Fingerprint identification matures to higher level

Crime scene investigators may be able to accelerate latent fingerprint identification with new automated identification technology.

Crime scene investigators may be able to speed latent fingerprint identification with new automated identification technology.

Recent tests at the National Institute of Standards and Technology (NIST) showed prototype systems performing better than expected, with half of the prototypes accurate at least 80 percent of the time and one with a near perfect score. These prototype systems would automate manual processes, freeing trained examiners to spend more time on very difficult images.

The research was funded by the Homeland Security Department’s Science and Technology Directorate and the FBI’s Criminal Justice Information Services Division. The prototype systems are based on an emerging technology named Automatic Feature Extraction and Matching.

In the evaluation, researchers used a data set of 835 latent prints and 100,000 fingerprints that have been used in real case examinations. The software extracted the distinguishing features of the latent prints, then compared them against 100,000 fingerprints. For each print, the software provided a list of 50 candidates that the fingerprint specialists compared by hand. Most identities were found in the top 10.

The technology supports counterterrorist or law-enforcement identification systems, such as the FBI's Integrated Automated Fingerprint Identification System, which compares latent prints against the 55 million sets of ten-print cards that are taken when a person is arrested.

NIST biometric researchers assessed prototypes that eight vendors are developing. In order of performance, the most accurate prototypes were furnished by NEC Corp., Cogent Inc., SPEX Forensics, Inc., Motorola, Inc. and L-1 Identity Solutions. Results ranged from nearly 100 percent for the most accurate product to around 80 percent for the last three listed.

Fingerprint identification accuracy of the technology was affected by the number of distinguishing features in a latent print and the quality of the image data.

A report on the results is titled "An Evaluation of Automated Latent Fingerprint Identification Technologies."