3D Camera with 12,616 Lenses

No, it’s not a get rich quick scheme by a company manufacturing lens cleaning tissues, but an idea for future camera technology from the brainy folk at Stanford University. A team led by Professor Abbas El Gamal are working on a camera with a multi-aperture image sensor. This is basically an image sensor with super-small pixels – several times smaller than the pixels on a regular camera. They are clustered together on the sensor chip in groups of 256 pixels, and each group has its own micro lens. It is like having a lot of cameras on a single chip; in effect the 3-megapixel chip the researchers are working on is equivalent to 12,616 separate cameras.

So far so good, but the really clever bit is that by selectively defocusing images captured by the cameras, the data can be processed to produce detailed 3D image maps of whatever it happens to be pointing at. Read more