Development Of 'Human-Like' Computer Vision
Intel Software Accelerates
Of 'Human-Like' Computer Vision
AUCKLAND, December 12, 2001 - Intel Corporation researchers have released at no cost software that allows developers to build computers that can view thee world the way people do - in stereo.
The software, which enables computers to recognise 'depth' as well as 'flat' images, enhances a wide variety of vision applications including gesture recognition, object tracking and face recognition. The software is available through the Open Source Computer Vision Library (OpenCV 2.1), a toolbox of over 500 imaging functions that helps researchers develop computer vision applications.
Existing 2D computer vision applications recognise scenes as single flat color fields. The new stereoscopic code will enable computers to use depth information to breakout individual objects and surfaces in a scene. In addition, the new 2.1 release includes an optional interface so all OpenCV functions can be imported into Matlab*, one of the most widely used software development tools for computer vision research.
"Intel wants to make computers aware of the visual world. Until we introduced OpenCV last year, the lack of common tools kept this from happening," said Justin Rattner, Intel Fellow and director of Microprocessor Research, Intel Labs. "With the addition of stereo capability and a Matlab interface, OpenCV 2.1 is certain to drive an explosion of vision-based applications, such as toys that respond to a child's movements and monitors that warn safety and security people of life-threatening behavior."
Accelerating Research Into New Uses
Faster microprocessors, falling camera prices and ten times more video capture bandwidth from technologies like USB 2 are all enabling real-time computer vision algorithms to run on standard PCs. The computer vision library is designed to increase innovation in this field by providing source code for a wide range of computer vision and imaging functions. In the first year since its release, OpenCV has seen over 75,000 downloads of code and has attracted more than 2,000 registered members to its user group.
"OpenCV implements a huge amount of standard and advanced image processing techniques. Until now most companies spent a lot of time implementing those well-known techniques. With OpenCV we now have access to a well-implemented version that is fast and reliable. Because it is open source, it gives us confidence in the code and we know we can make local modifications if necessary," said Hansjörg Gärtner, project manager, Inspeck*.
Developers are using OpenCV code in applications ranging from toys to industrial manufacturing. This open source software release includes C source code for all of the library's functionality and a royalty-free redistribution license. The OpenCV main website is located at: http://www.intel.com/research/mrl/research/opencv/. Anyone interested in joining the user group needs to register with Yahoo groups at http://groups.yahoo.com/ and then can subscribe by sending email to OpenCVemail@example.com.
Intel has developed a uniquely decentralised research model with over 80 labs situated around the world. The bulk of the OpenCV software team resides Intel's Software Development Centre in Nihzniy Novgorod, Russia. Established in 1999, the centre currently employees over 100 computer research scientists and engineers working in areas such as computer graphics, vision, media, Bayesian networks, compilers and tools research.
Intel, the world’s largest chip maker, is also a leading manufacturer of computer, networking and communications products. Additional information about Intel is available at www.intel.co.nz
* Other names and brands may be claimed as the property of others.