Iris Localisation and segmentation using Convolutional neural network

Old methods perform very well on a constrained environment, but in a less constrained environment, they perform poorly due to the eyelashes, eyelids. This paper proposes an efficient way to localize the region of interest that contains the iris by using the convolution neural network, and then the localization of the iris using the circular Hough transform. The model evaluated under CASIA Interval V4, IITD, MMU1, and UBIRIS V1.