Australia's #1 industrial directory for equipment & suppliers

Helping computers to see

07 November, 2012

Driverless cars, robotic mining, smart 'event alarm' CCTV and even at-home stroke rehabilitation - this is the world where computers can see, being made possible by researchers like new University of Adelaide Professor of Computer Science, Ian Reid.

Professor Reid has come to Adelaide from the University of Oxford where he led the Active Vision Laboratory, with research in computer vision and robot vision. His research will add to the activities of the University's well known Australian Centre for Visual Technologies within the School of Computer Science.

"The aim of computer vision is to understand the content of an image or video, to a variety of ends, trying to make computers see," Professor Reid said.

"Applications of computer vision are already changing our world, but the potential for the future is enormous - we've yet to come close to human sight abilities in most areas."

Professor Reid's work received national UK attention in the mid-1990s. He and an Oxford colleague showed, through an early application of visual geometry analysis (prior to HawkeyeTM), that the controversial third goal of the winning 1966 World Cup England soccer team - which helped Geoff Hurst to his hat-trick - should not have been awarded.

Much of his research is in the area of 'Active Vision' where immediate visual processing of video is used as part of a robotic system, with robots reacting to their environment, or with pan-tilt-zoom cameras for tracking and surveillance.

He and his team were the first to create real-time systems for building 3D maps from video and pinpointing the position of the camera within the environment so that live video can be combined with graphics realistically (live augmented reality), or used for robotic navigation. This has become known as visual SLAM (simultaneous localisation and mapping).

"Though successes in SLAM to date have mostly used other sensors such as laser range finders and radar, vision is attractive because cameras are cheap, versatile, and found everywhere, and video conveys a large amount of information about the world that is easily interpreted by humans," Professor Reid said.

Current research includes combining visual SLAM with radar to enable highly accurate mapping and surveillance of moving targets in a marine situation - used in security systems for the Weymouth sailing events at the 2012 Olympics.

Have your say...

We welcome thoughtful comments from readers
Reload characters
Type the characters you see in this box. This helps us prevent automated programs from sending spam.