基于人脸识别的图像美化系统设计与实现
|
摘要
随着社会的不断进步,科学技术的飞速发展,计算机应用领域也在不断拓展,于是数字图像处理技术作为一种全新的图像处理方法应运而生,而且应用范围越来越广。数字图像处理技术是利用计算机设备将图像转变成数字信息来进行保存、处理、传输和重现,通过计算机对图像进行去除噪声、增强、复原、分割、提取特征等处理的方法和技术。人脸识别技术是计算机模式识别技术在数字图像领域中的具体应用,广泛应用于安全验证系统、档案管理、视频会议、人机交互等方面,因而越来越成为当前模式识别和数字图像处理领域的一个非常热门的研究课题。
目前,数字图像处理与人脸识别技术相结合的应用范围越来越广,已经深入到人们的生活之中。但就目前计算机的应用水平而言,由于计算机对外部的感知能力还比较弱,自动化的处理能力还有待提高,仍然需要投入大量人力、物力来从事数字图像处理与人脸识别相结合的理论和应用的研究。
本课题论述了人脸识别技术在图像处理中的一个应用,采用几何特征的方法,包括人脸图像的预处理、人脸轮廓的检测、人脸图像中眉毛、眼睛、鼻子、嘴巴各部位的自动定位以及人脸图像的识别等几大功能。其中人脸检测部分,采用了基于haar-like特征的人脸检测算法,该算法在具有较高的检测率的同时也满足了实时检测的要求。算法利用OpenCV提供的库结合层叠式分类器并使用一系列haar-like特征来描述人脸。在人脸面部特征定位的研究中,采用了基于主动形状模型(ASM)的方法,能够很快的定位到人脸面部的各个特征部位。
本系统的出现为人脸识别技术在数字图像处理方面的应用寻找到了一个新的结合点,也使数字图像处理技术的应用范围更加广泛。
With the double-quick development of society and science technology, the field in computer application is also expanding as soon as possible.So digital image processing techonology has emerged as the times require as a bran-new image processing method.As the same time, it is also applied to more and more domains. Digital image processing is a method and technology which uses computer devices to convert image to digital informations in order to save,process,transmit and recur,and process image for wipe off yawp,enhance,recove,partition and extraction.Face recognition technology as a particular application of pattern recognition in image field is widely applied to surveillance and security, digital libraries,video meeting,and human-computer intelligent interaction and so on.
The application area combining digital image process with face recognition technology is more and more extensive and embedded to people’s life at present. But as the present computer level, the ability of computer in outside apperception is still very weak, so it needs more manpower and more material resource to do research on the theories and applications of combining digital image processing with face recognition. In this paper, we discuss one of the applications of face recognition among image processing which is based on geometry feature method. This system includes several functions such as pre-processing of facial images、face figure diction、automatic location of the eyebrow、eye、nose and mouth,and recognition of face images.In the part of face diction, the method used in this thesis is face diction algorithm based on haar-like feature. It can provide with high accurate, meet the request for real time detection, use the library provided by OpenCV combining with cascade classifier to describe the face using a serial of haar-like features. In our researehes for facial features localization,the method of active shape model(ASM) was used in order to locate each characteristic part of face quickly.
This system will offer a new combinable point, which is that face recognition technology is applied in digital image processing, and expand the range of digital image processing application.
目前,数字图像处理与人脸识别技术相结合的应用范围越来越广,已经深入到人们的生活之中。但就目前计算机的应用水平而言,由于计算机对外部的感知能力还比较弱,自动化的处理能力还有待提高,仍然需要投入大量人力、物力来从事数字图像处理与人脸识别相结合的理论和应用的研究。
本课题论述了人脸识别技术在图像处理中的一个应用,采用几何特征的方法,包括人脸图像的预处理、人脸轮廓的检测、人脸图像中眉毛、眼睛、鼻子、嘴巴各部位的自动定位以及人脸图像的识别等几大功能。其中人脸检测部分,采用了基于haar-like特征的人脸检测算法,该算法在具有较高的检测率的同时也满足了实时检测的要求。算法利用OpenCV提供的库结合层叠式分类器并使用一系列haar-like特征来描述人脸。在人脸面部特征定位的研究中,采用了基于主动形状模型(ASM)的方法,能够很快的定位到人脸面部的各个特征部位。
本系统的出现为人脸识别技术在数字图像处理方面的应用寻找到了一个新的结合点,也使数字图像处理技术的应用范围更加广泛。
With the double-quick development of society and science technology, the field in computer application is also expanding as soon as possible.So digital image processing techonology has emerged as the times require as a bran-new image processing method.As the same time, it is also applied to more and more domains. Digital image processing is a method and technology which uses computer devices to convert image to digital informations in order to save,process,transmit and recur,and process image for wipe off yawp,enhance,recove,partition and extraction.Face recognition technology as a particular application of pattern recognition in image field is widely applied to surveillance and security, digital libraries,video meeting,and human-computer intelligent interaction and so on.
The application area combining digital image process with face recognition technology is more and more extensive and embedded to people’s life at present. But as the present computer level, the ability of computer in outside apperception is still very weak, so it needs more manpower and more material resource to do research on the theories and applications of combining digital image processing with face recognition. In this paper, we discuss one of the applications of face recognition among image processing which is based on geometry feature method. This system includes several functions such as pre-processing of facial images、face figure diction、automatic location of the eyebrow、eye、nose and mouth,and recognition of face images.In the part of face diction, the method used in this thesis is face diction algorithm based on haar-like feature. It can provide with high accurate, meet the request for real time detection, use the library provided by OpenCV combining with cascade classifier to describe the face using a serial of haar-like features. In our researehes for facial features localization,the method of active shape model(ASM) was used in order to locate each characteristic part of face quickly.
This system will offer a new combinable point, which is that face recognition technology is applied in digital image processing, and expand the range of digital image processing application.
引文
[1] Rafael C. Gonzalez, Richard E. Woods.数字图像处理.第二版.阮秋琦,阮宇智译.北京:电于工业出版社, 2003
[2]孙文军.人脸模型美化: [硕士学位论文].杭州:浙江大学图书馆, 2008
[3] Kevin I-J Ho, Tung-Shou Chen, and Hising-Yi Su. Automatic image processing system for beautifying human faces, 2002
[4]邹北骥,吕格峰,周浩宇等.人脸整形与美容虚拟手术系统的设计与实现,系统仿真学报(Journalof System Stimulation), 2003, 15(6): 898-901
[5]彭群生,鲍虎军,金小刚.计算机真实感图形的算法基础.北京:科学出版社, 1999
[6] Watabe H, Arakawa Y. Nonlinear Filters for Multimedia Applications. Proc IEEE ICIP’99, 1999(10): 174-179
[7] Bledsoe W W. Man-machine facial recognition. Palo Alto, CA: Panoramic Res In. 1966
[8]王伟,马建光.人脸识别常用方法及其发展现状.自动检测技术, 2002, 21(1): 45-46
[9]张翠平,苏光大.人脸识别技术综述.中国图像图形学报, 2000, 11(5): 65-66
[10]刘小军.人脸识别技术研究: [硕士学位论文].北京:中国科学院电子所, 2001
[11]罗晓辉.阂值化图像分割算法研究.科技信息, 2008(7): 185-186
[12]张毓晋.图像分割.北京:科学出版社, 2001
[13] Roberto Brunelli, Tomaso Paggio. Face recognition: features versus templates. IEEE Trans on PAMI, 1993, 15(10): 1042-1052
[14]山世光.人脸识别中若干关键问题的研究: [博士学位论文].北京:清华大学图书馆, 2004
[15]宋刚,徐光枯.基于特征的人脸验证: [硕士学位论文].北京:清华大学图书馆, 2001
[16]田原.人脸识别技术及其相关方法的研究: [博士学位论文].西安:西安交通大学图书馆, 1998
[17]李士进.人脸检测与识别方法研究: [博士学位论文].南京:南京理工大学图书馆, 2000
[18]刘明宝.人脸检测与跟踪的研究: [博士学位论文].哈尔滨:哈尔滨工业大学图书馆, 1997
[19]荆晓远.模式分类技术在人脸识别中的应用: [博士学位论文].南京:南京理工大学图书馆, 1995
[20]刘志镜,李夏忠,武芒.基于二维模型的多姿态人脸识别.西安电子科技大学报, 2004, 31(2): 208-222
[21]马艳,王金城.基于颜色与模板匹配的人脸检测方法: [硕士学位论文].大连:大连理工大学图书馆, 2006
[22]张春雨,陈绵书,陈贺新.人脸正面图像的机器识别.计算机工程与应用, 2004(2): 62-65
[23]郑庆,闽帆.基于肤色人脸检测与定位: [硕士学位论文].成都:电子科技大学图书馆, 2004
[24]交锋,高文.人脸图像识别算法研究.中国科学院计算技术研究所: [硕士学位论文].成都:电子科技大学图书馆, 2006
[25]洪子泉,杨静宇.用于图像识别的图像代数特征抽取.自动化学报, 1992, 18(2): 233-237
[26]焦峰,山世光,催国勤.基于局部特征分析的人脸识别方法.计算计与图形学学报, 2003, 15(1): 21-23
[27] G. W. Cottrell, M. Fleming, Face recognition using unsupervised feature extractio, 1990: 322-325
[28] Kin Man Lam, Hong Yan. Location and extraction the eye in human face images. Pattern Recognition, 1996, 29(5): 771-779
[29]林维训,潘纲,吴朝晖等.脸部特征定位方法综述.中国图像图形学报, 2003(8): 849-859
[30] Y. Moses, Y. Adini, S. Ullman. Face Recognition: the Problem of Compensationgfor Changes in Illumination. IEEE Transaction on PAMI, 1997, 19(7): 721-732
[31]鲁秀青.人脸的特征提取与查询: [硕士学位论文].济南:山东大学图书馆, 2001
[32]李华胜,杨桦,袁保宗.人脸识别系统中的特征提取.北方交通大学学报, 2001, 25(2): 45-46
[33]王光辉,梁毅军,贺朋令.人脸识别中的眼睛定位.信号处理, 1999(15): 22-26
[34]陶亮,庄镇泉.复杂背景下的自动人眼定位.计算机辅助设计与图形学学报, 2003, 15(1): 38-42
[35]周东生,张强,魏小鹏.人脸动画中语音可视化算法研究进展.计算机工程与应用, 2007, 43(9): 36-39
[36]李培华,张田文.主动轮廓线模型(蛇模型)综述.软件学报, 2000, 11(6): 751-757
[37]侯云舒,付中华,张艳宁.基于改进ASM的人脸特征点提取.计算机应用研究, 2006(11): 255-257
[38] Cootes T F, Taylor C J, Cooper D H, et al. Active shape models-their training and application. Computer Vision and Image Understanding, 1995, 61(1): 38-59
[39] T Cootes, G Edwards, C Taylor. Comparing active shape models with active appearance models[A], In: British Ma-chine Vision Conference[C], 1999, 173-182
[40]石林英.基于改进的ASM方法的人脸特征点检测: [硕士学位论文],北京:中山大学图书馆, 2005
[41]聂云飞,彭兰英.脸的美学特征.专业*信息BEAUTY.
[42] Tommer Leyvand, Danel Cohen-Or, Gideon Dror, Dani Lischinski. Digital Face Beautifieation, ACM SIGGRAPH, Technical Sketches, 2006
[43] Donald Heam, M. Pauline Baker. Computer Graphics with OpenGL. Third Edition,北京:电子工业出版社, 2004
[2]孙文军.人脸模型美化: [硕士学位论文].杭州:浙江大学图书馆, 2008
[3] Kevin I-J Ho, Tung-Shou Chen, and Hising-Yi Su. Automatic image processing system for beautifying human faces, 2002
[4]邹北骥,吕格峰,周浩宇等.人脸整形与美容虚拟手术系统的设计与实现,系统仿真学报(Journalof System Stimulation), 2003, 15(6): 898-901
[5]彭群生,鲍虎军,金小刚.计算机真实感图形的算法基础.北京:科学出版社, 1999
[6] Watabe H, Arakawa Y. Nonlinear Filters for Multimedia Applications. Proc IEEE ICIP’99, 1999(10): 174-179
[7] Bledsoe W W. Man-machine facial recognition. Palo Alto, CA: Panoramic Res In. 1966
[8]王伟,马建光.人脸识别常用方法及其发展现状.自动检测技术, 2002, 21(1): 45-46
[9]张翠平,苏光大.人脸识别技术综述.中国图像图形学报, 2000, 11(5): 65-66
[10]刘小军.人脸识别技术研究: [硕士学位论文].北京:中国科学院电子所, 2001
[11]罗晓辉.阂值化图像分割算法研究.科技信息, 2008(7): 185-186
[12]张毓晋.图像分割.北京:科学出版社, 2001
[13] Roberto Brunelli, Tomaso Paggio. Face recognition: features versus templates. IEEE Trans on PAMI, 1993, 15(10): 1042-1052
[14]山世光.人脸识别中若干关键问题的研究: [博士学位论文].北京:清华大学图书馆, 2004
[15]宋刚,徐光枯.基于特征的人脸验证: [硕士学位论文].北京:清华大学图书馆, 2001
[16]田原.人脸识别技术及其相关方法的研究: [博士学位论文].西安:西安交通大学图书馆, 1998
[17]李士进.人脸检测与识别方法研究: [博士学位论文].南京:南京理工大学图书馆, 2000
[18]刘明宝.人脸检测与跟踪的研究: [博士学位论文].哈尔滨:哈尔滨工业大学图书馆, 1997
[19]荆晓远.模式分类技术在人脸识别中的应用: [博士学位论文].南京:南京理工大学图书馆, 1995
[20]刘志镜,李夏忠,武芒.基于二维模型的多姿态人脸识别.西安电子科技大学报, 2004, 31(2): 208-222
[21]马艳,王金城.基于颜色与模板匹配的人脸检测方法: [硕士学位论文].大连:大连理工大学图书馆, 2006
[22]张春雨,陈绵书,陈贺新.人脸正面图像的机器识别.计算机工程与应用, 2004(2): 62-65
[23]郑庆,闽帆.基于肤色人脸检测与定位: [硕士学位论文].成都:电子科技大学图书馆, 2004
[24]交锋,高文.人脸图像识别算法研究.中国科学院计算技术研究所: [硕士学位论文].成都:电子科技大学图书馆, 2006
[25]洪子泉,杨静宇.用于图像识别的图像代数特征抽取.自动化学报, 1992, 18(2): 233-237
[26]焦峰,山世光,催国勤.基于局部特征分析的人脸识别方法.计算计与图形学学报, 2003, 15(1): 21-23
[27] G. W. Cottrell, M. Fleming, Face recognition using unsupervised feature extractio, 1990: 322-325
[28] Kin Man Lam, Hong Yan. Location and extraction the eye in human face images. Pattern Recognition, 1996, 29(5): 771-779
[29]林维训,潘纲,吴朝晖等.脸部特征定位方法综述.中国图像图形学报, 2003(8): 849-859
[30] Y. Moses, Y. Adini, S. Ullman. Face Recognition: the Problem of Compensationgfor Changes in Illumination. IEEE Transaction on PAMI, 1997, 19(7): 721-732
[31]鲁秀青.人脸的特征提取与查询: [硕士学位论文].济南:山东大学图书馆, 2001
[32]李华胜,杨桦,袁保宗.人脸识别系统中的特征提取.北方交通大学学报, 2001, 25(2): 45-46
[33]王光辉,梁毅军,贺朋令.人脸识别中的眼睛定位.信号处理, 1999(15): 22-26
[34]陶亮,庄镇泉.复杂背景下的自动人眼定位.计算机辅助设计与图形学学报, 2003, 15(1): 38-42
[35]周东生,张强,魏小鹏.人脸动画中语音可视化算法研究进展.计算机工程与应用, 2007, 43(9): 36-39
[36]李培华,张田文.主动轮廓线模型(蛇模型)综述.软件学报, 2000, 11(6): 751-757
[37]侯云舒,付中华,张艳宁.基于改进ASM的人脸特征点提取.计算机应用研究, 2006(11): 255-257
[38] Cootes T F, Taylor C J, Cooper D H, et al. Active shape models-their training and application. Computer Vision and Image Understanding, 1995, 61(1): 38-59
[39] T Cootes, G Edwards, C Taylor. Comparing active shape models with active appearance models[A], In: British Ma-chine Vision Conference[C], 1999, 173-182
[40]石林英.基于改进的ASM方法的人脸特征点检测: [硕士学位论文],北京:中山大学图书馆, 2005
[41]聂云飞,彭兰英.脸的美学特征.专业*信息BEAUTY.
[42] Tommer Leyvand, Danel Cohen-Or, Gideon Dror, Dani Lischinski. Digital Face Beautifieation, ACM SIGGRAPH, Technical Sketches, 2006
[43] Donald Heam, M. Pauline Baker. Computer Graphics with OpenGL. Third Edition,北京:电子工业出版社, 2004