Automatic segmentation and recognition of iris images: With special reference to twins

Abstract

This paper describes an automatic approach for iris segmentation and recognition with focus on twins. The technique entails localizing and segmenting the iris, followed by iris normalization and obtaining distinctive features. Lastly, iris templates are matched to realize one to one and one to many recognition in twins. Further, effect of various template sizes on the accuracy and memory requirement are studied. To our knowledge, this is the first iris based authentication with special emphasis on twins.