Robust Iris Segmentation Algorithm in Non-Cooperative Environments Using Interleaved Residual U-Net

Iris segmentation plays an important and significant role in the iris recognition system. The prerequisite for accurate is correctness of segmentation. However, efficiency robustness traditional methods are severely challenged a non-cooperative environment because unfavorable factors, instance, occlusion, blur, low resolution, off-axis, motion, specular reflections. All above factors seriously reduce accuracy In this paper, we present novel algorithm that localizes outer inner boundaries image. We propose neural network model called “Interleaved Residual U-Net” (IRUNet) semantic mask synthesis. K-means clustering applied to select saliency points set order recover boundary iris, whereas border recovered by selecting another on side mask. Experimental results demonstrate proposed can achieve mean IOU value 98.9% 97.7% estimation, respectively, which outperforms existing approaches challenging CASIA-Iris-Thousand database.