FEATURES OF THE IRIS IMAGE PRE-PROCESSING ALGORITHM

Abstract

The article presents a brief overview of the not numerous issues of biometric identification based on iris. Usually, the original image corrupted by equipment scanning noise, discretization or data transmission channels, whe-reby images are uneven brightness and contrast. This leads to breaks in linear objects, masking of the analyzed complex objects, which leads to a large number of errors and failures, especially automatic identification and recog-nition systems. Preprocessing includes: equation of the general luminance background image, the elimination of the original image high-frequency noise and various arti-facts, contrast binary image and other functional trans-formations. Output description of the original informa-tion must be maximally adapted to storage, transmission and analysis of all possible solutions. But with all these transformations one of the most important indicators should be to preserve the maximum number of special «key» elements of the original information. Identified promising areas for further research in the field of iris image recognition. As the test data used the image of an open database.