Iris Image Deblurring Based on Refinement of Point Spread Function

Blurred iris images are inevitable during iris image acquisition due to limited depth of field and movement of subjects. The blurred iris images lose detailed texture information for accurate identity verification, so this paper proposes a novel iris image deblurring method to enhance the quality of blurred iris images. Our method makes full use of the prior information of iris images. Firstly, benefiting from the properties of iris images, a set of initialization methods for point spread function (PSF) is proposed to obtain a better start point than that of common deblurring methods. Secondly, only the most reliable iris image regions which are obtained by structure properties of iris images are used to refine the initial PSF. Finally, the more accurate PSF is used to reconstruct the clear iris texture for higher accuracy of iris recognition. Experimental results on both synthetic and real-world iris images illustrate that the proposed method is effective and efficient, and outperforms state-of-the-art iris image deblurring methods in terms of the improvement of iris recognition accuracy.