Novel Algorithm for Iris Biometrics Using Fast Wavelet Transforms.

Iris recognition is a potential tool in secure personal identification and authentication system which has the properties such as uniqueness, non-invasiveness and stability of human iris patterns. The method proposed in this paper, differed from the existing work in the segmentation phase where segmentation is done using a different morphological method. Also, comparison is done betw method, for their accuracy in detecting the pupil and iris region. The localized iris image is then normalized and Mallat’s fast wavelet transform is used for feature extraction. The obtained iris templa Hamming distance. The algorithm proposed in this method provides accurate features as well as simple and fast iris analysis. Digitized grayscale images from Chinese Academy of Sciences were used for determining the performance of the proposed system