13 July 2012
Iris recognition shows its age
by Will Parker
It was generally assumed that the iris was a "stable" biometric over a person's lifetime but new research from the University of Notre Dame shows the iris is susceptible to an aging process that causes recognition performance to degrade slowly over time. Iris recognition is already used in various airports and border crossings and the largest application of iris biometrics (known as the Unique ID program) is currently underway in India. The research paper was presented at the recent IEEE Computer Society Biometrics Workshop.
"The biometric community has long accepted that there is no 'template aging effect' for iris recognition, meaning that once you are enrolled in an iris recognition system, your chances of experiencing a false non-match error remain constant over time," explained Notre Dame's Kevin Bowyer. "Our experimental results show that, in fact, the false non-match rate increases over time, which means that the single enrollment for life idea is wrong."
The false match rate is how often the system says that two images are a match when in truth they are from different persons. The false non-match rate is how often the system says that two images are not a match when in truth they are from the same person.
Bowyer cites several reasons why misconceptions about iris biometric stability have persisted. "One reason is that because it was believed from the early days of iris recognition that there was no template aging effect, nobody bothered to look for the effect," he said. "Also, only recently have research groups had access to image datasets acquired for the same people over a period of several years."
In the new study, Bowyer analyzed a large dataset with more images acquired over a longer period of time. For one group of people in their dataset, they were able to analyze a year-to-year change over three successive years.
Bowyer believes any issues with recognition systems can be overcome, however. "I do not see this as a major problem for security systems going forward," he said. "One possibility is setting up a reenrollment interval. Another possibility is some type of 'rolling re-enrollment,' in which a person is automatically re-enrolled each time they are recognized. And, in the long run, researchers may develop new approaches that are 'aging-resistant.'"
Source: University of Notre Dame