EXAMINING INTRAVISIT IRIS STABILITY VISIT 4 Michael Boyle
EXAMINING INTRA-VISIT IRIS STABILITY (VISIT 4) Michael Boyle, Kyle Hurd, Mike Lozevski, Matthew Maclennan, Shreya Mohandas, Deep Patel, Chloe Pina, Alex Wolowiecki, Taylar Worrell, Kevin Chan, Steve Elliott, Ben Petry
WHAT IS BIOMETRICS? • Biometrics can be defined as “the automated recognition of an individual based on physiological or behavioral characteristics for the purposes of identification and/or verification” [1].
WHY IS BIOMETRICS IMPORTANT? • Security • Banks • Schools • Airports • Corporations • Solve inherent problem to token/password authentication • Unique identification that cannot be stolen, lost, or forged
BIOMETRICS • • Can be either physiological or behavioral Examples • • • Fingerprint Hand Geometry Facial Recognition Retina Scanners Iris Recognition Gait
IRIS RECOGNITION The benefits to implementing iris recognition as a primary biometric validator and identifier: • Unique to every person • Unobtrusive • Fast • Well protected from external forces • More resistant to change
STABILITY • It has been assumed that iris matching scores will remain relatively stable over time [2] • The research question is whether the iris is stable over a single visit of 10 or fewer minutes
WHAT IS AN IRIS? • The iris is the colored part of the eye in between the pupil and the sclera [3]
STABILITY SCORE INDEX •
AGING • Iris aging is a definitive change in the iris texture pattern due to human aging • Aging is one concern for implementing iris recognition • Research has shown that as an individual ages, they are less likely to match with the template they originally enrolled with • This phenomenon is known as template aging
RESEARCH QUESTION • This research will assess whether or not the template aging affect, as well as a possible iris aging, are an influential factor in the overall change in the Stability Score Index over a period of 10 or fewer minutes.
WAYS OF ANALYZING BIOMETRIC PERFORMANCE • ROC • DET • Zoo menagerie
DEFINE STABILITY OF THE IRIS • Iris recognition results can change due to many variables • Iris stability describes to what degree an iris’s score can vary • Used for more accurate individual acceptance tolerances
RESULTS
VISIT 1 AGE GROUPS
VISIT 1 GENDER
VISIT 1 – SELF DISCLOSED ETHNICITY
ANALYSIS • After running the different groupings of data sets through Minitab, using graphical summaries and Kruskal-Wallis Tests.
GROUP 1 • Alpha level of 0. 05 is less than the p value of 0. 841, fail to reject the null hypothesis. There wasn’t a statistical difference between the SID scores (H = 0. 35, p = 0. 814), with a mean of 0. 14973. This result means that there is stability in the iris.
GROUP 2 • Alpha level of 0. 05 is less than the p value of 0. 663, fail to reject the null hypothesis. There wasn’t a statistical difference between the SID scores (H = 0. 82, p = 0. 663), with a mean of 0. 14683. This result means that there is stability in the iris.
GROUP 3 • Alpha level of 0. 05 is less than the p value of 0. 343, fail to reject the null hypothesis. There wasn’t a statistical difference between the SID scores (H = 2. 14, p = 0. 343), with a mean of 0. 14973. This result means that there is stability in the iris.
GROUP 4 • Alpha level of 0. 05 is less than the p value of 0. 569, fail to reject the null hypothesis. There wasn’t a statistical difference between the SID scores (H = 1. 13, p = 0. 569), with a mean of 0. 15878. This result means that there is stability in the iris.
ANALYSIS • After running the different groupings of data sets through the Minitab software, using graphical summaries and Kruskal-Wallis Tests, there was no significant difference in the individual SSI scores as well as a difference between each of the groupings’ significance values.
RESULTS VISIT 1 N H DF P Group 1 60 0. 35 2 0. 814 Group 2 60 0. 82 2 0. 663 Group 3 60 2. 14 2 0. 343 Group 4 60 1. 13 2 0. 569 There was not a statistically significant difference between the median of the groupings, as indicated in the summary table. For this data, we can conclude that the iris is stable in this visit.
CONCLUSION • There was not a statistically significant difference between the median of the groupings, as indicated in the summary table. For this data, we can conclude that the iris is stable in this visit.
FUTURE WORK • With these finding we would want to extend the period of time that the iris’ are tested over to prove they continue to be stable. Along with this we would like to see if the stability of the iris starts to decrease over extended period of time.
BIBLIOGRAPHY [1] Woodward Jr, J. D. , Horn, C. , Gatune, J. , & Thomas, A. (2003). Biometrics: A look at facial recognition. RAND Corp, Santa Monica, CA. [2] Daugman, J. (2004). How iris recognition works. Circuits and Systems for Video Technology, IEEE Transactions on, 14(1), 21 -30. [3] Structure of the Eye, http: //www. uofmhealth. org/health-library/tp 9807 [4] Baker, S. E. , Bowyer, K. W. , & Flynn, P. J. (2009). Empirical evidence for correct iris match score degradation with increased time-lapse between gallery and probe matches. In Advances in Biometrics (pp. 1170 -1179). Springer Berlin Heidelberg. [5] Tome-Gonzalez, P. , Alonso-Fernandez, F. , & Ortega-Garcia, J. (2008, September). On the effects of time variability in iris recognition. In Biometrics: Theory, Applications and Systems, 2008. BTAS 2008. 2 nd IEEE International Conference on (pp. 1 -6). IEEE. [6] History of Biometrics. (n. d. ). Retrieved February 20, 2015, from http: //www. biometricupdate. com/201501/history-of-biometrics [7] Iris ID - Iris Recognition Technology : Iris Recognition Technology. (n. d. ). Retrieved February 20, 2015, from http: //www. irisid. com/irisrecognitiontechnology [8] Adler, F. H. , Physiology of the Eye (Chapter VI, page 143), Mosby (1953) [9] Daugman, J. (2004). How iris recognition works. Circuits and Systems for Video Technology, IEEE Transactions on, 14(1), 21 -30. [10] Daugman, J. (2006). Probing the uniqueness and randomness of Iris. Codes: Results from 200 billion iris pair comparisons. Proceedings of the IEEE, 94(11), 1927 -1935 [11] Doddington, G. , Liggett, W. , Martin, A. , Przybocki, M. , & Reynolds, D. (1998, November). Sheep, goats, lambs and wolves: an analysis of individual differences in speaker recognition performance. In the International Conference on Spoken Language Processing (ICSLP), Sydney. [12] O'Connor, K. J. (2013). Examination of stability in fingerprint recognition across force levels, MS. Thesis, Purdue University, West Lafayette, IN.
- Slides: 26