Utilizing Dark Features for Iris Recognition in Less Constrained Environments

We propose a novel approach for iris recognition in less constrained environments that takes into account imaging noise arising from image capture outside the Depth of Field (DOF) of cameras. The proposed approach utilizes stable dark regions in iris images for recognition and does not rely on special hardware or on computationally expensive image restoration algorithms. We have employed a Gabor-based model to establish that stable features, which are not sensitive to defocus, correspond to regions in iris images with low gray-level intensity. We will also present an approach to identify stable bits from the iris code representation, which correspond to dark regions in the enrolled image. Only these stable bits are used for recognition. Experimental results based on 15,000 images with varying degree of defocus show that the proposed method achieves an average recognition performance gain of up to 6% over a conventional method that relies on the entire code representation for iris recognition.