Cascading Statistical And Structural Classifiers For Iris Recognition (TA-L3)

Author(s) : Zhenan Sun (Center for Biometrics Authentication and Testing, National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences, China) Yunhong Wang (Center for Biometrics Authentication and Testing, National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences, China) Tieniu Tan (Center for Biometrics Authentication and Testing, National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences, China) Jiali Cui (Center for Biometrics Authentication and Testing, National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences, China)

Abstract : Reliable human identification using iris pattern has recently gained growing interests from pattern recognition researchers. In literature of iris recognition, almost all algorithms are based on a test of statistical independence. In this paper, a structural iris image analysis method is proposed, which provides complementary information to statistical classifier. In order to save computational cost, the structural matcher is not consulted unless the statistical classifier is uncertain of its decision. Three typical statistical iris signatures, i.e. multi-channel texture information, demodulated phase value and wavelet decomposition coefficients, are combined with the structural iris feature respectively to improve iris recognition accuracy. Extensive experimental results demonstrate that the cascaded classification system significantly outperforms single classifier.

Menu