Digital Camera Add-On Means the Light's Fantastic
KaleidoCamera is developed by Alkhazur Manakov of Saarland University in Saarbrücken, Germany, and his colleagues. It attaches directly to the front of a normal digital SLR camera, and the camera's detachable lens is then fixed to the front of the KaleidoCamera.
After light passes through the lens, it enters the KaleidoCamera, which splits it into nine image beams according to the angle at which the light arrives. Each beam is filtered, before mirrors direct them onto the camera's sensor in a grid of separate images, which can be recombined however the photographer wishes.
This set-up allows users to have far more control over what type of light reaches the camera's sensor. Each filter could allow a single colour through, for example, then colours can be selected and recombined at will after the shot is taken, using software. Similarly, swapping in filters that mimic different aperture settings allows users to compose other-worldly images with high dynamic range in a single shot.
And because light beams are split up by the angle at which they arrive, each one contains information about how far objects in a scene are from the camera. With a slight tweak to its set-up, the prototype KaleidoCamera can capture this information, allowing photographers to refocus images after the photo has been taken.
Roarke Horstmeyer at the California Institute of Technology in Pasadena says the device could make digital SLR photos useful for a range of visual tasks that are normally difficult for computers, like distinguishing fresh fruit from rotten, or picking out objects from a similarly coloured background. "These sorts of tasks are essentially impossible when applying computer vision to conventional photos," says Horstmeyer.
The ability to focus images after taking them is already commercially available in the Lytro – a camera designed solely for that purpose. But while Lytro is a stand-alone device which costs roughly the same as an entry-level digital SLR, KaleidoCamera's inventors plan to turn their prototype into an add-on for any SLR camera.
Manakov will present the paper at the SIGGRAPH conference in Anaheim, California, this month. He says the team is working on miniaturising it, and that much of the prototype's current bulk simply makes it easier for the researchers to tweak it for new experiments.
"A considerable engineering effort will be required to downsize the add-on and increase image quality and effective resolution," says Yosuke Bando, a visiting scientist at the MIT Media Lab. "But it has potential to lead to exchangeable SLR lenses and cellphone add-ons."
In fact, there are already developments to bring post-snap refocusing to smartphone cameras, with California-based start-up Pelican aiming to release something next year.
"Being able to convert a standard digital SLR into a camera that captures multiple optical modes – and back again – could be a real game-changer," says Andrew Lumsdaine of Indiana University in Bloomington.
By Hal Hodson, New Scientist
