June 20, 2012 -
The National Institute of Standards and Technology (NIST) recognizes 3M Cogent’s latent fingerprint algorithm as achieving the highest accuracy score.
In its latest Evaluation of Latent Fingerprint Technologies: Extended Feature sets (ELFT-EFS Evaluation #2) published last month, 3M Cogent achieved the highest accuracy score in the Image and Extended Feature (LE) subset of the NIST.
Matt Pepe, CEO of 3M Cogent, is very enthusiastic about the recognition. “The study demonstrated that our algorithm has the most matching candidates ranked in the number one position,” he said. “This is a key factor in improving productivity for our users and providing a scalable system over time in terms of accuracy.”
NIST is a non-regulatory federal agency within the U.S. Department of Commerce. Its mandate is to advance measurement science, standards and technology and publishes reports of results.
The latest NIST report shows the result of a latent matching capability test which is used by law enforcement agencies for national security and public safety. NIST evaluated the current state of the art in latent feature-based matching, by comparing the accuracy of searches using images alone with searches using different sets of features marked by experienced latent print examiners. These images are used for matching against fingerprints in a database. One of the purposes of ELFT-EFS is not a test of automatic EFS extraction but rather a test of data inoperability and how potentially useful such human marked features are when processed by an automated matcher.
NIST used additional extended feature in evaluating fingerprint images. The LE subset where 3M Cogent got the highest accuracy rating has additional extended features that include ridge quality maps, creases, dots, incipient ridges, ridge edge protrusions, and pores. These features are over the minutiae, ridge counts, cores and deltas, and pattern class included in other subsets.
According to Pepe, “3M Cogent’s state of the art algorithm is a result of 20 years of dedicated, applied technical research and development. We strive to push the accuracy and speed performance of our fingerprint identification technology.”
The algorithm used for the ELFT-EFS is now available as a system upgrade option for law enforcement customers globally.