Communications of the ACM
A Technological Eye on Criminal Behavior
Researchers are developing surveillance camera systems capable of autonomously detecting illegal activities, such as when a cashier is giving friends free items or if someone is about to commit a violent act. For example, the surveillance cameras at Big Y, a grocery chain in Massachusetts, can examine checkout lines to look for subtle tricks cashiers use to give free items to friends, including obscuring the bar code, slipping an item behind the scanner, or passing two items at once but charging for only one. Mathematical algorithms embedded in the store's new security system can autonomously identify these acts. When the system believes it has identified a problem it alerts management on a computer screen and provides the footage.
Such systems also could be used to identify abandoned packages on a train platform or alert an airline employee to possible terrorist activities. However, unlike the limited setting of a grocery, the challenge of monitoring behavior becomes far more difficult when trying to spot a hijacker on a plane full of passengers. A European consortium, including a team from the University of Reading, recently tested camera systems capable of identifying threats inside passenger planes. Some of the cameras focused on faces and upper torsos, looking for signs that someone may be planning something, such as heavy sweating, for example.
Reading's James Ferryman says a central computer is used to compile data from the cameras and other sources, including audio sensors and the plane itself. However, Ferryman says there are many variables involved in detecting a possible terrorist action. "A threat in one particular situation may not be in another," he says. "You don't want a system where the cabin crew is constantly being given false alerts."
From The Associated Press
View Full Article
Abstracts Copyright © 2009 Information Inc., Bethesda, Maryland, USA 
No entries found