International Biometric Group Delivers Iris Recognition Test Report to DHS

Test evaluated options from Iridian, LG, Oki and Panasonic


NEW YORK -- International Biometric Group announces the delivery of the Independent Testing of Iris Recognition Technology (ITIRT) Final Report to the US Department of Homeland Security. The ITIRT Final Report presents results from a groundbreaking evaluation of iris recognition accuracy, interoperability, and usability.

The ITIRT Final Report can be downloaded at http://www.biometricgroup.com/ITIRT.

Iris recognition has become an increasingly viable solution for border control, immigration, registered travelers, access control, and national ID initiatives. ITIRT represents the largest independently conducted test of iris recognition executed to date. Prior to ITIRT, independently generated data on iris recognition performance - a critical deployment consideration - was limited to results from small-scale tests.

ITIRT evaluated performance of the following state-of-the-art iris recognition hardware and software:

- Iridian KnoWho OEM SDK - LG IrisAccess 3000 - Oki IRISPASS-WG - Panasonic BM-ET300

Over 100,000 iris images were acquired through ITIRT devices over a several-week period. Dozens of images were acquired from each of 1,288 subjects varying in age and ethnicity. From this data, over 2 billion data points were generated for analysis. ITIRT presents results essential to a full understanding of biometric performance, including the following:

- False accept and false reject rates - Failure to enroll and failure to acquire rates - Transaction duration and level of effort - Performance over time

The Report also examines interoperability across iris recognition devices, a critical consideration for many deployers. Images from each iris recognition device were cross-compared to determine whether images captured through a given device could be matched against those captured through a different device.

The Report further examines whether certain test subjects are more likely than others to be falsely accepted or rejected by iris recognition devices, and whether failed enrollments occur across multiple devices.