August 21, 2013
Aging Effect Not An Issue For Using The Iris For Biometric Identification
April Flowers for redOrbit.com - Your Universe Online
Biometric researchers at the National Institute of Standards and Technology (NIST) have determined that no consistent change occurs in the distinguishing texture of the iris for at least a decade. The research team used data from thousands of frequent travelers enrolled in an iris recognition program at the US/Canada border. The findings, published in an IREX VI NIST Interagency Report, will inform identity program administrators on how often iris images should be recaptured to maintain accuracy.
Researchers looking for biometric identifiers other than fingerprints have been drawn to irises for decades, believing their one-of-a-kind texture meets the stability and uniqueness requirements for biometrics. Recent studies, however, have called that belief into question. Over a three-year period, a study looked at 217 participants, finding the recognition of the subjects' irises became increasingly difficult, consistent with an aging effect.
The NIST team used several methods to evaluate iris stability in order to learn more.
The scientists examined anonymous data from millions of transactions from NEXUS, the joint US and Canadian program used by frequent travelers to quickly cross the Canadian border. The irises of NEXUS members are placed into the system with an iris camera. Each time they cross the border, their irises are scanned and matched to system files. A larger, less well-controlled set of anonymous statistics collected over a six-year period were also examined.
The NIST team found no evidence of a widespread aging effect in either large-population study, according to Biometric Testing Project Leader Patrick Grother. A computer model developed by the team estimates after the initial enrollment, iris recognition of average people will typically be useable for decades.
"In our iris aging study we used a mixed effects regression model, for its ability to capture population-wide aging and individual-specific aging, and to estimate the aging rate over decades," said Grother. "We hope these methods will be applicable to other biometric aging studies such as face aging because of their ability to represent variation across individuals who appear in a biometric system irregularly."
The research team reanalyzed the images from the earlier study of 217 participants that evaluated the population-wide aspect. An increase in false rejection rates - the original, enrolled images taken in the first year of the study did not match those taken later – was reported in these studies. The rejection results did not necessarily demonstrate the iris texture itself was changing, although the numbers were high. A separate study conducted by another research team identified the primary cause of the rejections as pupil dilation, which led the NIST team to investigate.
In the original pool of subjects, the NIST researchers showed dilation increased in the second year of the test and decreased the next. They were not able to determine a cause, however. When the dilation effect was accounted for, no change was observed in the texture or aging effect. Some iris cameras already in use normalize dilation by using shielding or by varying the illumination.
The Iris Exchange Program (IREX) was established in 2008 by NIST to give quantitative support to iris recognition standardization, development and deployment.