Multiple Band Selection of Multispectral Dorsal Hand

  • David Zhang
  • Zhenhua Guo
  • Yazhuo Gong


The optimal single band for dorsal hand recognition is 890 nm based on the fusion of MFRAT and CompCode. The extracted vein information is limited within one spectrum only. One of the advantages of multispectral technique is to pursue higher recognition performance through the fusion of multiple bands. The main theoretical basis is that the images on different bands have complementary information, which is helpful to performance improvement. Obviously, adding more bands is similar to feature dimension increase, and the redundant information may also increase quickly especially when the bands are highly relevant to each other. So the multispetral band selection process is limited to high efficiency, and it demands that the selected bands have the maximum uncorrelation. Noting that recognition of visible region is obviously far worse and the texture of non-vein part is not reliable enough for efficient recognition, we plan to do the multiple band selection work in near infrared (NIR) region only. 422 dorsal hands of East Asians with different bands ranging from 700 to 1040 nm are used. This chapter tries to address two basic issues, the number of the bands for optimal group and which bands can explain the multispectral model more precisely. Unlike optimal single band selection, exhaustive method is not practicable for this task. The number of possible combination is immeasurable, especially when the optimal band number is unknown. Our scheme is to realize the task in two steps: First, divide the NIR region into several band classes according to special rules so that the number of optimal band is fixed; secondly, choose the bands from these classes to represent them with proper estimation criterion.


Dorsal hand recognition Multispectral image Band clustering Band fusion 


  1. Arnau TJ, Sotoca JM, Pla F (2004) Non agressive orange acid and sugar indexes estimation system. Image analysis for agricultural products and processes, Vol 69, pp 170–174Google Scholar
  2. Assaleh K, Qaddoumi N, Shanableh T, Adel M (2011) A novel biometric via hand structure using near-field microwave imaging. In: IEEE international conference on automatic face and gesture recognition and workshops, pp 167–172 (doi: 10.1109/FG.2011.5771392)
  3. Chen K, Zhang D (2011) Band selection for improvement of dorsal hand recognition. In: IEEE proceedings of international conference on hand-based biometrics, pp 1–4 (doi: 10.1109/ICHB.2011.6094333)
  4. Chen S, Zhang R, Su H, Tian J, Xia J (2010) SAR and multispectral image fusion using generalized IHS transform based on a trous wavelet and EMD decompositions. IEEE Sens J 10(3):737–745. doi: 10.1109/JSEN.2009.2038661 CrossRefGoogle Scholar
  5. Ding Y, Zhuang D, Wang K (2005) A study of hand vein recognition method. In: IEEE proceedings of international conference on mechatronics and automation, vol 4, pp 2106–2110 (doi: 10.1109/ICMA.2005.1626888)
  6. Ekenel HK, Stiefelhagen R (2010) Automatic frequency band selection for illumination robust face recognition. In: 20th international conference on pattern recognition, pp 2684–2687 (doi: 10.1109/ICPR.2010.658)
  7. Foote J (2000) Automatic audio segmentation using a measure of audio novelty. In: IEEE conference on multimedia and expo, vol 1, pp 452–455 (doi: 10.1109/ICME.2000.869637)
  8. Huang R, Li X (2008) Band selection based on evolution algorithm and sequential search for hyperspectral classification. In: Proceeding of international conference on audio, language and image processing, pp 1270–1273Google Scholar
  9. Jia W, Huang D-S, Zhang D (2008) Palmprint verification based on robust line orientation code. Pattern Recogn 41(5):1504–1513. doi: 10.1016/j.patcog.2007.10.011 MATHCrossRefGoogle Scholar
  10. Johnson RA, Wichern DW (1998) Applied multivariate statistical analysis, 4th edn. Prentice-Hall, Upper Saddle RiverGoogle Scholar
  11. Kumar A, Prathyusha KV (2009) Personal authentication using hand vein triangulation and knuckle shape. IEEE Trans Image Process 18(9):2127–2136. doi: 10.1109/TIP.2009.2023153 MathSciNetCrossRefGoogle Scholar
  12. Ozawa K (1983) CLASSIC: a hierarchical clustering algorithm based on asymmetric similarities. Pattern Recogn 16(2):201–211. doi: 10.1016/0031-3203(83)90023-7 MATHCrossRefGoogle Scholar
  13. Pearson K (1896) Mathematical contributions to the theory of evolution, regression, heredity and panmixia. Philos Trans R Soc Lond 187:253–318MATHCrossRefGoogle Scholar
  14. Sanchit, Ramalho M, Correia P, Soares L (2011) Biometric identification through palm and dorsal hand vein patterns. In: IEEE international conference on computer as a tool (EUROCON), pp 1–4 (doi: 10.1109/EUROCON.2011.5929297)
  15. Theodoridis S, Koutroumbas K (2006) Pattern recognition, 3rd edn. Elsevier, AmsterdamMATHGoogle Scholar
  16. Wang L, Leedham G, Cho S-Y (2007) Infrared imaging of hand vein patterns for biometric purposes. IET Comput Vision 1(3–4):113–122. doi: 10.1049/iet-cvi:20070009 MATHMathSciNetCrossRefGoogle Scholar
  17. Wang K, Zhang Y, Yuan Z, Zhuang D (2008) Hand vein recognition based on multi supplemental features of multi-classifier fusion decision. In: IEEE proceedings of international conference on mechatronics and automation 1790–1795 (doi: 10.1109/ICMA.2006.257486)
  18. Wu X, Gao E, Tang Y, Wang K (2010) A novel biometric system based on hand vein. In: 5th international conference on frontier of computer science and technology, pp 522–526 (doi: 10.1109/FCST.2010.65)
  19. Yang L, Liu X, Liu Z (2010) A skeleton extracting algorithm for dorsal hand vein pattern. In: International conference on computer application and system modeling 13, pp 92–95 (doi: 10.1109/ICCASM.2010.5622671)
  20. Yuksel A, Akarun L, Sankur B (2010) Biometric identification through hand vein patterns. In: International workshop on emerging techniques and challenges for hand-based biometrics, pp 1–6 (doi: 10.1109/ETCHB.2010.5559295)

Copyright information

© Springer International Publishing Switzerland 2016

Authors and Affiliations

  1. 1.Biometrics Research CentreThe Hong Kong Polytechnic UniversityHung HomHong Kong SAR
  2. 2.Shenzhen Key Laboratory of Broadband Network & Multimedia, Graduate School at ShenzhenTsinghua UniversityShenzhenChina
  3. 3.University of Shanghai for Science and TechnologyShanghaiChina

Personalised recommendations