Coarse Iris Classification by Learned Visual Dictionary

  • Xianchao Qiu
  • Zhenan Sun
  • Tieniu Tan
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4642)


In state-of-the-art iris recognition systems, the input iris image has to be compared with a large number of templates in database. When the scale of iris database increases, they are much less efficient and accurate. In this paper, we propose a novel iris classification method to attack this problem in iris recognition systems. Firstly, we learned a small finite dictionary of visual words(clusters in the feature space), which are called Iris-Textons, to represent visual primitives of iris images. Then the Iris-Texton histograms are used to represent the global features of iris textures. Finally, K-means algorithm is used for classifying iris images into five categories. Based on the proposed method, the correct classification rate is 95% in a five-category iris database. By combining this method with traditional iris recognition algorithm, our system shows better performance in terms of both speed and accuracy.


Global Feature Iris Image Iris Recognition Visual Dictionary Iris Database 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


  1. 1.
    Daugman, J.: High confidence visual recognition of persons by a test of statistical independence. IEEE TRANS. PAMI 15(11), 1148–1161 (1993)Google Scholar
  2. 2.
    Ma, L., Tan, T., Wang, Y., Zhang, D.: Personal identification based on iris texture analysis. IEEE TRANS. PAMI 25(12), 1519–1533 (2003)Google Scholar
  3. 3.
    Wildes, R.P.: Iris recognition: An emerging biometric technology. Proceedings of the IEEE 85(9), 1348–1363 (1997)CrossRefGoogle Scholar
  4. 4.
    Karu, K., Jain, A.K.: Fingerprint classification. Pattern Recognition 29(3), 389–404 (1996)CrossRefGoogle Scholar
  5. 5.
    Wu, X., Zhang, D., Wang, K., Huang, B.: Palmprint classification using principal lines. Pattern Recognition 37, 1987–1998 (2004)zbMATHCrossRefGoogle Scholar
  6. 6.
    Yu, L., Zhang, D., Wang, K., Yang, W.: Coarse iris classification using box-counting to estimate fractal dimensions. Pattern Recognition 38, 1791–1798 (2005)CrossRefGoogle Scholar
  7. 7.
    Fu, J., Caulfield, H.J., Yoo, S.-M., Atluri, V.: Use of artificial color filtering to improve iris recognition and searching. Pattern Recognition Letters 26, 2244–2251 (2005)CrossRefGoogle Scholar
  8. 8.
    Zhu, S.-C., Guo, C.-E., Wang, Y., Xu, Z.: What are textons? International Journal of Computer Vision 62(1-2), 121–143 (2005)CrossRefGoogle Scholar
  9. 9.
    Winn, J., Criminisi, A., Minka, T.: Object categorization by learned universal visual dictionary, vol. 2, pp. 1800–1807 (2005)Google Scholar
  10. 10.
    CASIA iris image database (2003),
  11. 11.
    University of Bath iris image database,
  12. 12.
    Jain, A.K., Ross, A.: Learning user-specific parameters in a multibiometric system, pp. 57–60 (2002)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2007

Authors and Affiliations

  • Xianchao Qiu
    • 1
  • Zhenan Sun
    • 1
  • Tieniu Tan
    • 1
  1. 1.Center for Biometrics and Security Research, National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences, P.O. Box 2728, Beijing, 100080P.R. China

Personalised recommendations