Iris Recognition on Images Reconstructed with Gradient-based Algorithm
Tijana Radojicic, Milena Lipovina Božović, Nina Blagojevic · 2020
Iris recognition is a procedure used for identifying persons based on unique patterns that are placed around pupil of the human eye. The aim is to localize the inner and outer boundaries of the iris in an image of the eye. Further, software detects and removes eyelids, eyelashes and specular reflections that often interfere with parts of the iris. In this paper we are discussing possibilities of different iris recognition software to perform recognition on images that are subject to compressive sensing approach. Namely, image is intentionally damaged according to the Compressive Sensing. Damaged (missing) image pixels are recovered by applying recovery algorithm. For the reconstruction the gradient-based algorithm is used. The reconstruction is tested considering different percentage of the pixels as damaged (missing).