Researchers at Michigan State University have created a set of algorithms and software that could greatly simplify crimefighting.
The new system automatically matches hand-drawn facial sketches to mug shots that are stored in law enforcement databases.
Traditionally, forensic artists' work has often been hit or miss. A police sketch isn't always a realistic depiction of a suspect. And manually searching through mug shots for a match eats up vital investigation time.
The MSU team is led by Professor Anil Jain and doctoral student Brendan Klare. The project is the first large-scale attempt to match forensic sketches with photographs. And the implications are huge.
“We’re dealing with the worst of the worst here,” Klare said. “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 software works by matching high-level features in the sketch and the photo, such as structural distribution and the shape of the eyes, nose and chin, Jain said.
Although there are a few commercial forensic sketching software programs available, the resulting sketches are usually less accurate than those drawn by a trained forensic artist.
“We improved significantly on one of the top commercial face-recognition systems,” Klare said. “Using a database of more than 10,000 mug shot photos, 45 percent of the time we had the correct person.”
The team plans to field test the software next year sometime.