Software Developed to Match Police Sketches, Mug Shots
MSU doctoral student Brendan Klare, who was one of the leaders of the research team, said,
“We’re dealing with the worst of the worst here. Police sketch artists aren’t called in because someone stole a pack of gum. A lot of time is spent generating these facial sketches so it only makes sense that they are matched with the available technology to catch these criminals.”The algorithms MSU developed compare eye, nose, and chin shapes, as well as other facial structures, between pairs of images.
The software was 45 percent correct when comparing sketches against a database of more than 10,000 mug shots. The sketches used were from actual crimes in which the criminal was later identified.
While 45 percent is a small percentage when compared about the 100 percent accuracy of a DNA match, DNA isn't always available, and anyone who has done research knows that eyewitness testimony is often flawed.
The MSU team intends to field test the system in about a year.