New Fingerprint Identification Technique Devised

US experts from the National Institute of Standards and Technology (NIST) have recently announced that they are in the latter stages of developing a new investigation technique that will help forensics specialists in the manual portion of the latent fingerprint identification. This would free up a lot of time for the professionals, which could then spend more of their precious minutes analyzing images that computer software will never be able to thoroughly scan. According to the people at NIST, half of the prototypes they have created thus far have been at least 80 percent accurate in identifying the correct fingerprints, whereas one got a nearly perfect score from the first attempt.

Latent fingerprints are the ones left behind when a villain touches something at their crime scene. They can be found on everything, from glasses and windows to furniture tops and door knobs. However, they are left on textured materials at times, which makes identifying their traits a painstakingly difficult procedure. This process involves the forensics expert carefully marking the area where the full or partial fingerprint has been left, and then meticulously drawing the contours of the ridges, where they join together, and where they separate.

The data is then inserted in a local or federal law-enforcement database, which contains millions of fingerprint sets, taken at arrests over the years. One such database is known as the Integrated Automated Fingerprint Identification System (IAFIS), and is operated by the US Federal Bureau of Investigation. This one alone contains 55 million sets of ten-print cards, which can be shuffled through with relative ease.

The new system developed at NIST, dubbed the Automatic Feature Extraction and Matching (AFEM), is a biometric research device, which is able to perform all the tasks associated with latent fingerprint identification faster than a forensics expert. Prototypes from constructors NEC Corp., Cogent Inc., SPEX Forensics Inc., Motorola Inc. and L1 Identity Solutions have been assessed and made to identify a number of 935 latent prints, against a 100,000-large database.

“While the testing has demonstrated accuracy beyond pre-test expectations, the potential of the technology remains undefined and further testing is required. In the future, we will look at lower quality latent images to understand the technology’s limitations and we will support development of a standardized feature set that extends the one currently used by examiners for searches,” Patrick Grother, who is a computer scientist at NIST, said.