Feature extraction and recognition of iris based on biorthogonal multiwavelets
基于双正交多小波的虹膜特征提取与识别

Biorthogonal multiwavelets filter characterized with self-affine was proposed to extract iris texture feature and local with global feature was applied to recognize iris. After using biorthogonal multiwavelets filter to process iris images, local coarse quantization encoding was adopted in the low frequency parts of coefficients, and Hamming distance was taken as the classifier. When the Hamming distance was uncertain of its decision due to the influence of eyelids, eyelashes and iris deformations, mean and variance were extracted from eoefficients of multiwavelets transform, and Euclidean distance of covarianee reciprocal with weight value was designed as the classifier. The results show that this approach is able to identify iris quickly and reliably.