Australian scientists have developed complex new surveillance technology that can spot a suspicious object at a crowded airport and sound a security alarm within seconds.
The breakthrough technology is a world-first, using robotic learning techniques to train computers to rapidly screen vast amounts of footage from security cameras and radar systems.
The hi-tech image analysis program, developed by the Cooperative Research Centre for Sensor Signal and Information Processing at the University of Adelaide, is already being used to watch activity on Sydney's Harbour Bridge and Anzac Bridge.
A computer network simultaneously reads streams of video footage from dozens of cameras positioned on both bridges. If anything suspicious is spotted, it immediately sounds an alarm and illuminates the activity onscreen. ''With most security systems, you've got teams of people looking at a bank of monitors. That's not very efficient, because people get visually exhausted quite quickly and it can be difficult for the human eye to spot small changes - like a suitcase or package left behind in a busy area at an airport,'' the head of the University of Adelaide's school of computer sciences, Professor Michael Brooks, said.
''We've trained the computers to look for any suspicious changes in the scenes they're watching, to distinguish between different types of behaviour and objects.'' The system has already won awards around the globe, and made Australia a leader in smart surveillance technology. The technology is also being tested to sort diagnostic images, sorting through millions of PAP smear images to detect cancerous or pre-cancerous cells.
