Technology Sectors
On the perimeter: Keeping watch with video analytics
 |
|
Guy Lorman |
For government and critical infrastructure settings, perimeter protection is especially crucial and often the first line of defense for security-conscious environments. Where fences once did the job, now video analytics – sometimes referred to as a Perimeter Intrusion Detection Solution (PIDS) -- is stepping in, either working alongside physical barriers or doing the job solo, as a virtual fence.
The PIDS evolution
There was a time when fences or other physical barriers were the only option to protect a perimeter from intrusion. Often, such barriers had to be manned by guards, which could be prohibitively expensive. Then CCTV arrived on the scene, giving security operators the ability to monitor an area or perimeter from a remote location. But that came with its own set of challenges, primarily information overload.
It has been shown that as the number of cameras in a video surveillance network increases, a security operator’s ability to accurately scan those feeds actually decreases. Research further shows that after approximately 20 minutes of viewing video, the average security operator takes in only about five percent of the information being screened.
Video analytics addresses this challenge of information overload. The right video analytics-enabled PIDS can alert security operators in real-time to suspicious activity, allowing them to verify the alert and then allocate resources and respond immediately.
Video analytics -- the backbone of an effective PIDS
Video analytics works by analyzing captured video for certain data, behavior and objects, using a set of defined algorithms. Once the algorithms identify a potential security breach, personnel are alerted to the issue for immediate verification and action.
Today’s PIDS have the ability not only to identify potential threats, but also to continuously track evolving threats. Once a fixed camera detects a suspicious object or person, pan tilt zoom (PTZ) cameras are activated and begin tracking all movements. The cameras, both fixed and PTZ, have hand-off capabilities, meaning that as the object or person goes out of the range of one camera, another one seamlessly picks up the tracking. With their wide field of view, a single PTZ camera can be linked to eight fixed cameras, or other sensors.
Furthermore, with the latest generation of PIDS, intrusion detection accuracy is greatly improved. Research and field testing has shown that with this latest PIDS technology, referred to as 4th generation, or 4D, false alarms can be reduced to fewer than one false alarm per camera, per day, and can achieve 90-95 percent detection rates.
Other advancements, such as day, night and weather condition filters, along with state-of-the-art human classifier capabilities, are also contributing to greater accuracy and efficiency. Both the reduction in false alarms and the increased detection rates are a substantial improvement from the previous generation of PIDS.
Automated, video-analytics enabled PTZ cameras -- a force multiplier
Depending on the environment and location, there are several options for deploying an effective PIDS, the first being a virtual fence. When a physical barrier is not an option due to esthetics or the need for public access, a video analytics-enabled surveillance system can be used as a virtual fence. Using a combination of fixed and PTZ cameras, enabled with video analytics and value-added video capabilities to filter weather conditions or detect camera-tampering, a government organization can create a virtual fence as an effective, pro-active first-line of defense.
In a situation where a physical fence already exists, it can also be combined with a PIDS for even greater fortification. The fence serves as a physical barrier, while the PIDS provides coverage, early detection and verification.
A third option would be to deploy a smart fence. A smart fence consists of a physical fence, combined with a PIDS, and a fiber optic sensor to detect vibration, flexion, compression and cutting of the physical fence. For highly sensitive facilities, such as a nuclear plant or a tier-1 chemical facility, the combined technologies reinforce one another to improve overall security.
Making PIDS effective
The success of any PIDS depends on its ability to integrate within an existing infrastructure, and the available supporting technology and features that boost its efficacy. Open architecture supporting any analog or IP camera gives organizations flexibility and choice, and independence from any one vendor or technology.
Equally important is the experience and knowledge of the team implementing and designing the solution. In fact, this seems to be one of the biggest influencers on a successful deployment, performance and customer satisfaction. Every environment is unique and it takes real know-how to design a solution to meet expectations.
With so many options available, careful examination is a must. Look for the following when considering a PIDS:
- 4G capabilities;
- Field-proven video analytics;
- Open architecture support for complete flexibility;
- An experienced/knowledgeable design/deployment team.
Guy Lorman is product manager for NICE’s video analytics products. He can be reached at:
