My first taste of high tech photo detective stuff came via Blade Runner, when Blade Runner Rick Deckard (Harrison Ford) zooms in and enhances a photo to work out the location of rogue androids.

Now in a case of tech mirroring science fiction, boffins from York University have worked out how to capture images of bystanders captured in the reflections in the eyes of people nearby posing for photographs. It's a clever technique that could play a huge role in fighting crime in the near future.

The technology works like thus. The human cornea has a black mirror-like surface that is surprisingly good at reflecting objects and people. The researchers worked out that digital photography if sufficiently high-resolution will allow them to zoom in and obtain usable images from the reflections in a person's eye.

Reflections of human faces were found by zooming in on hi-res photos of eyes of people, which revealed pixelated (but crucially still identifiable) images of faces using a 38 megapixel camera. Using a range of test subjects, over 70% of the reflected faces were able to be identified.


With high-res multi-megapixel cameras becoming increasingly commonplace in smartphones and steady improvements being made with biometric facial recognition technologies, the potential for crime fighting could be huge.