision and Image Understanding,Sept.2001,vol.83,110.3,pp.236-274.
[7].G Yang,T S Huang.Human Face Detection in Complex Background Pattern Recognition[J].1994,27(1):53-63.
[8].卢春雨,张长水,闻方等.基于区域特征的快速人脸检测法[J].清华大学学报(自然科学版),1999,39(1):101-105.
[9].C.Kotropoulos and I.Pitas.Rule-Based Face Detection in Frontal Views Proc[C].Int'l Conf.Acoustics,Speech and Signal Processing,1997,4:2537-2540
[10].冯建强,刘文波,于盛林.基于灰度积分投影的人眼定位[J].计算机仿真,2005.4,第22卷,第4期.
[11].H P Graf,et al.Locating Faces and Facial Parts[A].Proc.1st Workshop Automatic Face and Gesture Recognition[C].1995,41-46.
[12].Kin Chong Yow,Roberto Cipolla.Feature-Based Human Face Detection.CUED/FINFENG/I'R,August 1996,249.
[13].T.K.Leung,M.C.Burl,and P.Perona.Finding Faces in Clutered Scenes Using Random Labeled Graph Matching.Proc.Fifth IEEE Conf.Computer Vision,1995,5:637-644.
[14].S.A.Sirohey.Human Face Segmentation and Identification.Technical Report CS-TR-3176.Univ.of Maryland.1993.
[15].Wang J G,Tan T N.A new face detection method based on shape information.Pattern Recognition Letters,2000,21(6-7):463-471.
[16].M.F.Augusteijn and T.L.Skujca,Identification of Human Faces through Texture-Based Feature Recognition and Neural Network Techonology,Proc.IEEE Conf.Neural Networks,1993,pp.392-398.
[17].Y.Dai and Y.Nakano,Face-Texture Model Based on SGLD and Its Application in Face Detection in a Color Scene,Pattern Recognition,1996,vol.29,no.6,pp.1007-1017.
[18].J.C.Terrillon,M.N.Shirazi,H.Fukamach,S.Akamatsu.Comparative Performance of different skin chrominance models and chrominance spaces for the automatic detection of human faces in color images[J],IEEE Proceeding of the International Conference on Face and Gesture Recognition,2000,54-61.
[19].Akamatsu:The IEEE Third International Conference on Automatic Face and Gesture Recognition[J], Nara, Japan, April 1998,14-16.
[20].Rein-Lien Hsu, M.Addel-Motaleb, and A.K.Jain.Face Detection in Color Images.Transactions on Patern Analysis and Machine Intelligence.2002,24(5): 696-706.
[21].T.Sakai, M.Nagao, and S.Fujibayashi.Line Extraction and Pattern Detection in a Photograph[J],Pattern Recognition, 1969,1:233-248.
[22].Craw , Elljs H, Lishman J.Automatic extraction of face features.Pattern Recognition Letters,1987,5(2): 183-187.
[23].Yuille A,Hallinan P,Cohen D.Feature Extraction from Faces Using Deformable Templates[J].International Journal of Computer Vision.1992,8.2:99-111.
[24].M Turk.A Pentland.Eigenfaces for Recognition [J].J.cognitive Neuroscience, 1991, 3(1),71-86.
[25].B Moghaddam, A Pentland.Probabilistic visual learning for object representation [J].IEEE Trans.Pattern Analysis and Machine Intelligence, 1997,19(7): 696-710.
[26].Kah-Kay Sung, Tomaso Poggio.Example-Based Learning for View-Based Human Face Detection.IEEE Translation on Patem Analysis and Machine Intelligence, Jan 1998, 20(1): 39-50.
[27].H M El-Bakry, H Stoyan.Fast neural networks for sub-matrix (object/face) detection, Circuits and Systems [A].ISCAS '04.Proceedings of the 2004 International Symposium [C], 2004,764-767.
[28].Edgar Osuna, Robert Freund and Federico Girosi.Training Support Vector Machines: an Application to Face Detection.Proceedings of CVPR'97,1997,130-136.
[29].J C Platt.Sequential minimal optimization: A fast algorithm for training support vector machines [R].MSR: Technical Report MSR-TR-98-14,1998.
[30].Nefian A V.A Hidden Markov Model-Based Approach for Face Detection and Recognition[J].school of Electrical and Computer Engineering, Georigia Institute of Technology, 1998.
[31].A.V.Nefian, M.H.Hayes Ⅲ.An embedded HMM based approach for face detection and recognition [C], IEEE Inter.Conf.on Acoustics, Speech and Signal Processing,Phoenix AZ.1999.
[32].R Meir, G Raetsch.An introduction to Boosting and Leveraging [A].Advanced Lectures on Machine Learning, LNCS [C].2003.119-184.
[33].P Viola, M Jones.Robust real time object detection [A].8th IEEE International Conference on Computer Vision (ICCV) [C].Vancouver, British Columbia, 2001.
[34].Elli Angelopoulou.Understanding the Color of Human Skin.In Proceedings of the SPIE Conference on Human version and Electronic Imaging VI (SPIE) 2001,SPIE Vol.4229,SPIE Press,2001,243-251
[35].S.L.Phung, D.Chai, A, Bouzerdounm.A novel skin color medel in YcrCb color space and its application to human face detection.IEEE International Conference on Image Processing(ICIP/2002), Rochester, USA, Sep.2002,22-25.
[36].J.S.Tang, S.Acton Locating human faces in a complex background including non-face skin color Journal of Electronic Imaging 2003,12(3), 423-430.
[37].N.Tsapatsoulis, S.Kollias Face Detection in color images and video sequence Electrotechnical Conf., 10th Mediterranean, 2002,Vol.2 pp:498-502.
[38].Henry Chang and Ulises Robles.Face Detection.Internet Publication, available at http://www-cs- students.stanford.edu/robles/ee368/main.html.
[39].MENSER, B., AND WIEN, M.2000.Segmentation and Tracking of Facial Regions in Color Image Sequences.In Proceeding SPIE Visual Communications and Image Processing, 2000, 731-740.
[40].lausi D A.K-means Iterative Fisher(KIF) unsupervised clustering algorithm applied to image texture segmentation[J].Pattern Recognition,2002,35:1959-1972.
[41].于殿泓.图像检测与处理技术[M].西安电子科技大学出版社,2006,p182-183.
[42].Y.Freund and R.E.Schapire.Experiments with a New Boosting Algorithm.In Proc.13th International Conference on Machine Learning,1996,pp.148-156.
[43].Freund,Y.and Schapire,R.E.A decision theoretic generalization of on-line learning and an application to boosting,Journal of Computer and System Science,1997,55(1):pages:119-139.
[44].Viola Paul,Jones Michael.Rapid Object Detection using a Boosted Cascade of Simple Features.In:Proc.IEEE Conf.on Computer Vision and Pattern Recognition,Kauai,Hawaii,USA,2001.
[45].赵楠.基于AdaBoost算法的人脸检测[D].北京大学,2005.6.
[46].罗明刚.基于adaboost算法的人脸检测系统的设计与实现[D].昆明理工大学,2006.
[47].王爱玲 叶明生 邓秋香.MATLAB R2007图像处理技术与应用[M].电子工业出版社p166-171.
[48].赵瑞.基于肤色和模板的人脸检测算法研究[D].西北工业大学,2004.3
[49].Krishnan Nallaperumal,Senior Member,IEEE,Ravi Subban Member,IEEE,Krishnaveni3.K,Member,IEEE,Lenin Fred4,Member,IEEE,R.K.Selvakumar5 Member,IEEE.Human Face Detection in Color Images Using Skin Color and Template Matching Models for Multimedia on the Web.
[50].David Rademacher.Face Detection.Colorado School of Mines.Introduction to Computer Vision and Image.Processing.EGGN 498C December 6,2001.