Eye On Analytics
- By John Whiteman
- Jun 01, 2009
When acts of violence occur at schools or on college campuses, or threats of terrorism force the emptying of the local shopping mall or a downtown office building, everyone goes on alert. The idea that significant security threats can come as close to home as school or work raises the question: What can be done to protect these areas and safeguard people?
The video analytics industry has traditionally and successfully secured hard targets -- facilities such as airports, critical infrastructure, government buildings and military bases -- with fixed perimeters and checkpoints.
But soft targets, such as malls, schools and workplaces, where people typically circulate more freely, also can be protected by the deployment of video surveillance enhanced with analytic capabilities. Intelligent video has the wherewithal to identify and track suspicious persons or objects in these scenarios as well.
For several decades in the United States, schools and college campuses have been the scene of tragic events. Dating as far back as the 1966 rampage on the University of Texas at Austin campus, in which 16 people were killed, and the shooting deaths of 32 people at Virginia Tech in 2007, educational institutions have proven to be vulnerable to attack. In 2007, there were four deadly college shooting incidents and an additional two since the beginning of 2008.
How can video analytics be used as a security measure to protect schools?
Schools have areas that can be identified for detection and protection. On school property, surveillance areas can include perimeters such as surrounding roads, sidewalks and parking lots. Video also can be captured at entrances, hallways and stairwells. Within school settings, video surveillance with analytics can be present in dorms, playgrounds, classrooms and cafeterias.
The mere presence of cameras in these locations will often be enough to deter some people from committing acts of violence or vandalism on school property. But both students and visitors soon come to realize that cameras aren’t completely effective if no one is watching or analyzing the video. Cameras also generate hours of video, and if the goal is to prevent a event from occurring, not just capture it for forensic purposes that someone must watch it live and react to what is happening.
The benefit of intelligent video is that it has the capability to detect predetermined events, which allows security to be alerted when out-of-the-ordinary or suspicious behaviors occur. Whether a campus has 25 cameras or 250, intelligent video will watch for, detect, identify and alert personnel of potential threats. This can result in fewer guards, reduced costs and better use of those human resources.
Some of these campus-related scenarios that can be defined and detected through the use of video analytics include fence climbing, trespassing or loitering, theft and vandalism. Even illegal parking can be identified through the use of specific analytic algorithms.
For example, trespass detection can be employed to detect someone approaching school property from an abnormal area, such as a wooded area behind the school or a neighboring construction site. Through the use of tripwire or virtual fence programming, video analytics can detect if someone is crossing a pre-determined invisible barrier and ignore movement that runs outside of the designated area, thus reducing the chance of false alarms and improving the probability of detection.
On the grounds of college campuses, the use of PTZ cameras equipped with autonomous tracking provide the ability to track an intruder’s movements through handoff detection from fixed cameras, so security personnel can see where an attacker is at all times and follow him or her continuously. Because video analytics can perform in inclement weather and low-light conditions, there isn’t the concern of system failure at critical moments.
At Israel’s Technion University, potential threatening behavior from students, such as vandalism and theft, are teamed with the real concern for suicide bombers, snipers and terrorist rocket attacks.
While cameras were a big part of the university’s security plan, personnel soon realized that additional video needed to be put to the best possible use, as the human eye just isn’t capable of processing that much information. Intelligent video appliances were added to perimeter and gate area cameras to detect acts such as fence jumping, as well as for remote surveillance capabilities, such as when the school was under rocket attack in 1996. Although the campus had been evacuated, by using the enhanced video system, school officials were able to observe the attack from a safe distance and determine if additional emergency response was required.
Having a system that can be set to detect various behaviors at different times of day also allowed Technion University officials to conduct business as usual during peak times while watching for packages left behind, idling vehicles or removed items at other times of day.
Rethinking Retail Security
Just as with school settings, the need for enhanced security has never been greater within the retail community. The 2007 RAND Report, it identified more than 60 terrorist attacks against shopping centers in 21 countries since 1998. The RAND Corp. is a nonprofit research organization.
Within a mall environment, security personnel find themselves better equipped to respond to events when they have video analytics at the ready. Intelligent video -- which works over IP networks and allows for remote analysis and visual confirmation -- delivers information into the hands of security guards on patrol. Using a wireless network, guards can carry PDAs that allow them to see the information coming from the cameras and control them as well. This form of portable visualization helps security make the best decisions on responding to an event, as well as helping to evacuate or secure people and property during a crisis.
If a gunman is in the mall, for example, guards can view the exits and determine which is the safest means of egress for the thousands of workers and shoppers. The intelligence in the video can alert law enforcement to additional potential hazards -- for example, a possible bomb left by a doorway or a second gunman waiting outside.
If a mall closes at 10 p.m., analytics can be used to detect and alert when someone breaches the loading dock area in the middle of the night or, via the use of a different algorithm, sends an alarm if someone idles a car outside an entrance after closing time for longer than a predefined duration.
There also is the opportunity to link the information from analytics with access control systems so doors are released or locked as needed or an alarm is sounded.
A More Secure Workplace
Workplaces will benefit from the addition of analytics to their video surveillance systems, especially in the current climate of workplace-related violence. The Bureau of Labor Statistics’ Census of Fatal Occupational Injuries reported 11,613 workplace homicide victims between 1992 and 2006. The types of violent incidents taking place on the job can include criminal acts by outsiders, customer- or client-related events, coworker incidents and acts of domestic violence.
According to the BLS Survey of Workplace Violence in 2005, employees can be affected by workplace violence in a number of ways, including increased fear, lower morale and higher absenteeism.
Within the workplace, analytics helps take conventional camera systems beyond the basic capabilities and into the realm of unmanned surveillance monitoring. The price of cameras continues to decline and, because IP networks reduce infrastructure costs, analytics is becoming a more workable and affordable solution for mid-range through large-scale employers. The systems are easy to install and use, which enhances their attractiveness to end users who may not be as tech-savvy as government or critical infrastructure counterparts.
While the security of schools, malls and other workplace soft targets can be significantly improved through the use of video analytics, the system isn’t meant as a substitute for a thorough threat assessment and a well-constructed security and emergency response plan. But it can serve as a key component in a proactive security program and allow students, shoppers and office workers alike to go about their day, providing them with a heightened sense of security.
John Whiteman is vice president and general manager of ioimage North America.