Smarter CCTV system to be used to recognize and prevent crime
By Jeff Salton
September 29, 2009
The negative impact surrounding terrorism, crime and anti-social behavior has resulted in an escalation in the amount of remote surveillance undertaken around the world, but especially in the UK, which, according to the Centre for Secure Information Technologies (CSIT), has deployed more than 4 million CCTV cameras. Putting aside privacy issues for another article, the increase in CCTV usage has had very little success in preventing crime. The main problem seems to lie in the amount of video captured versus the amount that can be viewed and interpreted by trained staff. To overcome these shortcomings, UK researchers are investigating the use of computer technology that recognizes suspicious behavior in live Internet-enabled CCTV feeds from buses and trains, allowing control room staff to intervene and protect drivers and passengers from assaults, thefts and other incidents.
The system automatically analyzes video and audio information – such as shouting, abusive language, physical gestures and other threatening behavior – that is captured by every camera/sensor in a CCTV network and instantly sends the information it judges as most suspicious to four screens in the control room.
The controller can then decide what to do, such as: notify a police car nearest the bus; send the relevant CCTV feed to a small screen in the bus driver’s cab (the driver can then activate a recorded warning saying a police car is on its way); link the bus with the police car so that the police can issue a warning; or the controller can speak to suspicious individuals via the video screen.
Prevention better than cure
It is envisioned the technology will not only alert CCTV control room staff that an incident might be about to occur, but also equip them to intervene and prevent it from happening.
The ISIS (Integrated Sensor Information System) project was initially developed at the Queen’s University Belfast and funded by the Engineering and Physical Sciences Research Council (EPSRC). CSIT will now progress the work in two key areas: increasing the sophistication of the data analysis techniques which the system incorporates; and equipping the system with the artificial intelligence needed to connect small-scale events, draw conclusions about their significance and prioritize information displayed in CCTV control rooms.
“Our system will prioritize feeds but still ensure it’s the controller who makes the decisions as to what action to take,” says research director Dr Paul Miller. “The system will instantly give every live feed a score, based on factors such as time of day, crime statistics for the location, a threat assessment of the people shown, and so on. This score will determine where each feed is placed in the queue for the controller’s attention.
“We aim to develop a system which helps to make crime-free buses, trains, stations and airports a reality,” says Dr Miller. “Ultimately, it could be adapted to protect many other kinds of critical infrastructure, too.”
It is hoped the computer system could be deployed in around five years.