Engineers build compact gigapixel camera

Researchers have previously developed cameras with very high resolutions, but such cameras are usually for research. However, a new gigapixel technology could eventually become suitable for consumers.

Photos with a resolution of tens of gigapixels had been made before, but these were composite images made with normal DSLRs. The US military has already had a camera with a few gigapixels developed and a telescope will have an advanced 3.2GP camera in a few years’ time. However, the camera technology developed by researchers from the Duke, California and Arizona universities and the company DistantFocus in America would eventually also be suitable for consumers.

For now, the camera is not ready for the consumer market; the device is about 50 centimeters deep and about 75 centimeters wide and high. However, the bulk of the device does not consist of optical elements and sensors, but of the necessary processors and electronics to form the 50 gigapixel image. That hardware has to merge the image that the separate 14 megapixel cameras produce into a single image. The camera uses an FPGA board that controls two cameras. The individual cameras have a resolution of 1384 by 3288 pixels and operate at 10fps.

The gigapixel camera uses a single objective lens that distributes the light among the small cameras. The composition of the camera resembles the facet eye of some insects. Each camera can be individually focused and controlled. In order not to miss any details, the ‘fields of view’ of the cameras always overlap a bit. The field of view of the complete camera, limited by the housing and lens, is 120 degrees horizontally and 40 degrees vertically.

The camera, co-developed for the American defense research branch Darpa, has the project name Aware. Several prototypes have already been shown, including the latest version with 226 cameras and a resolution of 2 gigapixels. an earlier version had 98 cameras and, after deducting the overlap, a resolution of 1 gigapixel. Images shot with the cameras can be viewed online see†

The camera should eventually be able to produce an image of a maximum of 50 gigapixels, or 50,000 megapixels. According to one of the developers, David Brady of Duke University, the further increase in computing power should make the camera smaller. The cooling of the components also leads to relatively large dimensions. In five years’ time, however, the technology would have progressed far enough to be used in consumer cameras.