多指指纹分割算法研究
|
摘要
指纹识别技术提供了高度安全的个人身份鉴别手段,具有重要的研究与应用价值,是生物识别研究的热点与重点内容之一。如何有效提高指纹识别的精度是指纹识别的核心任务,多指指纹融合识别可提高单一生物特征识别的精度,且无需增加采集成本和软件处理复杂度,从而多指指纹融合识别成为研究和应用的必然,而多指指纹分割是多指指纹融合识别的前提与基础。
根据多指指纹硬件采集与多指指纹图像特点,结合多指指纹识别系统的实际工程需要,本文系统研究了多指指纹分割算法,提出了一种基于方向投影的多指指纹分割算法。该方法首先采用双线性插值算法、中值滤波处理、canny边缘检测等预处理过程,降低了图像畸变、采集环境不同程度污染、背景光照不同等方面的影响;然后利用X与Y方向投影和Otsu阈值分割算法,实现了主方向获取,然后利用每个手指的垂直主方向投影与阈值化,获取每个手指指纹的坐标范围和指纹类型,实现多指指纹分割;最后利用指纹面积、总体灰度均值和灰度方差、块内的灰度均值和灰度方差作为指纹图像质量评价的主要指标,实现分割后指纹图像质量评价,判断是非满足要求。实验过程,将上述算法利用VC 6.0和Matlab进行系统实现和仿真试验,实验结果表明该方法足有效的,能满足实际分割识别的需要。本文的研究可为多指指纹分割提供参考与借鉴。
With the increasing requirements for security, biometrics based personal identification methods have been received extensive attention. Recently, Fingerprint recognition technology is becoming an active topic in biometrics due to its high reliability for personal identification, which can overcome the drawbacks of the traditional identification and widely used in the customs, airports, and other special high-security channel region. Accordingly, Fingerprint recognition technology has the important value of the study and application, and is an active topic in biometrics recognition.
How to effectively improve the accuracy of fingerprint recognition is the core of fingerprint recognition. Multi-biometric fusion identification can effectively improve the accuracy of a single model biometric recognition, and multi-fingerprint recognition can improve the accuracy of a single fingerprint recognition, which wills not additional acquisition cost and software complexity of treatment, so this will become an important application in future. As the multi-fingerprint segmentation method is the foundation of multi-fingerprint recognition technology, according to the characteristics of fingerprint images and fingerprint acquisition hardware, the multi-fingerprint segmentation method would be studied by actual works required, and a new multi-finger fingerprint segmentation algorithm based on direction projection is provided.
The fingerprint physiological characteristics, fingerprint acquisition hardware system, acquisition requirements, and characteristics of fingerprint images, which are foundation of fingerprint segmentation and provide the priori knowledge to design the segmentation algorithm, are introduced in the paper. Based on the above characteristics, the design, including image correction, image filtering processing, edge detection, the direction of projection, the main direction of acquisition and vertical projection, to determine the type of fingerprint segmentation and image quality evaluation part of the multi-finger fingerprint technology route, which includes pre-processing stage image correction, image filtering processing, edge detection process, are separated with the basic premise; separated stages, including the direction of projection, the main direction of acquisition and vertical projection, fingerprint segmentation process and the type of judge, etc., are the core of the algorithm, the fingerprint image quality evaluation After the partition of the fingerprint on a qualitative assessment, in order to enhance image recognition accuracy of the latter to lay the foundation. The above treatment refers to the process for multi-fingerprint segmentation provides an important technical support.
Preprocess of multi-finger fingerprint fingerprints are many segmentation refers to the premise and foundation, and its concrete realization methods include bilinear interpolation algorithm, median filtering treatment, canny edge detection processes, bilinear interpolation algorithm will be collected in 2940×1840 of the original image data standardization for the 1812 * 1476 new images, reduces the effects of light aberration; median filter processing algorithms to reduce image noise and other interference; Canny edge detection to obtain fingerprint refers to the brink of many features, provide the basis for the direction of projection reduces the image acquisition process, the background light is different from the pressure difference, such as finger implications. Therefore, some pretreatment reduces image distortion, collecting environmental pollution to varying degrees, the background illumination of different areas such as the impact of many means for follow-up fingerprint image segmentation has laid a foundation.
Multi-fingerprint segmentation algorithms are the core of this article and provide a basis for later identification. Multi-fingerprint segmentation process includes X direction's projection, Y direction's projection, threshold of X direction's projection, threshold of Y direction's projection, the main direction acquisition, each finger perpendicular to the main direction of projection and threshold, multi-fingerprint image segmentation type and fingerprint identification process. X direction of projection statistics of its vertical edge points of the number of fingerprints to obtain horizontal direction refers to many fingerprint projection distribution; Otsu threshold segmentation algorithm is used by X direction of projection threshold, the realization of the adaptive threshold selection, access to the horizontal direction refers to many fingerprints on the distribution of projection area, then use the projector with the merger of the region bound algorithm to obtain the horizontal direction on many of the projection distribution refers to the number of fingerprints and fingerprint of each of the space coordinates of the scope of the projector; Y direction of projection statistics of its horizontal direction the number of edge points fingerprint, access to many means a vertical projection of the distribution of fingerprints; Otsu threshold segmentation algorithm is used by Y direction projection threshold, the realization of the adaptive threshold selection, access to many means a vertical projection of the distribution of the fingerprint region , then the region ruled out the use of projection algorithm and combined to obtain the vertical projection of the many means the distribution of the number of fingerprints and fingerprint of each projection space coordinates of reference; X and Y direction in accordance with the distribution of projection, combined with many referring to the characteristics of fingerprint images, many means to obtain the main direction of the fingerprint image, and then statistics for each finger perpendicular to the main direction of the projection distribution, using Otsu threshold segmentation algorithm and projection region bound algorithm with the merger of its main direction of the vertical projection of the distribution of thresholds, access to each fingerprints fingers the scope of the space coordinates. According to the characteristics of multi-fingerprint images, each fingerprint image types are recognized automatically and automatic segmentation process of multi-fingerprints are completed.
Fingerprint image quality evaluation of the partition after the treatment in the identification before the image quality for pre-evaluation to determine whether they meet the necessary requirements of image recognition, which can effectively improve the accuracy of fingerprint identification. In this paper, the fingerprint image of the fingerprint area and the overall mean gray level and gray-scale variance, block the gray mean and variance of gray-scale fingerprint image quality evaluation of key indicators, after the realization of fingerprint image segmentation quality evaluation, to judge whether meet the requirements, if it is necessary to meet the identification or not, they are required to re-acquisition.
Using VC 6.0 developed many fingerprint segmentation refers to the experimental system, combined with the Matlab simulation analysis, collected for different circumstances, for many online and offline fingerprint segmentation test, experimental results show that this article refers to many fingerprint segmentation method are effective and can meet the needs of the actual partition identification. Therefore, this study refers to many fingerprints can be divided to provide a sense of reference.
根据多指指纹硬件采集与多指指纹图像特点,结合多指指纹识别系统的实际工程需要,本文系统研究了多指指纹分割算法,提出了一种基于方向投影的多指指纹分割算法。该方法首先采用双线性插值算法、中值滤波处理、canny边缘检测等预处理过程,降低了图像畸变、采集环境不同程度污染、背景光照不同等方面的影响;然后利用X与Y方向投影和Otsu阈值分割算法,实现了主方向获取,然后利用每个手指的垂直主方向投影与阈值化,获取每个手指指纹的坐标范围和指纹类型,实现多指指纹分割;最后利用指纹面积、总体灰度均值和灰度方差、块内的灰度均值和灰度方差作为指纹图像质量评价的主要指标,实现分割后指纹图像质量评价,判断是非满足要求。实验过程,将上述算法利用VC 6.0和Matlab进行系统实现和仿真试验,实验结果表明该方法足有效的,能满足实际分割识别的需要。本文的研究可为多指指纹分割提供参考与借鉴。
With the increasing requirements for security, biometrics based personal identification methods have been received extensive attention. Recently, Fingerprint recognition technology is becoming an active topic in biometrics due to its high reliability for personal identification, which can overcome the drawbacks of the traditional identification and widely used in the customs, airports, and other special high-security channel region. Accordingly, Fingerprint recognition technology has the important value of the study and application, and is an active topic in biometrics recognition.
How to effectively improve the accuracy of fingerprint recognition is the core of fingerprint recognition. Multi-biometric fusion identification can effectively improve the accuracy of a single model biometric recognition, and multi-fingerprint recognition can improve the accuracy of a single fingerprint recognition, which wills not additional acquisition cost and software complexity of treatment, so this will become an important application in future. As the multi-fingerprint segmentation method is the foundation of multi-fingerprint recognition technology, according to the characteristics of fingerprint images and fingerprint acquisition hardware, the multi-fingerprint segmentation method would be studied by actual works required, and a new multi-finger fingerprint segmentation algorithm based on direction projection is provided.
The fingerprint physiological characteristics, fingerprint acquisition hardware system, acquisition requirements, and characteristics of fingerprint images, which are foundation of fingerprint segmentation and provide the priori knowledge to design the segmentation algorithm, are introduced in the paper. Based on the above characteristics, the design, including image correction, image filtering processing, edge detection, the direction of projection, the main direction of acquisition and vertical projection, to determine the type of fingerprint segmentation and image quality evaluation part of the multi-finger fingerprint technology route, which includes pre-processing stage image correction, image filtering processing, edge detection process, are separated with the basic premise; separated stages, including the direction of projection, the main direction of acquisition and vertical projection, fingerprint segmentation process and the type of judge, etc., are the core of the algorithm, the fingerprint image quality evaluation After the partition of the fingerprint on a qualitative assessment, in order to enhance image recognition accuracy of the latter to lay the foundation. The above treatment refers to the process for multi-fingerprint segmentation provides an important technical support.
Preprocess of multi-finger fingerprint fingerprints are many segmentation refers to the premise and foundation, and its concrete realization methods include bilinear interpolation algorithm, median filtering treatment, canny edge detection processes, bilinear interpolation algorithm will be collected in 2940×1840 of the original image data standardization for the 1812 * 1476 new images, reduces the effects of light aberration; median filter processing algorithms to reduce image noise and other interference; Canny edge detection to obtain fingerprint refers to the brink of many features, provide the basis for the direction of projection reduces the image acquisition process, the background light is different from the pressure difference, such as finger implications. Therefore, some pretreatment reduces image distortion, collecting environmental pollution to varying degrees, the background illumination of different areas such as the impact of many means for follow-up fingerprint image segmentation has laid a foundation.
Multi-fingerprint segmentation algorithms are the core of this article and provide a basis for later identification. Multi-fingerprint segmentation process includes X direction's projection, Y direction's projection, threshold of X direction's projection, threshold of Y direction's projection, the main direction acquisition, each finger perpendicular to the main direction of projection and threshold, multi-fingerprint image segmentation type and fingerprint identification process. X direction of projection statistics of its vertical edge points of the number of fingerprints to obtain horizontal direction refers to many fingerprint projection distribution; Otsu threshold segmentation algorithm is used by X direction of projection threshold, the realization of the adaptive threshold selection, access to the horizontal direction refers to many fingerprints on the distribution of projection area, then use the projector with the merger of the region bound algorithm to obtain the horizontal direction on many of the projection distribution refers to the number of fingerprints and fingerprint of each of the space coordinates of the scope of the projector; Y direction of projection statistics of its horizontal direction the number of edge points fingerprint, access to many means a vertical projection of the distribution of fingerprints; Otsu threshold segmentation algorithm is used by Y direction projection threshold, the realization of the adaptive threshold selection, access to many means a vertical projection of the distribution of the fingerprint region , then the region ruled out the use of projection algorithm and combined to obtain the vertical projection of the many means the distribution of the number of fingerprints and fingerprint of each projection space coordinates of reference; X and Y direction in accordance with the distribution of projection, combined with many referring to the characteristics of fingerprint images, many means to obtain the main direction of the fingerprint image, and then statistics for each finger perpendicular to the main direction of the projection distribution, using Otsu threshold segmentation algorithm and projection region bound algorithm with the merger of its main direction of the vertical projection of the distribution of thresholds, access to each fingerprints fingers the scope of the space coordinates. According to the characteristics of multi-fingerprint images, each fingerprint image types are recognized automatically and automatic segmentation process of multi-fingerprints are completed.
Fingerprint image quality evaluation of the partition after the treatment in the identification before the image quality for pre-evaluation to determine whether they meet the necessary requirements of image recognition, which can effectively improve the accuracy of fingerprint identification. In this paper, the fingerprint image of the fingerprint area and the overall mean gray level and gray-scale variance, block the gray mean and variance of gray-scale fingerprint image quality evaluation of key indicators, after the realization of fingerprint image segmentation quality evaluation, to judge whether meet the requirements, if it is necessary to meet the identification or not, they are required to re-acquisition.
Using VC 6.0 developed many fingerprint segmentation refers to the experimental system, combined with the Matlab simulation analysis, collected for different circumstances, for many online and offline fingerprint segmentation test, experimental results show that this article refers to many fingerprint segmentation method are effective and can meet the needs of the actual partition identification. Therefore, this study refers to many fingerprints can be divided to provide a sense of reference.
引文
[1]王阳生.中国生物特征识别发展重大进展[J].中国安防产品信息,2005(02):36-36.
[2]International Biometric Group.Biometrics Market and Industry Report 2007-2012 2007.
[3]孙冬梅,裘正定.生物特征识别技术综述[J].电子学报.2001,(S1):1744-1748.
[4]景英娟,董育宁.生物特征识别技术综述[J].桂林电子工业学院学报.2005(02):27-32.
[5]明星.虹膜识别技术中小波变换的应用原理与方法:吉林大学;2006.
[6]Anon.Partial fingerprint pattern classification[J].Journal of the Institution of Electronics and Telecommunication Engineers.1989,35(1):28-33.
[7]Moayer Bijan,Fu King-Sun.TREE SYSTEM APPROACH FOR FINGERPRINT PATTERN RECOGNITION[J].Ieee Transactions on Pattern Analysis and Machine Intelligence.1986,PAMI-8(3):376-387.
[8]Simmleit Norbert,Schulten Hans-RolE Pattern recognition of spruce trees.An integrated,analytical approach to forest damage[J].Environmental Science and Technology.1989,23(8):1000-1006.
[9]Hrechak Andrew K.,McHugh James A.Automated fingerprint recognition using structural matching[J].Pattern Recognition.1990,23(8):893-904.
[10]Wilkinson T.D.,Mears R.J.Compact optical correlator using silicon backplane FLC SLMs for fingerprint recognition[J].Ferroelectrics.1996,181(1-4 pt 4):47-52.
[11]Chong Michael M.S.,Ngee Tan Han,Jun Liu.Geometric framework for fingerprint image classification[J].Pattern Recognition.1997,30(9):1475-1488.
[12]Germain Robert S.,Califano Andrea,Colville Scott.Fingerprint matching using transformation parameter clustering[J].IEEE Computational Science &Engineering.1997,4(4):42-49.
[13]Klima Milos,Rott Jiri,Dvorak Pavel.Model of 2D optical correlator for fingerprint identification[J].IEEE Aerospace and Electronic Systems Magazine.1997,12(7):3-9.
[14]Maio Dario,Maltoni Davide.Direct gray-scale minutiae detection in fingerprints[J].Ieee Transactions on Pattern Analysis and Machine Intelligence. 1997,19(1):27-40.
[15]Roddy Andrea R.,Stosz Jonathan D.Fingerprint features-statistical analysis and system performance estimates[J].Proceedings of the IEEE.1997,85(9):1390-1421.
[16]Rowe Robert K.,Nixon Kristin A.,Corcoran Stephen P.Multispectral fingerprint biometrics.In;2005;Piscataway,NJ 08855-1331,United States:Institute of Electrical and Electronics Engineers Computer Society;2005.p.14-20.
[17]Rowe Robert K.,Uludag Umut,Demirkus Meltem.A multispectral whole-hand biometric authentication system.In;2007;Piscataway,NJ 08855-1331,United States:Institute of Electrical and Electronics Engineers Computer Society;2007.p.4430532.
[18]Slot Krzysztof,Gozdzik Marek.Fingerprint alignment with deformable templates.In;2008;Piscataway,NJ 08855-1331,United States:Institute of Electrical and Electronics Engineers Computer Society;2008.p.395-398.
[19]祝恩,殷建平,胡春风等.指纹纹路方向计算与图像分割的系统方法[J].同济大学学报(自然科学版).2007(05).
[20]Yang Jianwei,Liu Lifeng,Jiang Tianzi.A modified Gabor filter design method for fingerprint image enhancement[J].Pattern Recognition Letters.2003,24(12):1805-1817.
[21]Zhang Weipeng,Van Tang Yuan,You Xinge.Fingerprint enhancement using wavelet transform combined with gabor filter[J].International Journal of Pattern Recognition and Artificial Intelligence.2004,18(8):1391-1406.
[22]Wang Wei,Li Jianwei,Huang Feifei.Design and implementation of Log-Gabor filter in fingerprint image enhancement[J].Pattern Recognition Letters.2008,29(3):301-308.
[23]Delcorte A.,Bertrand P.Energy distributions of hydrocarbon secondary ions from thin organic films under keV ion bombardment:correlation between kinetic and formation energy of ions sputtered from tricosenoic acid[J].Nuclear Instruments & Methods in Physics Research,Section B:Beam Interactions with Materials and Atoms.1996,117(3):235-242.
[24]Ding Y.,Clarkson T.G.Fingerprint recognition by pRAM neural networks[J].Neural Network World.1996,6(4):535-543.
[25]Fitz A.P.,Green R.J.Fingerprint classification using a hexagonal fast Fourier transform[J].Pattern Recognition.1996,29(10):1587-1597.
[26]Halici Ugur,Ongun Guclu.Fingerprint classification through self-organizing feature maps modified to treat uncertainties[J]. Proceedings of the IEEE. 1996, 84(10): 1497-1512.
[27] Ratha Nalini K., Karu Kalle, Chen Shaoyun. Real-time matching system for large fingerprint databases[J]. Ieee Transactions on Pattern Analysis and Machine Intelligence. 1996, 18(8): 799-813.
[28] Whiteley James R., Davis James F., Mehrotra Amit. Observations and problems applying ART2 for dynamic sensor pattern interpretation[J]. IEEE Transactions on Systems, Man, and Cybernetics Part A:Systems and Humans. 1996, 26(4): 423-437.
[29] Jain Anil, Hong Lin, Bolle Ruud. On-line fingerprint verification[J]. Ieee Transactions on Pattern Analysis and Machine Intelligence. 1997, 19(4): 302-314.
[30] Ratha Nalini K., Connell Jonathan H., Bolle Ruud M. Biometrics break-ins and band-aids[J]. Pattern Recognition Letters. 2003, 24(13): 2105-2113.
[31] Jain Anil K., Ratha Nalini K., Lakshmanan Sridhar. Object detection using Gabor filters[J]. Pattern Recognition. 1997, 30(2): 295-309.
[32] Isenor D. K., Zaky S. G. FINGERPRINT IDENTIFICATION USING GRAPH MATCHING[J]. Pattern Recognition. 1986, 19(2): 113-122.
[33] Wang Anna, Chen Yu, Zhang Xinhua. Iris recognition based on 2D wavelet and adaboost neural network[J]. Lecture Notes in Electrical Engineering. 2008, 6: 117-128.
[34] Besbes F., Trichili H., Solaiman B. Multimodal Biometric System Based on Fingerprint Identification and Iris Recognition[J]. 2008 3rd International Conference on Information and Communication Technologies: From Theory to Applications, Vols 1-5.2008: 1224-1228.
[35] Anon. Facial recognition - Part 1[J]. Biometric Technology Today. 2003, 11(11): 10-11.
[36] Abate Andrea F., Nappi Michele, Riccio Daniel. 2D and 3D face recognition: A survey[J]. Pattern Recognition Letters. 2007, 28(14): 1885-1906.
[37] Lee Kangrok, Park Kang Ryoung, Jang Jain. A study on multi-unit fingerprint verification. In; 2005; Heidelberg, D-69121, Germany: Springer Verlag; 2005. p. 141-150.
[2]International Biometric Group.Biometrics Market and Industry Report 2007-2012 2007.
[3]孙冬梅,裘正定.生物特征识别技术综述[J].电子学报.2001,(S1):1744-1748.
[4]景英娟,董育宁.生物特征识别技术综述[J].桂林电子工业学院学报.2005(02):27-32.
[5]明星.虹膜识别技术中小波变换的应用原理与方法:吉林大学;2006.
[6]Anon.Partial fingerprint pattern classification[J].Journal of the Institution of Electronics and Telecommunication Engineers.1989,35(1):28-33.
[7]Moayer Bijan,Fu King-Sun.TREE SYSTEM APPROACH FOR FINGERPRINT PATTERN RECOGNITION[J].Ieee Transactions on Pattern Analysis and Machine Intelligence.1986,PAMI-8(3):376-387.
[8]Simmleit Norbert,Schulten Hans-RolE Pattern recognition of spruce trees.An integrated,analytical approach to forest damage[J].Environmental Science and Technology.1989,23(8):1000-1006.
[9]Hrechak Andrew K.,McHugh James A.Automated fingerprint recognition using structural matching[J].Pattern Recognition.1990,23(8):893-904.
[10]Wilkinson T.D.,Mears R.J.Compact optical correlator using silicon backplane FLC SLMs for fingerprint recognition[J].Ferroelectrics.1996,181(1-4 pt 4):47-52.
[11]Chong Michael M.S.,Ngee Tan Han,Jun Liu.Geometric framework for fingerprint image classification[J].Pattern Recognition.1997,30(9):1475-1488.
[12]Germain Robert S.,Califano Andrea,Colville Scott.Fingerprint matching using transformation parameter clustering[J].IEEE Computational Science &Engineering.1997,4(4):42-49.
[13]Klima Milos,Rott Jiri,Dvorak Pavel.Model of 2D optical correlator for fingerprint identification[J].IEEE Aerospace and Electronic Systems Magazine.1997,12(7):3-9.
[14]Maio Dario,Maltoni Davide.Direct gray-scale minutiae detection in fingerprints[J].Ieee Transactions on Pattern Analysis and Machine Intelligence. 1997,19(1):27-40.
[15]Roddy Andrea R.,Stosz Jonathan D.Fingerprint features-statistical analysis and system performance estimates[J].Proceedings of the IEEE.1997,85(9):1390-1421.
[16]Rowe Robert K.,Nixon Kristin A.,Corcoran Stephen P.Multispectral fingerprint biometrics.In;2005;Piscataway,NJ 08855-1331,United States:Institute of Electrical and Electronics Engineers Computer Society;2005.p.14-20.
[17]Rowe Robert K.,Uludag Umut,Demirkus Meltem.A multispectral whole-hand biometric authentication system.In;2007;Piscataway,NJ 08855-1331,United States:Institute of Electrical and Electronics Engineers Computer Society;2007.p.4430532.
[18]Slot Krzysztof,Gozdzik Marek.Fingerprint alignment with deformable templates.In;2008;Piscataway,NJ 08855-1331,United States:Institute of Electrical and Electronics Engineers Computer Society;2008.p.395-398.
[19]祝恩,殷建平,胡春风等.指纹纹路方向计算与图像分割的系统方法[J].同济大学学报(自然科学版).2007(05).
[20]Yang Jianwei,Liu Lifeng,Jiang Tianzi.A modified Gabor filter design method for fingerprint image enhancement[J].Pattern Recognition Letters.2003,24(12):1805-1817.
[21]Zhang Weipeng,Van Tang Yuan,You Xinge.Fingerprint enhancement using wavelet transform combined with gabor filter[J].International Journal of Pattern Recognition and Artificial Intelligence.2004,18(8):1391-1406.
[22]Wang Wei,Li Jianwei,Huang Feifei.Design and implementation of Log-Gabor filter in fingerprint image enhancement[J].Pattern Recognition Letters.2008,29(3):301-308.
[23]Delcorte A.,Bertrand P.Energy distributions of hydrocarbon secondary ions from thin organic films under keV ion bombardment:correlation between kinetic and formation energy of ions sputtered from tricosenoic acid[J].Nuclear Instruments & Methods in Physics Research,Section B:Beam Interactions with Materials and Atoms.1996,117(3):235-242.
[24]Ding Y.,Clarkson T.G.Fingerprint recognition by pRAM neural networks[J].Neural Network World.1996,6(4):535-543.
[25]Fitz A.P.,Green R.J.Fingerprint classification using a hexagonal fast Fourier transform[J].Pattern Recognition.1996,29(10):1587-1597.
[26]Halici Ugur,Ongun Guclu.Fingerprint classification through self-organizing feature maps modified to treat uncertainties[J]. Proceedings of the IEEE. 1996, 84(10): 1497-1512.
[27] Ratha Nalini K., Karu Kalle, Chen Shaoyun. Real-time matching system for large fingerprint databases[J]. Ieee Transactions on Pattern Analysis and Machine Intelligence. 1996, 18(8): 799-813.
[28] Whiteley James R., Davis James F., Mehrotra Amit. Observations and problems applying ART2 for dynamic sensor pattern interpretation[J]. IEEE Transactions on Systems, Man, and Cybernetics Part A:Systems and Humans. 1996, 26(4): 423-437.
[29] Jain Anil, Hong Lin, Bolle Ruud. On-line fingerprint verification[J]. Ieee Transactions on Pattern Analysis and Machine Intelligence. 1997, 19(4): 302-314.
[30] Ratha Nalini K., Connell Jonathan H., Bolle Ruud M. Biometrics break-ins and band-aids[J]. Pattern Recognition Letters. 2003, 24(13): 2105-2113.
[31] Jain Anil K., Ratha Nalini K., Lakshmanan Sridhar. Object detection using Gabor filters[J]. Pattern Recognition. 1997, 30(2): 295-309.
[32] Isenor D. K., Zaky S. G. FINGERPRINT IDENTIFICATION USING GRAPH MATCHING[J]. Pattern Recognition. 1986, 19(2): 113-122.
[33] Wang Anna, Chen Yu, Zhang Xinhua. Iris recognition based on 2D wavelet and adaboost neural network[J]. Lecture Notes in Electrical Engineering. 2008, 6: 117-128.
[34] Besbes F., Trichili H., Solaiman B. Multimodal Biometric System Based on Fingerprint Identification and Iris Recognition[J]. 2008 3rd International Conference on Information and Communication Technologies: From Theory to Applications, Vols 1-5.2008: 1224-1228.
[35] Anon. Facial recognition - Part 1[J]. Biometric Technology Today. 2003, 11(11): 10-11.
[36] Abate Andrea F., Nappi Michele, Riccio Daniel. 2D and 3D face recognition: A survey[J]. Pattern Recognition Letters. 2007, 28(14): 1885-1906.
[37] Lee Kangrok, Park Kang Ryoung, Jang Jain. A study on multi-unit fingerprint verification. In; 2005; Heidelberg, D-69121, Germany: Springer Verlag; 2005. p. 141-150.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |