logo

SCIENCE CHINA Information Sciences, Volume 61, Issue 5: 052103(2018) https://doi.org/10.1007/s11432-016-9037-0

Finger vein recognition based on deformation information

More info
  • ReceivedJul 22, 2016
  • AcceptedJan 26, 2017
  • PublishedAug 23, 2017

Abstract

The measurement of the vessel pattern in fingers is a superior method for identifying individuals owing to its convenience and the security it offers. We introduce in this paper a new perspective to accomplish finger vein recognition. This method, which regards deformations as discriminative information, is distinct from existing methods that attempt to prevent the influence of deformations. The proposed technique is based on the observation that regular deformation, which corresponds to a posture change, can only exist in genuine vein patterns. In terms of methodology, we incorporate optimized matching to generate pixel-based 2D displacements that correspond to deformations. The texture of uniformity extracted from the displacement fields is taken as the final matching score. Evaluated on two publicly available databases, PolyU and SDU-MLA, extensive experiments demonstrated that the discriminability of the new feature derived from deformations is preferable. The equal error rate (EER) achieved is the lowest compared to that of state-of-the-art techniques.


Acknowledgment

The work was supported by National Science Foundation of China (Grant Nos. 61573219, 61472226), NSFC Joint Fund with Guangdong under Key Project (Grant No. U1201258), Natural Science Foundation for the Youth of Shandong Province (Grant No. ZR2016FQ18), and Fostering Project of Dominant Discipline and Talent Team of Shandong Province Higher Education Institutions.


References

[1] IEEE Transactions on Circuits and Systems for Video Technology Information for authors. IEEE Trans Circuits Syst Video Technol, 2004, 14: c4-c4 CrossRef Google Scholar

[2] Ross A A, Nandakumar K, Jain A K. Handbook of Multibiometrics. Berlin: Springer Science & Business Media, łinebreak 2006. 6. Google Scholar

[3] Yanagawa T, Aoki S, Ohyama T. Human finger vein images are diverse and its patterns are useful for personal identification. MHF Prepr Ser, 2007, 12: 1--7. Google Scholar

[4] Hashimoto J. Finger vein authentication technology and its future. In: Proceedings of 2006 Symposium on VLSI Circuits, Digest of Technical Papers, Honolulu, 2006. 5--8. Google Scholar

[5] Liu Z, Yin Y, Wang H. Finger vein recognition with manifold learning. J Network Comput Appl, 2010, 33: 275-282 CrossRef Google Scholar

[6] Wu J D, Ye S H. Driver identification using finger-vein patterns with Radon transform and neural network. Expert Syst Appl, 2009, 36: 5793-5799 CrossRef Google Scholar

[7] Kiyomizu H, Miura N, Miyatake T, et al. Finger vein authentication device. US Patent 8811689, 2014-08-19. Google Scholar

[8] Sato H. Finger vein authentication apparatus and finger vein authentication method. US Patent 8229179, 2012-07-24. Google Scholar

[9] Lee E C, Jung H, Kim D. New finger biometric method using near infrared imaging.. Sensors, 2011, 11: 2319-2333 CrossRef PubMed Google Scholar

[10] Meng X, Yang G, Yin Y, et al. Finger vein recognition based on local directional code. Sensors, 2010, 12: 14937--14952. Google Scholar

[11] Song W, Kim T, Kim H C. A finger-vein verification system using mean curvature. Pattern Recognition Lett, 2011, 32: 1541-1547 CrossRef Google Scholar

[12] Miura N, Nagasaka A, Miyatake T. Feature extraction of finger-vein patterns based on repeated line tracking and its application to personal identification. Machine Vision Appl, 2004, 15: 194-203 CrossRef Google Scholar

[13] Kumar A, Zhou Y. Human identification using finger images. IEEE Trans Biomed Eng, 2012, 21: 2228--2244. Google Scholar

[14] Yu C B, Qin H F, Cui Y Z. Finger-vein image recognition combining modified Hausdorff distance with minutiae feature matching.. Interdiscip Sci Comput Life Sci, 2009, 1: 280-289 CrossRef PubMed Google Scholar

[15] Liu F, Yang G, Yin Y. Singular value decomposition based minutiae matching method for finger vein recognition. Neurocomputing, 2014, 145: 75-89 CrossRef Google Scholar

[16] Wu J D, Liu C T. Finger-vein pattern identification using principal component analysis and the neural network technique. Expert Syst Appl, 2011, 38: 5423-5427 CrossRef Google Scholar

[17] Yang G, Xi X, Yin Y. Finger vein recognition based on (2d)$^{2}$pca and metric learning. J Biomed Biotech, 2012, 2012: 324249. Google Scholar

[18] Guan F, Wang K, Liu J, et al. Bi-direction weighted (2d)$^{2}$pca with eigenvalue normalization one forefinger vein recognition. Pattern Recogn Art Intell, 2011, 24: 417--424. Google Scholar

[19] Wu J D, Liu C T. Finger-vein pattern identification using svm and neural network technique. Expert Syst Appl, 2011, 38: 14284--14289. Google Scholar

[20] Yang L, Yang G, Yin Y, et al. A survey of finger vein recognition. In: Proceedings of Chinese Conference on Biometric Recognition. Berlin: Springer, 2014. 234--243. Google Scholar

[21] Lee E C, Lee H C, Park K R. Finger vein recognition using minutia-based alignment and local binary pattern-based feature extraction. Int J Imag Syst Technol, 2009, 19: 179-186 CrossRef Google Scholar

[22] Lee E C, Park K R. Image restoration of skin scattering and optical blurring for finger vein recognition. Optics Lasers Eng, 2011, 49: 816-828 CrossRef ADS Google Scholar

[23] Mulyono D, Jinn H S. A study of finger vein biometric for personal identification. In: Proceedings of International Symposium on Biometrics and Security Technologies, Islamabad, 2008. 1--8. Google Scholar

[24] Qin H, Qin L, Xue L. Finger-vein verification based on multi-features fusion.. Sensors, 2013, 13: 15048-15067 CrossRef PubMed Google Scholar

[25] Yang J, Shi Y. Finger-vein ROI localization and vein ridge enhancement. Pattern Recognition Lett, 2012, 33: 1569-1579 CrossRef Google Scholar

[26] Lu Y, Xie S, Yoon S. Robust finger vein ROI localization based on flexible segmentation.. Sensors, 2013, 13: 14339-14366 CrossRef PubMed Google Scholar

[27] Yang L, Yang G, Yin Y. Sliding window-based region of interest extraction for finger vein images.. Sensors, 2013, 13: 3799-3815 CrossRef PubMed Google Scholar

[28] Miura N, Nagasaka A, Miyatake T. Extraction of finger-vein patterns using maximum curvature points in image profiles. IEICE Trans Inf Syst, 2007, 90: 1185--1194. Google Scholar

[29] Ojala T, Pietikainen M, Maenpaa T. Multiresolution gray-scale and rotation invariant texture classification with local binary patterns. IEEE Trans Pattern Anal Machine Intell, 2002, 24: 971-987 CrossRef Google Scholar

[30] Pang S, Yin Y, Yang G, et al. Rotation invariant finger vein recognition. In: Proceedings of Chinese Conference on Biometric Recognition. Berlin: Springer, 2012. 151--156. Google Scholar

[31] Rosdi B A, Shing C W, Suandi S A. Finger vein recognition using local line binary pattern.. Sensors, 2011, 11: 11357-11371 CrossRef PubMed Google Scholar

[32] Baochang Zhang , Yongsheng Gao , Sanqiang Zhao . Local Derivative Pattern Versus Local Binary Pattern: Face Recognition With High-Order Local Pattern Descriptor. IEEE Trans Image Process, 2010, 19: 533-544 CrossRef PubMed ADS Google Scholar

[33] Xi X, Yang G, Yin Y. Finger vein recognition with personalized feature selection.. Sensors, 2013, 13: 11243-11259 CrossRef PubMed Google Scholar

[34] Liu C, Yuen J, Torralba A, et al. Sift flow: dense correspondence across different scenes. In: Proceedings of European Conference on Computer Vision. Berlin: Springer, 2008. 28--42. Google Scholar

[35] Lowe D G. Distinctive Image Features from Scale-Invariant Keypoints. Int J Comput Vision, 2004, 60: 91-110 CrossRef Google Scholar

[36] Gonzalez R C, Woods R E, Eddins E L. Digital Image Processing Using MATLAB. Princeton: Pearson Education Inc., 2004. Google Scholar

[37] Yin Y, Liu L, Sun X. Sdumla-hmt: a multimodal biometric database. In: Proceedings of Chinese Conference on Biometric Recognition. Berlin: Springer, 2011. 260--268. Google Scholar

[38] Si X, Feng J, Zhou J. Detection and Rectification of Distorted Fingerprints.. IEEE Trans Pattern Anal Mach Intell, 2015, 37: 555-568 CrossRef PubMed Google Scholar

[39] Jongsun Kim , Jongmoo Choi , Juneho Yi . Effective representation using ICA for face recognition robust to local distortion and partial occlusion.. IEEE Trans Pattern Anal Machine Intell, 2005, 27: 1977-1981 CrossRef PubMed Google Scholar

  • Figure 1

    Various kinds of deformations of the same finger. (a) and (b) Finger vein images affected by in-plane translation and rotation; (c) and (d) examples of out-of-plane rotation; (e) bent finger; (f) captured after exercise, and exhibits low intensity.

  • Figure 2

    (Color online) Displacements generated in genuine and imposter matching show discriminative information. Yellow arrows indicate the pixel-based displacements. (a) Genuine matching: the four matched pixels have similar displacements, i.e., the displacements of the four positions are small and similar in direction; (b) imposter matching: the displacements of the four matched pixels vary, i.e., the displacements are large in distance and in various directions.

  • Figure 3

    Flowchart of the proposed method.

  • Figure 4

    (Color online) Illustration of the ROI extraction. (a) Center line detection of the finger vein image, according to which the skew can be adjusted; (b) finger joints detection; (c) internal tangents of the edges of the finger; (d) the ROI of the sample finger.

  • Figure 5

    Displacement matrices of (a) genuine and (b) imposter matching.

  • Figure 6

    Description of uniformity feature. (a) The normalized histogram of the horizontal direction; (b) the normalized histogram of the vertical direction; (c) the merged histogram of both directions.

  • Figure 7

    (Color online) Distribution of matching scores on the PolyU database. Score distribution of (a) proposed method and (b) MO function.

  • Figure 8

    (Color online) Distribution of matching scores on the SDU-MLA database. Score distribution of (a) proposed method and (b) MO function.

  • Figure 9

    (Color online) ROC curves for the two databases: (a) PolyU database and (b) SDU-MLA database.

  • Figure 10

    (Color online) Cumulative match curves by different methods: (a) PolyU database and (b) SDU-MLA database.

  • Figure 11

    (Color online) Performance when parameter $d$ varies.

  • Figure 12

    (Color online) ROC curves for different methods.

  • Table 1   Verification performance by different methods on PolyU and SDU-MLA database
    Method PolyU database SDU-MLA database
    EER FRR at-zero-FAR FAR at-zero-FRR EER FRR at-zero-FAR FAR at-zero-FRR
    MO function 0.0310 0.2511 0.9183 0.0684 0.8936 0.9982
    Proposed method 0.0053 0.2182 0.6674 0.0268 0.7152 0.9989
  • Table 2   Time efficiency analysis of the proposed method
    Method Preprocessing (ms) Feature Optimized Feature Total (ms)
    extraction (ms) matching (ms) extraction (ms)
    MO function 37.2 9.4 287.4 334
    Proposed method 37.2 9.4 287.4 0.6 334.6
  • Table 3   Identification performance by different methods on PolyU and SDU-MLA database
    Method PolyU database SDU-MLA database
    Rank-one Lowest rank of Rank-one Lowest rank of
    recognition rate perfect recognition recognition rate perfect recognition
    MO Function 0.9382 ($\pm$0.0222) 230 0.8483 ($\pm$0.0158) 633
    Proposed method 0.9940 ($\pm$0.0041) 25 0.9400 ($\pm$0.0074) 477
  • Table 4   Performance comparison with and without smooth term
    Method Verification mode Recognition mode
    EER FRR at-zero-FAR FAR at-zero-FRR Rank-one Lowest rank of
    recognition rate perfect recognition
    Without smooth 0.0145 0.0754 0.9499 0.9810 ($\pm$0.0034) 224
    Proposed method 0.0053 0.2182 0.6674 0.9940 ($\pm$0.0041) 25
  • Table 5   Performance of different methods on the PolyU database
    Method EER T-test
    $H$ $P$
    LBP[29] 0.0690 ($\pm$ 0.0277) 1 2.90E$-$05
    LLBP[31] 0.0427 ($\pm$ 0.0286) 1 3.65E$-$04
    LDP[32] 0.2241 ($\pm$ 0.0327) 1 1.76E$-$07
    LDC[10] 0.0359 ($\pm$ 0.0331) 1 0.0018
    MeanC[11] 0.1064 ($\pm$ 0.0615) 1 2.05E$-$04
    MaxC[28] 0.0265 (–)
    RLT[12] 0.0825 (–)
    EGM[13] 0.0065 (–)
    SIFT[30] 0.0472 ($\pm$ 0.0256) 1 1.18E$-$04
    SPM[33] 0.0357 ($\pm$ 0.0275) 1 6.77E$-$04
    SVDMM[15] 0.0245 ($\pm$ 0.0147) 1 5.06E$-$04
    Proposed method 0.0010 ($\pm$ 0.0035)

    – means the value is not provided in the reference paper.

Copyright 2019 Science China Press Co., Ltd. 《中国科学》杂志社有限责任公司 版权所有

京ICP备18024590号-1