The emergence of software that enables hiding digital code in photographic images has given criminals a new way to disguise their activities, but researchers are working to give law enforcement advanced tools of their own to sniff out such code.
Researchers at the Midwest Forensics Resource Center at the Department of Energy's Ames Laboratory and Iowa State University are working on tools for what's called steganalysis. Steganography refers to the concealing of files in other files, such as JPEG images (the colors of a picture might be slightly changed to hide data).
"There's lots of room within a digital image to hide data," says Clifford Bergman, an Iowa State math professor and researcher, in a statement. Bergman and his fellow researchers are searching for statistical measures to spot hidden data and are using a pattern recognition system called an artificial neural net to tell stego images from clean ones. The system in one early test was able to spot 92 percent of the stego images in a sample database.
The Iowa Department of Criminal Investigation is field-testing the program and a GUI is in the works to make the system easier to use.
The hope, Bergman says, is to come up with strong enough algorithms that forensic experts can collect evidence that will be admissible in court.