ISU researchers develop software to detect hidden messages

Andrea Fier

A picture is worth a thousand words, but two ISU researchers are working to find out what pictures also hide a thousand secrets.

Jennifer Davidson, associate professor of mathematics, and Clifford Bergman, professor of mathematics, have been working together to design software able to identify computer images that contain hidden data.

The technology, known as steganography, allows individuals to use software available on the Internet to insert hidden data into seemingly normal image files.

Davidson said at its most basic level, the technology works on mathematical principles familiar to undergraduates in mathematics or engineering, thus discovering hidden images is also a matter of number crunching.

“The way the data is found is based on mathematics and image processing,” Davidson said.

Although a person would need specific software to be able to do steganography, there are sites on the Internet where software can be downloaded for free.

“It was fashionable about five to six years ago to write these programs,” Bergman said.

There isn’t really any legal use for steganography outside of espionage, Bergman said.

“The example I always use, is for a spy in enemy territory,” Bergman said. “The trick is to send a message so the enemy doesn’t know that they sent anything.”

The spy could, for example, take a photo of himself in front of the Eiffel Tower, then use steganography to hide data within the photo a receiver could then extract. A third party observer or interceptor would have no clue that sensitive information was being transferred; the spy would look like a tourist, Bergman said.

The researchers have been working with the Iowa Division of Criminal Investigation in the area.

“[Steganography] doesn’t really have any legal purpose, per se, which is why law enforcement is so interested in it,” Davidson said.

Steganography could be used to hide information within a simple picture of a can opener on eBay, Bergman said. The information could be passed back and forth with no suspicion.

“There are so many images, and then how do you check it? We’re writing the software to check it,” Bergman said.

The problem they are working on with the software is designing it so it would be able to detect all kinds of steganography techniques quickly, Bergman said.

“It’s just not fast enough; we really just started on this,” Bergman said.