A New local and Global Model to Iris Recognition

摘要

Traditional iris recognition systems transfer iris images to polar coordinates, normalize the images and achieve rotation invariance by rotating the feature vector. In order to decrease the complexity of the typical iris recognition method, we propose a new method of iris recognition based on global and local model that are extracted from preprocessed iris image without normalizing. Firstly, we applied a bank of no-tensor product wavelet filters to extract the global features of the iris. Secondly, we used a SIFT method to extract the local features points of the selected regions. Finally, we tested the similarity distances of local and global features with different weights. Experimental results show that the proposed method in this paper has the recognition accuracy of 99.5% when the equal error rate is 0.94%. Without normalizing the iris images, the proposed approach can obtain very good recognition performance.