Abstract:

In order to overcome the weakness of wavelet transform in two dimensional spatial analysis，an improved algorithm based on fast discrete Curvelet transform for iris recognition was proposed.Curvelet transform which can effectively capture the image edge information was introduced to decompose iris image.Mean and variance of low frequency sub-band coefficients and the energy of high frequency sub-band were extracted.Then the feature vectors were weighted according to the difference of classification ability of sub-band feature.Fuzzy least square support vector machine with optimal binary tree was developed to implement classification and recognition.The simulation results show that the proposed algorithm has higher recognition performance than the present method.

Key words: iris recognition, feature weighting, fast discrete Curvelet transform, fuzzy least square support vector machine, optimal binary tree