Self-Organizing Map Application for Iris Recognition

Main Article Content

Anna Sedrak Hovakimyan Siranush Gegham Sargsyan Arshak Nazaryan

Abstract

Human iris is  a good subject of biometrical identification, since  iris patterns are unique like fingerprints. Iris is well protected against damage, unlike fingerprints, which can be harder to recognize after years of certain types of manual labor.

A problem of iris recognition is considered in the paper. In machine learning, pattern recognition is the assignment of a label to a given input value. Pattern classification is an example of pattern recognition: it attempts to assign each input value to one of a given set of classes. Nowadays various techniques are used for this purpose, and in particular artificial neural networks.

For iris recognition problem solving  Kohenen Self Organizing Maps are suggested to use. The software for iris recognition is developed  which is customizable and allows to select the appropriate parameters of the neural network to obtain the most satisfactory results. The developed Self-Organizing Map Library of classes can be used for various kinds of object classification problem solving as well as for any problems suitable to solve with Self-Organizing Maps.

Article Details

How to Cite
HOVAKIMYAN, Anna Sedrak; SARGSYAN, Siranush Gegham; NAZARYAN, Arshak. Self-Organizing Map Application for Iris Recognition. Journal of Communications and Computer Engineering, [S.l.], v. 3, n. 2, p. 10-13, mar. 2014. ISSN 2090-6234. Available at: <http://m-sciences.com/index.php?journal=jcce&page=article&op=view&path%5B%5D=760>. Date accessed: 25 may 2017. doi: https://doi.org/10.20454/jcce.2013.760.
Section
Articles