Technology

Camera Develops Pictures With Algorithms, Not Lenses

March 31, 2014 | by Lisa Winter

Photo credit: Stork and Gill, 2014

Since the advent of photography, cameras have been getting progressively smaller, but there is one thing they all have in common: a lens. A new system developed by Rambus uses only a computer chip to create pictures. The process was described in a self-published paper which took the prize of “best paper” at Sencomm 2013. 

The light isn’t focused through the lens like on a traditional camera; it is done with very tiny a computer chip, just like the ones used in CMOS cameras. The image becomes etched into the glass, which the chip then scans and analyzes. The etching of the glass impacts how light is refracted through, which is then run through an algorithm to make sense of it. The design itself cannot be seen with our eyes, until the chip has converted it.

Currently, the chip is able to produce images that are 128x128, which is an admittedly low resolution. Though the pictures aren’t crisp and clear, they are able to be recognized, as evidenced by this image of the Mona Lisa. 

Impressively, the chip and the glass can fit comfortably on top of the head of a pin. Without the need for a lens, cameras will be able to get even smaller. This technology could be integrated into almost any other digital device, giving it the ability to capture visual information.

As far as applications for this technology go, the sky is pretty much the limit. There are innumerable uses for this device, though they can essentially be broken down into three categories:

Imaging Obviously, the camera’s primary function would be used for creating images. It would be able to create low- to mid-resolution pictures relatively cheaply where only one image is required. This could be used during safety testing, environmental disasters, or even used on the battlefield. The developers note that there will have to be decisions made regarding the design of the device, in order to be most useful for the application.

Motion Detection A small device capable of detecting motion could be safety mechanisms such as airbag deployment, security items such as motion lights, or the technology could even be integrated into toys to make them more interactive with the child.

Pattern recognition The camera could be used to read information encoded with a pattern, such as that hidden inside a barcode or a QR code. The applications of this could be widely varied, depending on what is needed.

Of course, with the possibility of such miniature cameras that can be implanted nearly anywhere, there will be obvious concerns of where security ends and invasion of privacy begins. 

All images credited to Stork and Gill

Tags