IRIS Recognition Using Multi-Directional Wavelets: A Novel Approach

The increasing requirement of security due to advances in information technologies, especially e-Commerce have led to rapid development of personnel identification /recognition systems based on biometrics. Iris is one of the most reliable biometrics because of its uniqueness, stability and non-invasive nature. A remarkable and important characteristic of the iris is the randomly distributed irregular texture details in all directions. In this paper, the authors have proposed a novel approach of feature extraction of iris image using a new method, 2D redundant rotated complex wavelet transform (RCWT) which obtains the features in 12 different directions, when used in conjunction with 2D Dual Trace Complex wavelet Transform(DT-CWT) against 3 and 6 directions in Discrete Wavelet Transform (DWT) and Complex Wavelet Transform (CWT) respectively. Iris features are obtained by computing 36 energies and 36 standard deviation of detailed coefficients in 12 directions per stage, at 3 level of decomposition. Canbera distance is used for matching. The results are obtained using DWT, CWT and combination of CWT and RCWT on UBIRIS database of 2400 images. The performance measure, ZeroFAR is reduced from 6.3 using DWT to 2.7 using the proposed method. The results are also comparable with the Daughman method. The method is also computationally efficient as compared to Gabor Filters.