Photographs are turning into the digital equivalent of fingerprints, allowing law enforcement to search through a collection of images to help track down the identity of photo-taking criminals, such as smartphone thieves and child pornographers. Prior investigation has shown that a digital photo can be paired with the very camera that took it by examining the unique noise pattern that its sensor imprints onto photos, and now researchers have begun applying that to social networks, grabbing photos from Facebook, Flickr, Tumblr, Google+, and personal blogs to see whether one individual image could be matched to a specific user's account.
In a paper published earlier this year, researchers say that they were able to match a photo with a specific person 56 percent of the time in their studied circumstance — examining 10 different people's photos found on two separate websites each. The researchers, Riccardo Satta and Pasquale Stirparo from the European Commission's Institute for the Protection and Security of the Citizen, acknowledge that this performance is far from perfect, but they argue that it's still much better than random guess and could at the least help to pinpoint persons of interest in a criminal investigation. Analayzing photos by what's known as their "sensor pattern noise" is still a relatively new field, however, so those figures are likely to rise with more research.