Algorithm for Extraction of the Iris Region in an Eye Image

The iris recognition system depends heavily on the accuracy of the selected region of the iris. The paper considers the task of selecting the iris region in an eye image. To solve the problem, an algorithm has been developed that allows selecting the iris inner and outer boundaries in an acceptable time. The algorithm differs from the known ones by resistance to clutters in the form of glare, eyelashes, shadows and eyelids, as well as by high-speed operation, and allows working with images recorded under different conditions. Static results are given that determine the accuracy of the results and the time of the operation algorithm. To analyze the effectiveness of the proposed algorithms, the CASIA V4 database consisting of 52034 images of various positions and quality was used.