指纹识别技术作为目前主要的身份识别方式之一，以其较高的安全性、可靠性以及便利性被广泛应用于社会各个领域。指纹识别领域在大面积指纹识别、高质量指纹识别、成年指纹识别以及单一数据库指纹识别等方面的技术已经非常成熟，而在小面积指纹识别、低质量指纹识别、婴幼儿指纹识别和交叉库指纹识别等方面，识别精度还有待提高，识别算法和相关理论还需要深入探索。交叉库指纹识别是指将来自于不同采集仪的指纹进行匹配的过程。由于指纹图像来自不同原理和规格的传感器，导致待匹配指纹间存在较大的非线性形变、尺度缩放和旋转平移变化，相对于常规的指纹匹配，交叉库指纹之间的识别率较低。交叉库指纹识别已经成为目前指纹识别领域的研究热点之一。

本文针对交叉库指纹间较大非线性形变的问题，深入研究基于细节点空间拓扑关系的匹配方法，主要包括：

(1)提出了一种基于细节点的混合特征算子。在经典细节点方向算子的基础上，融合了指纹频率信息，并将其用以指纹匹配。通过特征融合，增加了指纹细节点匹配的准确性，为后续的交叉库指纹匹配打下了良好的基础。

(2)提出一种细节点空间特征融合的交叉库指纹识别算法。该算法针对交叉库指纹图像间较大非线性形变的问题，将细节点混合特征描述子与传播算法相结合，在匹配过程中使用细节点空间拓补结构作为约束条件，实现了交叉库指纹间的精确匹配。

(3)实现了交叉库指纹细节点特征和脊线特征的结合。在传播算法的基础上，添加受非线性形变影响较小的脊线计数特征，并获取脊线相似度。将细节点和脊线的相似度分数相融合作为图像的匹配分数，使细节点信息和脊线信息能够相互补充，达到较低的EER值。

本文对提出的创新算法在Fingerpass的三个子数据库上进行了交叉匹配实验，结果显示：针对较大形变的交叉库指纹匹配，本文算法的最优EER值为2.01%，平均EER达到2.28%，相对于其他算法，性能有明显提高，验证了本文方法的有效性。

fingerprint identification technology is widely used in all aspects of human social life as a popular identity identification method with its stable security, reliability and convenience. at present, fingerprint recognition has been very mature in terms of large-area fingerprint recognition, high-quality fingerprint recognition, adult fingerprint recognition, and fingerprint recognition between single databases, its application in various fields of society is also increasing. however, in areas such as small-area fingerprint recognition, low-quality fingerprint recognition, infant fingerprint recognition, and fingerprint cross-matching identification, the recognition accuracy still needs to be improved, the recognition algorithms and related theories have to be explored. in the field of fingerprint cross-matching identification, fingerprint images come from sensors of different principles and specifications, resulting in large non-linear deformation, scaling and rotation translation changes between the fingerprints to be matched. compared with conventional fingerprint matching, the recognition rate between cross-library fingerprints is low. fingerprint cross-matching has become one of the research hotspots in the field of fingerprint recognition.

 

in this paper, we have done some exploratory work in the field of fingerprint cross-matching recognition for the problem of large nonlinear deformation, including the following aspects.

 

(1) this paper proposes a minutia-based hybrid feature operator. on the basis of the classic minutiae direction operator, the fingerprint frequency information is integrated and used for fingerprint matching. through feature fusion, the accuracy of fingerprint minutiae matching is increased, which lays a good foundation for subsequent fingerprint cross- matching.

 

(2) this paper proposes a fingerprint cross-matching algorithm based on the detailed spatial feature fusion. this algorithm combines the minutiae hybrid feature deor with the propagation algorithm for the problem of large nonlinear deformations between cross-matching images. the minutia space topology is used as a constraint condition in the matching process to realize the fingerprint cross- matching.

(3) in this paper, the combination of fingerprint minutiae feature and ridge feature of the cross-matching is realized. based on the propagation algorithm, adding the ridge counting features which are less affected by the nonlinear deformation, then obtain the ridge similarity. the similarity score of the minutiae and ridges are merged as the matching score of the images, so that the minutiae information and the ridge information can complement each other and a lower eer value is achieved.

 

in this paper, the cross-match experiments on the three sub-databases of fingerpass are carried out on the proposed innovative algorithm. the results show that the best eer value of the proposed algorithm is 2.01% and the average eer reaches 2.28%. compared with other algorithms, performance has improved significantly, which verifies the effectiveness of the proposed method.

[1] 田捷, 杨鑫等. 生物特征识别理论与应用[m]. 北京: 清华大学出版社, 2009.
[2] 庞辽军, 赵伟强, 李岩等. 生物特征加密基础[m]. 北京: 电子工业出版社, 2016.
[3] zhou j, gu j. a model-based method for the computation of fingerprints' orientation field[j]. image processing ieee transactions on, 2004, 13(6): 821.
[4] arun r, rohan n. a thin-plate spline calibration model for fingerprint sensor interoperability[m]. ieee educational activities department, 2008
[5] jihyeno j, stephen j e, hakil k. on improving interoperability of fingerprint recognition using resolution compensation based on sensor evaluation. icb 2007. 2007, 4642. 455-463.
[6] senior a, bolle r. improved fingerprint matching by distortion removal. ieice trans. on information and system, special issue on biometrics, e84-d(7):825-831, 2001.
[7] watson c i, grother p j, and casasent d p. distortion-tolerant filter for elastic-distorted fingerprint matching, in proceedings of spie optical pattern recognition, florida usa, 2000, pp.166-174.
[8] kovacs-vajna, z. m. a fingerprint verification system based on triangular matching and dynamic time warping. ieee trans. pattern anal. mach. intell. 22(11): 1266-1276,2000.
[9] he y l, tian j, ren q, et al. maximum-likelihood deformation analysis of different-sized fingerprints. avbpa 2003: 421-428
[10] lim c, kim s, lee c. geochemical fingerprint of the primary magma composition in the marine tephras originated from the baegdusan and ulleung volcanoes[j]. journal of asian earth sciences, 2014, 95(95):266-273.
[11] cao k, yang x, chen x, et al. a novel ant colony optimization algorithm for large-distorted fingerprint matching[j]. pattern recognition, 2012, 45(1):151-161
[12] liu e y, zhao h, et al. fingerprint segmentation based on an adaboost classifier[j]. frontiers of computer science in china, 2011, 5(2):148-157.
[13] jain a k, feng j, nandakumar k. fingerprint matching[j]. computer. 2010, 43(2), pp 36-44
[14] maltoni d, maio d, jain a k, and prabhaker s. handbook of fingerprint recognition[m]. chicago: library of congress control number, 2009.
[15] prabhakar s, ivanisov a, and jain a k. biometric recognition: sensor characteristics and image quality, ieee instrumentation & measurement magazine. 2011, 14(3)，pp 10-16
[16] jain a k, hong l, bolle r. on-line fingerprint verification[j]. ieee trans. on pattern analysis and machine inteligence, 1997, 19(4), pp 302-314
[17] jain a k, ross a, prabhakar s. an introduction to biometric recognition. ieee transactions on circuits and systems for video technology. 2004, 14 (1), pp 4 -20.
[18] hong, l. wan, a. k. jain. fingerprint enhancement: algorithm and performance evaluation[j]. ieee trans. on pattern analysis and machine intelligence, 1998, 20(8): 777-789.
[19] gragnaniello d, poggi g, sansone c, verdoliva l. local contrast phase deor for fingerprint liveness detection. pattern recognition. 2015, 48(4), pp 1050-1058
[20] bazen a m. directional field computation for fingerprints based on the principal component analysis of local gradients[j]. ieee trans. on image processing. 2000, 17(6), pp 23-29.
[21] dass s c. fingerprint-based recognition. wiley .2013,81(2)，pp 175-187
[22] jain a k, nandakumar k, ross a.50 years of biometric research: accomplishments, challenges, and opportunities. pattern recognition letters.2016, 79(1), pp 80-105
[23] bazen a m, gerez s h. segmentation of fingerprint images. workshop on circuits systems and signal processing, the netherlands, november 2001.
[24] chikkerur s. cartwright a n, govindaraju v. fingerprint enhancement using stft analysis. pattern recognition. 2007, 40(1), pp 198-211
[25] sherlock b g, monro d m, and millard k. fingerprint enhancement by directional fourier filtering[j]. iee proceedings - vision, image and signal processing.1994, 141(2), pp 87-94.
[26] luo p, tian j. knowledge based fingerprint image enhancement. pattern recognition. 2000, 4, pp 783-736
[27] cheng j, tian j, chen h, ren q, and yang x. fingerprint enhancement using oriented diffusion filter[c]. proceedings of 4th international conference on audio- and video-based biometric person authentication, spirnger, 2003:164-171
[28] cheng j, tian j. fingerprint enhancement with dyadic scale-space. pattern recognition letters. 2004, 25(11), pp 1273-1284
[29] mehtre b m, chatterjee b. segmentation of fingerprint image-a composite method. pattern recognition. 1989,22(4), pp 381－385
[30] feng j. combining minutiae deors for fingerprint matching[j]. pattern recognition.2008, 41(1), pp 342-352.
[31] choi k, choi h, lee s, and kim j h, fingerprint image mosaicking by recursive ridge mapping, ieee transaction on systems, 2007, 37(5), pp.1191-1203.
[32] ramoser h, wachmann b, bischof h. ramoser, efficient alignment of fingerprint images[c]. international conference on pattern recognition, 2002. proceedings ieee, 2002:748-751 vol.3.
[33] gil y, ahn d, ryu c, et al. user enrollment using multiple snapshots of fingerprint. lecture notes in computer science 3316(2004):344-349.
[34] toh k a, yau w y, jiang x. minutia data synthesis for fingerprint identification applications, image process, 2001, vol. 3, pp.262-265.
[35] 梅园, 孙怀江, 夏德深. 一种基于梯度的健壮的指纹方向场估计算法[j]. 计算机研究与发展, 2007, 44(6): 1022-1031.
[36] xie w h, tian j, he y. embedded fingerprint identification system based on dsp[j]. proceedings of spie - the international society for optical engineering, 2005, 5779.
[37] tico m, kuosmanen p. fingerprint matching using an orientation-based minutia deor[j]. ieee transactions on pattern analysis & machine intelligence, 2003, 25(8): 1009-1014.
[38] cappelli r, ferrara m, maltoni d. minutia cylinder-code: a new representation and matching technique for fingerprint recognition[j]. ieee trans. on pattern analysis and machine intelligence. 2010, 32(12), pp 2128-2141
[39] zhang w w, and wang y. core-based structure matching algorithm of fingerprint verification. 16 th international conference on pattern recognition ieee computer society, 2002:10070.
[40] do a d, dang q h. an optimal minutiae-based combining method for multiple fingerprint templates[c]. international conference on advanced technologies for communications ieee, 2016:97-102.
[41] jain a k, feng j. latent fingerprint matching[j]. ieee trans. on pattern analysis and machine intelegence.2011, 33(1): 88-100.
[42] ramoser h, wachmann b, efficient alignment of fingerprint images[c]. international conference on pattern recognition, 2002. proceedings ieee, 2002:748-751 vol.3.
[43] mao k m, zhu z l, deng z f. fingerprint core point detection based on svm and complex filter[j]. psychophysiology, 2011, 3(8): 85-96.
[44] liu e, jain a k, tian j. a coarse to fine minutiae-based latent palmprint matching[j]. pattern analysis and machine intelligence, ieee transactions on, 2013, 35(10):2307 – 2322
[45] hu c, yin j, zhu e, et al. fingerprint alignment using special ridges[c]. international conference on pattern recognition, 2008, 10: 1-4.
[46] cheng j, tian j, chen h. fingerprint minutiae matching with orientation and ridge[c]. dblp, 2004: 351-358.
[47] xu h, veldhuis r n j, kevenaar t a m, et al. a fast minutiae-based fingerprint recognition system[j]. systems journal ieee, 2010, 3(4): 418-427.
[48] zhang d, liu f, zhao q, et al. selecting a reference high resolution for fingerprint recognition using minutiae and pores[j]. ieee transactions on instrumentation & measurement, 2011, 60(3): 863-871.
[49] marasco e, ross a. a survey on antispoofing schemes for fingerprint recognition systems[j]. acm computing surveys, 2014, 47(2): 1-36.
[50] bossen a, lehmann r, meier c. internal fingerprint identification with optical coherence tomography[j]. photonics technology letters ieee, 2010, 22(7): 507-509.
[51] fang g, srihari s n, srinivasan h. use of ridge points in partial fingerprint matching[j]. proc spie, 2007, 6539: 65390d-65390d-9.
[52] boshra m. processing a fingerprint for fingerprint matching[j]. u.s. patent, 2016, 12: 7-9.
[53] mousmi s, meyappan d t. fingerprint identification using minutiae matching[j]. international journal of computer trends & technology, 2013, 4(5): 127-131
[54] divan t, gulhane v. a fingerprint matching technique using minutiae based algorithm for voting system: a survey[c]. ieee international conference on electrical, computer and communication technologies. ieee, 2015: 1-5.
[55] arun ross, rohan nadgir, a thin-plate spline calibration model for fingerprint sensor interoperability, ieee transactions on knowledge and data engineering, 2008, 20(8):1097-1110
[56] chen j, pan c, wu c, et al. reproducing kernel particle methods for large deformation analysis of non-linear structures. computer methods in applied mechanics and engineering, 1996, 139:195–227
[57] lee d, choi k, lee s, et al. fingerprint fusion based on minutiae and ridge for enrollment[c]. audio-and video-based biometrie person authentication, international conference, avbpa 2003, guildford, uk, june 9-11, 2003 proceedings dblp, 2003:478-485.
[58] 姜红超, 张高伟, 基于脊线跟踪的指纹细节点特征提取[j]. 计算机工程, 2007, 33(8): 76-85.
[59] jain a k, arun r, prabhakar s. fingerprint matching using minutiae and texture features. image processing, 2001. proc. 2001 int. conf. on. 2001, 3. 282-285.
[60] xie w h, tian j, he y. embedded fingerprint identification system based on dsp[j]. proceedings of spie - the international society for optical engineering, 2005, 5779.
[61] arun r, jain a k, qian j z. information fusion in biometrics[c]. the 3rd int. conf. on audio and video based biometric person authentication. 2002, 4. 354-360.