While the biometrics industry touts the reliability of using scans to tie people to their identification records, new research suggests that the aging process may have an unintended effect on biometrics.
As reported on Nature’s website, Kevin Bowyer, a professor of computer science at Indiana’s University of Notre Dame, did research that compared iris scans taken about a month apart and iris scans that were taken one, two or three years apart. The match failure rate when comparing scans over longer periods of time grew by 153% over the three-year span. This shows that the natural aging process may change the appearance of an iris.
This study suggests that iris-recognition may not be the end all, be all in terms of identification verification and may require the industry to come up with algorithms that are age-resistant.
Read more here.