Friday, December 16, 2011

New standard aims to accelerate image processing

It goes without saying that computer vision has become an essential ingredient of many modern systems, where it has been used for numerous purposes including gesture tracking, smart video surveillance, automatic driver assistance, visual inspection, and robotics.

Many modern consumer computer-based devices -- from smart phones to desktop computers -- can be capable of running vision applications, but to do so, they often require hardware-accelerated vision algorithms to enable them to work in real time.

Consequently, many hardware vendors have developed accelerated computer vision libraries for their products: CoreImage by Apple, IPP by Intel, NPP by Nvidia, IMGLIB and VLIB by TI, the recently announced FastCV by Qualcomm.

As each of these companies develops its own API, however, the market fragments, creating a need for an open standard that will simplify the development of efficient cross-platform computer vision applications.

Now, Khronos' (Beaverton, OR, USA) vision working group aims to do just that, by developing an open, royalty-free cross-platform API standard that will be able to accelerate high-level libraries, such as the popular OpenCV open-source vision library, or be used by applications directly.

The folks at the Khronos Group say that any interested company is welcome to join the group to make contributions, influence the direction of the specification, and gain early access to draft specifications before a first public release within 12 months.

The vision working group will commence work during January 2012. More details on joining Khronos can be found at or e-mailing

No comments:

Post a Comment