In years of economic recession, the level of property crime tends to grow quickly. If one has concerns about safeguarding the well-being of your employees and protecting your property, then installing systems to improve security can be a wise move. Theft and fraud cost American business more than $ 40 billion a year.
Security surveillance use has become widespread as more people want to protect themselves from harm. An ABC News / Washington Post poll in 2007 found that 71 percent of Americans favour increased video surveillance. Already in the United States, there are more than 30 million surveillance cameras shooting in excess of 4 billion hours of video recording every week.
When planning a security camera system, one needs to ask: what is the objective? Do we want a video record so that it can be reviewed in the event of an anomaly, or do we want the ability to react real-time (to an event in progress)? Either objective can be accomplished, but the system will need to be more sophisticated if the objective is to react real-time.
Mounting video cameras are cheap, but funding human resources to observe the output is expensive. Like so many problems, the solution can be computer technology. Computers can not do the job by themselves, but they can analyze the imagery and automatically alert a human operator to any suspicious event.
“It’s impossible for mere mortals with eyesballs and brains to process all the information we’re gathering,” says Stephen Russell, CEO of 3VR.
Progress with artificial intelligence (AI) techniques have largely been driven by our efforts to combat terrorism. This technology is called “Video Analytics”. Facial recognition software is already in use hotels, banks, airports, etc. This software is far from perfect, but it is a “force-multiplier” that is, instead of a 100 humans monitoring 20 video screens each, the computer software screens for suspicious behaviour, and a single human can monitor 2000 cameras. The system can be integrated with the Internet so that the human need not be on-site.
The technology used by Homeland Security is trickling down to systems being employed to monitor businesses. It is a much simpler task for the software to spot suspicious activity in a loading dock or warehouse during a period when it should be idle than it is to determine suspicious activity in a busy environment.
Formerly dominated by burglar alarm companies and camera experts, the video surveillance field is rapidly becoming an industry of information technology integrators, who can offer a more sophisticated approach. Revenue generated from surveillance software is anticipated to increase from $ 245 million last year (2008) to more than $ 900 million in 2013, according to a report by ABI Research.
It is important that we are realistic with our expectations of video analytics. Computers have limited visual intelligence and can perform some video analytic functions reliably, but only by constraining the application. Computers are good at locating possible events, and humans are good at validating those events.