Error Rates in Fingerprint Matching

with Jennifer L. Mnookin (UCLA), Philip J. Kellman (UCLA), Itiel Dror (CCI), Dave Charlton (CCI), Gennady Erlikhman (UCLA) and Patrick Garrigan (SJU)

Funding: US National Institute of Justice (PI: Jennifer L. Mnookin; co-PI: Philip J. Kellman, Itiel Dror )

For which of the above pairs is it easier to match the latent to the tenprint? Why is matching one of the pairs a lot easier?

Goal

The example above shows that visual features play an essential role in deciding the difficulty of task for matching latents to tenprints. More often than not the fingerprint examiners are presented with partial or degraded latent prints. The goal of this project is to identify visual features in a given pair of fingerprints that determine the difficulty of match for that pair.

Importance

Even though fingerprint evidence has been used in court for more than hundred years not much scientific literature is present on the error rates of the task of fingerprint matching. In this project we also plan to investigate the accuracy and reliability of latent fingerprint identification.

Fingerprint chocolates from Los Angeles Police Department

Related Links 
  • From UCLA's newsroom: 
    --Major federal grant awarded to UCLA-based project in collaboration with Cognitive Consultants International