MegaMatcher On Card Fingerprint Algorithm For Smart Cards Succeeds In NIST MINEX III Evaluation
The Neurotechnology MegaMatcher On Card algorithm for fingerprint matching on a smart card is compliant with the same NIST MINEX III criteria used to evaluate much more resource-intensive PC-based algorithms.
VILNIUS, Lithuania, Nov. 21, 2016 /PRNewswire/ -- Neurotechnology, a provider of high-precision biometric identification and object recognition technologies, today announced that the MegaMatcher On Card fingerprint matching algorithm implementation for smart cards was submitted for the NIST Minutiae Interoperability Exchange (MINEX) III evaluation and successfully passed all the mandatory steps of the evaluation protocol. The MINEX III evaluation is a continuing test of fingerprint templates used to establish compliance of template generators and template matchers for the SP 800-76-2 Biometric Specifications for Personal Identity Verification (PIV) U.S. Government program for interoperability, native accuracy and ANSI INCITS 378 standard compliance requirements. Because there is not yet a specific NIST MINEX evaluation for fingerprint matching on smart cards, the template matcher named Neurotechnology+0204 was evaluated as a PC-based algorithm. It passed the same requirements as algorithms that usually run on much more resource-intensive processors. MegaMatcher On Card implements the exact same fingerprint matching technology as Neurotechnology+0204, which is so efficient it can also run on the tiny microcontrollers in smart cards. "The algorithms submitted for NIST MINEX III evaluation are normally designed to use the resources of standard PCs and servers," said Antonello Mincone, business development manager for Neurotechnology. "Meeting the same criteria with our MegaMatcher on Card algorithm affirms to our customers that high quality, accurate and reliable fingerprint recognition can run efficiently on very small, secure microcontrollers embedded in smart cards or other tokens used in fields like National ID projects or payment applications products." While being proven very accurate by the NIST Minex III criteria, this algorithm is also fast. For example, it can verify the true card holder's fingerprint data in less than 0.5 seconds, on average, on a NXP P60 chip. Moreover, the library size fits in less than 16 Kb, including Java level on Java Card OS. An average size fingerprint template with 48 minutiae and mandatory public data may use less than 256 bytes for persistent storage.