基于MBF200指纹图像的识别算法研究
|
摘要
本文以MBF200指纹采集芯片所获取的指纹图像为研究对象，针对当前自动指纹识别技术中存在的问题，重点研究MBF200指纹图像的识别算法，目的是提高自动指纹识别系统的识别率和精确度，以最终开发自主的指纹识别产品。本文的研究工作对当前蓬勃发展的生物特征识别领域，具有重要现实意义和参考价值。
本文在阅读和分析国内外大量相关文献的基础上，对指纹预处理和指纹特征匹配方面进行深入的研究。首先研究和分析MBF200指纹图像的特征，总结出该类型图像的三个特点。针对这三个特点，基于灰度均衡法的指纹预处理方案被提出，以此来解决自动指纹识别系统中的指纹图像质量问题。其中的指纹分割算法，应用灰度均衡化取代传统的灰度归一化，改进灰度归一化对MBF200指纹图像分割阈值的非确定性。同时，灰度均衡法的引入改变了传统Gabor滤波增强的步骤，以适应MBF200指纹图像的增强；在中心点定位方面，改进基于方向场的块垂直累积中心点检测法，搜索各行累积值取代搜索各列累积值以得到最大累积向量，通过确定该向量中的最大与次大峰值点的关系来正确定位中心点，从而改进以往只通过确定最大值对中心点检测的非准确性。
指纹预处理后，本文采用基于细节特征的匹配算法来匹配指纹图像。通过指纹预处理中正确的中心点位，解决指纹匹配中的平移问题；利用特征点在极坐标下的转化，解决指纹图像匹配的旋转问题；借鉴可变最小限界盒思想，本文提出线性等级可变限界盒方法，解决指纹匹配的弹性扭曲问题。通过对指纹识别技术中存在问题的解决，从而提高自动指纹识别系统的识别率和精确度。
最后，应用Microsoft Visual C++6.0开发平台，设计和研发出自动指纹识别系统软件，对本研究室自行设计的MBF200指纹采集器所获取的指纹图像进行测试和评估，进而验证本文识别算法的有效性。
To aimed at some problems which exist in the current automation fingerprint identification technology, this paper research main the identification algorithm with the MBF200 fingerprint images to be an object of study The purpose is to improve systematic recognition rate and precision of AFIS, to develop the fingerprint identification product acting on selfs own ultimately. The research works of this paper to currently vigorous characteristic identification field has important practical significance and consults value.
This paper is carried out thorough research in the respect of the fingerprint preprocessing and the fingerprint characteristic matching on the basis reading and analyzing home and abroad large amount of interrelated documents. The research and analyses of the MBF200 fingerprint images firstly, this paper sums up three characters of the type images. To aim at these, the scheme based on gray balance fingerprint preprocessing is proposed to resolve the quality problems of fingerprint images in AFIS. In the aspect of fingerprint segmentation, this paper uses gray balance instead of the, traditional gray normalizing, in order to improving the certainness of threshold in the images segmentation. Meanwhile, it changes the steps of traditional Gabor filter enhancement to adapt MBF200 fingerprint images. In the localization of central point, the paper improves the inspecting central point based on the field of directions block vertical accumulation, and starches the accumulating value of each rows instead of ranks for getting the maximum accumulation vector, and detects the area of central point by checking the relation of maximum and sub maximum peak point in the vector, thereby gets better accuracy of the central point detection than that was only by ascertaining maximum value in the past.
After fingerprint preprocessing, this paper adopt the algorithm based on detail characteristic matching. By the correctly central point location, the fingerprint translatory
problem is resolved in matching. The image changing is also resolved by making use of the fingerprint characteristic point transform in polar coordinates. Reference by minimal changeable boundary box thought, this paper proposes the linearity grade of changeable boundary box method, which is resolved the fingerprint matching elasticity warping problem. By solving having the problems to the fingerprint identification technology, the systematic recognition rate and precision of automation fingerprint identification are improved thereby.
Finally, applying Microsoft Visual C++6.0 exploitation platform, the automation fingerprint identification system software is designed and studied. The images are carried out the validity testing and distinguishing appraising, which gained by the MBF200 fingerprint collection station that this labs designs independently, in order to verify the identification algorithmic by this paper proposed validity.
本文在阅读和分析国内外大量相关文献的基础上，对指纹预处理和指纹特征匹配方面进行深入的研究。首先研究和分析MBF200指纹图像的特征，总结出该类型图像的三个特点。针对这三个特点，基于灰度均衡法的指纹预处理方案被提出，以此来解决自动指纹识别系统中的指纹图像质量问题。其中的指纹分割算法，应用灰度均衡化取代传统的灰度归一化，改进灰度归一化对MBF200指纹图像分割阈值的非确定性。同时，灰度均衡法的引入改变了传统Gabor滤波增强的步骤，以适应MBF200指纹图像的增强；在中心点定位方面，改进基于方向场的块垂直累积中心点检测法，搜索各行累积值取代搜索各列累积值以得到最大累积向量，通过确定该向量中的最大与次大峰值点的关系来正确定位中心点，从而改进以往只通过确定最大值对中心点检测的非准确性。
指纹预处理后，本文采用基于细节特征的匹配算法来匹配指纹图像。通过指纹预处理中正确的中心点位，解决指纹匹配中的平移问题；利用特征点在极坐标下的转化，解决指纹图像匹配的旋转问题；借鉴可变最小限界盒思想，本文提出线性等级可变限界盒方法，解决指纹匹配的弹性扭曲问题。通过对指纹识别技术中存在问题的解决，从而提高自动指纹识别系统的识别率和精确度。
最后，应用Microsoft Visual C++6.0开发平台，设计和研发出自动指纹识别系统软件，对本研究室自行设计的MBF200指纹采集器所获取的指纹图像进行测试和评估，进而验证本文识别算法的有效性。
To aimed at some problems which exist in the current automation fingerprint identification technology, this paper research main the identification algorithm with the MBF200 fingerprint images to be an object of study The purpose is to improve systematic recognition rate and precision of AFIS, to develop the fingerprint identification product acting on selfs own ultimately. The research works of this paper to currently vigorous characteristic identification field has important practical significance and consults value.
This paper is carried out thorough research in the respect of the fingerprint preprocessing and the fingerprint characteristic matching on the basis reading and analyzing home and abroad large amount of interrelated documents. The research and analyses of the MBF200 fingerprint images firstly, this paper sums up three characters of the type images. To aim at these, the scheme based on gray balance fingerprint preprocessing is proposed to resolve the quality problems of fingerprint images in AFIS. In the aspect of fingerprint segmentation, this paper uses gray balance instead of the, traditional gray normalizing, in order to improving the certainness of threshold in the images segmentation. Meanwhile, it changes the steps of traditional Gabor filter enhancement to adapt MBF200 fingerprint images. In the localization of central point, the paper improves the inspecting central point based on the field of directions block vertical accumulation, and starches the accumulating value of each rows instead of ranks for getting the maximum accumulation vector, and detects the area of central point by checking the relation of maximum and sub maximum peak point in the vector, thereby gets better accuracy of the central point detection than that was only by ascertaining maximum value in the past.
After fingerprint preprocessing, this paper adopt the algorithm based on detail characteristic matching. By the correctly central point location, the fingerprint translatory
problem is resolved in matching. The image changing is also resolved by making use of the fingerprint characteristic point transform in polar coordinates. Reference by minimal changeable boundary box thought, this paper proposes the linearity grade of changeable boundary box method, which is resolved the fingerprint matching elasticity warping problem. By solving having the problems to the fingerprint identification technology, the systematic recognition rate and precision of automation fingerprint identification are improved thereby.
Finally, applying Microsoft Visual C++6.0 exploitation platform, the automation fingerprint identification system software is designed and studied. The images are carried out the validity testing and distinguishing appraising, which gained by the MBF200 fingerprint collection station that this labs designs independently, in order to verify the identification algorithmic by this paper proposed validity.
引文
[1] 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，2005，25(2)：27-32．
[2] Uludag U, Pankanti S, Prabhakar S, JainA K. Biometric cryptosystems: issues and challenges[J]. Proceedings of the IEEE, 2004, 92 (6): 948-960.
[3] 罗棻．生物识别技术的发展展望[J]．重庆工商大学学报(自然科学版)，2005，22(3)：249-252．
[4] Jain A K, Ross A, Prabhakar S. An introduction to biometric recognition[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2004, 14 (1):4-20.
[5] 张海春，回文博，林立忠．指纹识别技术研究进展[J]．石家庄学院学报，2005，7(3)：25-28．
[6] 段少雄，张堂辉，田捷．指纹识别技术漫谈[EB/OL]. http://industry.ccidnet.com/art/1101/20011210/6358_1.html, 2005-11-21.
[7] 詹小四．生物认证-人类未来身份认证的主要方式[J]．阜阳师范学院学报(自然科学版)，2004，21(2)：52-56．
[8] 霍宏涛．数字图像处理[M]．北京：北京理工大学出版社，2002．9．
[9] 于秀霞．指纹识别技术及其应用[J]．长春大学学报，2005，15(2)：30-32．
[10] 吕知辛，李兴江．指纹识别系统中的模式识别技术[J]．山东师范大学学报(自然科学版)，2004，19(4)：20-23．
[11] 李徐周．自动指纹识别技术的应用[J]．山东省青年管理干部学院学报，2005，2：143-144．
[12] 王崇文．自动指纹识别方法研究[D]．重庆：重庆大学，2002．
[13] 李见为．指纹识别技术的研究和发展[J]．重庆职业技术学院学报，2004，13(2)：4-6．
[14] 王卫东，平西建．自动指纹识别技术研究[J]．信息工程大学学报，2005，6(2)：47-50．
[15] 刘欣．基于图像内容的指纹识别系统的研究及其DSP实现[D]．陕西：西安理工大学，2004．
[16] 张成，周媛媛，林嘉宇．指纹采集技术及其产品发展趋势[J]．电子技术应用，2004，(8)：1-3．
[17] Tabassi E, Wilson C L. A novel approach to fingerprint image quality[C]. Image Processing, IEEE International Conference on ICIP, 2005, 2: 37-40.
[18] Xue-jun Tan, Bhanu B, Ying-qiang Lin. Fingerprint classification based on learned features[J]. IEEE Transactions on Systems Man and Cybemetics, Part C, 2005, 35(3):287-300.
[19] 林喜荣，黄析伟，苏晓生，周斌，裁晓清．生物特征识别技术的标准化进程[J]．清华大学学报(自然科学版)，2006，46(2)：194-198．
[20] R W Frischholz, U Diechmann. Bioid: a multimodal biometric identification system[J]. Computer, 2000, 33 (32): 64-68.
[21] 张亮．指纹图像传感器技术关键及发展趋势[J]．传感器技术，2005，24(9)：1-3．
[22] Rowe R K, Nixon KA, Corcoran S P. Multispectral fingerprint biometrics [C]. Proceedings of the Sixth Annual IEEE Workshop on SMC. NY: United States Military Academy, 2005: 14-20.
[23] Sato N, Shigematsu S, Morimura H, et al. Novel surface structure and its fabrication process for MEMS fingerprint sensor[J]. IEEE Transactions on Electron D v ces, 2005, 52 (5): 1026 1032.
[24] Gian Luca Marcialis, Fabio Roli. Fingerprint verification by fusion of optical and capacitive sensors [J]. Pattern Recognition Letters, 2004, (25): 1315-1322.
[25] 派斯数码．MBF200指纹芯片[EB／OL]．http://www.epasstech.com，2005．
[26] 富士通公司.MBF200 Solid State Fingerprint Sensor datasheet[EB/OL].http://www.fma.fujitsu.com,2001.
[27] 徐俊，龙占超．MBF200型指纹传感器的原理与应用[J]．电子世界，2005，(2)：40-41．
[28] 黄铝文．基于DSP的指纹识别系统的研究[D]．陕西：西安理工大学，2005．
[29] 余锦华，陈建华，施心陵．指纹图像的预处理[J]．计算机工程与设计，2004，25(12)：2210-2212．
[30] Jiang Xu-dong. On orientation and anisotropy estimation for online fingerprint authentication[J]. IEEE Trans on Signal Processing, 2005, 53(10): 4038-4049.
[31] Jain A K, Prabhakar S, Hong L, Pankanti S. Filter bank-based fingerprint matching [J]. IEEE Trans on Image Processing, 2000, 9(5): 846-859.
[32] Sherlock B G., Monro D M. A model for interpreting fingerprint topology[J]. Pattern Recognition, 1993, 26 (7): 1047-1055.
[33] 苏彦华等．Visual C++数字图像识别技术典型按例[M]．北京：人民邮电出版社，2004．
[34] 何斌，马天予等．Visual C++数字图像处理(第二版)[M]．北京：人民邮电出版社，2002．
[35] Alonso-Fernandez F, et al. An enhanced Gabor Filter-Based Segmentation Algorithm for Fingerprint Recognition Systems [C]. Image and Signal Processing and Analysis, Proceedings of the 4th International Symposium on ISPA, 2005: 239-244.
[36] Gabor D. Theory of communication [J]. Journal of the Institute of Electrical Engineers, 1946, 93(26): 429-457.
[37] 陆颖．指纹自动识别原理与方法综述[J]．工程数学学报，2004，21(6)：1003-1010．
[38] 田捷，杨鑫．生物特征识别技术理论与应用[M]．北京：电子工业出版社，2005．
[39] Jain A K, L Hong, S Prabhakar. On-line identity authentication system using fingerprints [J]. Proceedings of the IEEE, 1997, 85: 1365-1388.
[40] 罗希平，田捷．自动指纹识别中的图像增强与细节匹配[J]．软件学报，2002，13(5)：946-956．
[41] P. Jonathon Phillips, Alvin Martin, C. L. Wilson et al. An introduction to evaluating biometric systems. IEEE Computer, 2000: 56-63.
[2] Uludag U, Pankanti S, Prabhakar S, JainA K. Biometric cryptosystems: issues and challenges[J]. Proceedings of the IEEE, 2004, 92 (6): 948-960.
[3] 罗棻．生物识别技术的发展展望[J]．重庆工商大学学报(自然科学版)，2005，22(3)：249-252．
[4] Jain A K, Ross A, Prabhakar S. An introduction to biometric recognition[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2004, 14 (1):4-20.
[5] 张海春，回文博，林立忠．指纹识别技术研究进展[J]．石家庄学院学报，2005，7(3)：25-28．
[6] 段少雄，张堂辉，田捷．指纹识别技术漫谈[EB/OL]. http://industry.ccidnet.com/art/1101/20011210/6358_1.html, 2005-11-21.
[7] 詹小四．生物认证-人类未来身份认证的主要方式[J]．阜阳师范学院学报(自然科学版)，2004，21(2)：52-56．
[8] 霍宏涛．数字图像处理[M]．北京：北京理工大学出版社，2002．9．
[9] 于秀霞．指纹识别技术及其应用[J]．长春大学学报，2005，15(2)：30-32．
[10] 吕知辛，李兴江．指纹识别系统中的模式识别技术[J]．山东师范大学学报(自然科学版)，2004，19(4)：20-23．
[11] 李徐周．自动指纹识别技术的应用[J]．山东省青年管理干部学院学报，2005，2：143-144．
[12] 王崇文．自动指纹识别方法研究[D]．重庆：重庆大学，2002．
[13] 李见为．指纹识别技术的研究和发展[J]．重庆职业技术学院学报，2004，13(2)：4-6．
[14] 王卫东，平西建．自动指纹识别技术研究[J]．信息工程大学学报，2005，6(2)：47-50．
[15] 刘欣．基于图像内容的指纹识别系统的研究及其DSP实现[D]．陕西：西安理工大学，2004．
[16] 张成，周媛媛，林嘉宇．指纹采集技术及其产品发展趋势[J]．电子技术应用，2004，(8)：1-3．
[17] Tabassi E, Wilson C L. A novel approach to fingerprint image quality[C]. Image Processing, IEEE International Conference on ICIP, 2005, 2: 37-40.
[18] Xue-jun Tan, Bhanu B, Ying-qiang Lin. Fingerprint classification based on learned features[J]. IEEE Transactions on Systems Man and Cybemetics, Part C, 2005, 35(3):287-300.
[19] 林喜荣，黄析伟，苏晓生，周斌，裁晓清．生物特征识别技术的标准化进程[J]．清华大学学报(自然科学版)，2006，46(2)：194-198．
[20] R W Frischholz, U Diechmann. Bioid: a multimodal biometric identification system[J]. Computer, 2000, 33 (32): 64-68.
[21] 张亮．指纹图像传感器技术关键及发展趋势[J]．传感器技术，2005，24(9)：1-3．
[22] Rowe R K, Nixon KA, Corcoran S P. Multispectral fingerprint biometrics [C]. Proceedings of the Sixth Annual IEEE Workshop on SMC. NY: United States Military Academy, 2005: 14-20.
[23] Sato N, Shigematsu S, Morimura H, et al. Novel surface structure and its fabrication process for MEMS fingerprint sensor[J]. IEEE Transactions on Electron D v ces, 2005, 52 (5): 1026 1032.
[24] Gian Luca Marcialis, Fabio Roli. Fingerprint verification by fusion of optical and capacitive sensors [J]. Pattern Recognition Letters, 2004, (25): 1315-1322.
[25] 派斯数码．MBF200指纹芯片[EB／OL]．http://www.epasstech.com，2005．
[26] 富士通公司.MBF200 Solid State Fingerprint Sensor datasheet[EB/OL].http://www.fma.fujitsu.com,2001.
[27] 徐俊，龙占超．MBF200型指纹传感器的原理与应用[J]．电子世界，2005，(2)：40-41．
[28] 黄铝文．基于DSP的指纹识别系统的研究[D]．陕西：西安理工大学，2005．
[29] 余锦华，陈建华，施心陵．指纹图像的预处理[J]．计算机工程与设计，2004，25(12)：2210-2212．
[30] Jiang Xu-dong. On orientation and anisotropy estimation for online fingerprint authentication[J]. IEEE Trans on Signal Processing, 2005, 53(10): 4038-4049.
[31] Jain A K, Prabhakar S, Hong L, Pankanti S. Filter bank-based fingerprint matching [J]. IEEE Trans on Image Processing, 2000, 9(5): 846-859.
[32] Sherlock B G., Monro D M. A model for interpreting fingerprint topology[J]. Pattern Recognition, 1993, 26 (7): 1047-1055.
[33] 苏彦华等．Visual C++数字图像识别技术典型按例[M]．北京：人民邮电出版社，2004．
[34] 何斌，马天予等．Visual C++数字图像处理(第二版)[M]．北京：人民邮电出版社，2002．
[35] Alonso-Fernandez F, et al. An enhanced Gabor Filter-Based Segmentation Algorithm for Fingerprint Recognition Systems [C]. Image and Signal Processing and Analysis, Proceedings of the 4th International Symposium on ISPA, 2005: 239-244.
[36] Gabor D. Theory of communication [J]. Journal of the Institute of Electrical Engineers, 1946, 93(26): 429-457.
[37] 陆颖．指纹自动识别原理与方法综述[J]．工程数学学报，2004，21(6)：1003-1010．
[38] 田捷，杨鑫．生物特征识别技术理论与应用[M]．北京：电子工业出版社，2005．
[39] Jain A K, L Hong, S Prabhakar. On-line identity authentication system using fingerprints [J]. Proceedings of the IEEE, 1997, 85: 1365-1388.
[40] 罗希平，田捷．自动指纹识别中的图像增强与细节匹配[J]．软件学报，2002，13(5)：946-956．
[41] P. Jonathon Phillips, Alvin Martin, C. L. Wilson et al. An introduction to evaluating biometric systems. IEEE Computer, 2000: 56-63.