Statistics of Visual Features in the Human Iris

Published

August 16, 2021

Author(s)

George W. Quinn, James Matey, Patrick J. Grother, Edward Watters

Abstract

In most current applications of iris recognition, matching is done by computer algorithms. The dominant algorithms are based on the work of John Daugman and are well understood because of the extensive analysis in the literature of iris2pi (the shorthand name for the Daugman algorithms). However, the internals of the dominant algorithms are not intuitive – the features that lead to the determination of a match between two iris images are not obvious to a casual observer. In forensic cases, evidence may be presented to non-technical people. Hence there is interest in the forensic community in human adjudication of matches between iris images, where the human relies upon easily understood visible features. This paper presents an analysis of the distribution of visible features in the human eye that may be of use in assessing the likelihood of a set of visible features occurring in two eyes by chance.

Citation

NIST Interagency/Internal Report (NISTIR) - 8386

Report Number

8386

NIST Pub Series

NIST Interagency/Internal Report (NISTIR)

Pub Type

NIST Pubs

Download Paper

https://doi.org/10.6028/NIST.IR.8386

Local Download

Keywords

Biometrics, Forensics

Citation

Quinn, G. , Matey, J. , Grother, P. and Watters, E. (2021), Statistics of Visual Features in the Human Iris, NIST Interagency/Internal Report (NISTIR), National Institute of Standards and Technology, Gaithersburg, MD, [online], https://doi.org/10.6028/NIST.IR.8386, https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=932452 (Accessed May 20, 2025)