Application-oriented Principles And Methodology Of Wavelet Transform In Iris Recognition

Biometrics makes use of the inherent physiological or behavioral characteristics of human beings to identify individuals. Due to the unique, stable and live physiological properties of the iris and non-invasive to users, most stable and reliable for identification in practical applications of iris-based system, iris recognition is receiving extensive attention and becoming an active topic in biometrics.As an application-oriented research project, iris recognition integrates mathematic, computer science, optics, electronics and physiology etc. Based on wavelet transform, aimed to improve the recognition performance, centered at wavelet-based iris feature representation and pattern classification, we review and explore the iris sequence image quality assessment, iris image pre-processing, iris recognition performance evaluation and several other linked topics. We mainly investigate on the principles and application methodology of wavelet transform for iris feature representation and iris pattern classification methods. Meanwhile, we indicate the key issues and directions of the wavelet-based iris recognition technology.Our main work and contributions in this thesis are as follows:(1) An iris image acquirement device is designed and the experimental platform for testing the iris recognition algorithm is constructed. Our home-made iris acquirement device obtained the national patent. Utilizing this device, we collected 2,464 iris images from 176 eyes at different time, which form the JLUBR-IRIS iris database. Except the acquirement system and JLUBR-IRIS database, the experimental platform includes iris pre-processing, iris feature extraction and recognition. This platform provides the open environment for new algorithms to be studied and analyzed.(2) The wavelet properties and their effect on the iris feature representation are analyzed. The principles of the wavelet transform for iris feature representation are probed and some key points related to wavelet-based iris recognition are summarized as: (a) how to represent the iris features with the wavelet coefficients; (b) how to select more effective wavelet base to analyze the iris texture; (c) how to overcome the shift sensitivity of the discrete wavelet transform; (d) how to choose the wavelet fast decomposition algorithm; (e) how to decide the decomposition scales of...