MIT Presents Computational Photography Processor

MIT’s Microsystems Technology Laboratory presents a processor optimized for computational photography tasks. In one example, the chip can build a 10MP HDR picture combined out of 3 frames "in a few hundred milliseconds". Other examples include combination of two pictures with and without flash, adaptive brightening the dark parts of the picture, noise reduction with a bilateral filter, and more.

The chip offers a hardware solution to some important problems in computational photography, says Michael Cohen at Microsoft Research in Redmond, Wash. "As algorithms such as bilateral filtering become more accepted as required processing for imaging, this kind of hardware specialization becomes more keenly needed," he says.

The power savings offered by the chip are particularly impressive, says Matt Uyttendaele, also of Microsoft Research. "All in all [it is] a nicely crafted component that can bring computational photography applications onto more energy-starved devices," he says.

The work was funded by the Foxconn Technology Group, based in Taiwan