Page 46 - Security Today, March 2018

technologies you can use to detect its presence.
Commercial ground-based radar can detect people and objects at
a distance even in heavy fog, rain and snow. They come in various de- tection ranges, from several feet to many miles. These emerging-based radars are filling a security need in both the consumer and critical infrastructure markets. Unlike their military counterparts, they are very affordable. As IP-based technology, they can send geolocation data to a high-resolution IP video camera to turn its focus to the area in question to verify and auto-track the threat. As the threat moves across the camera’s field of view, tracking automatically can be handed off to the next adjacent camera and so on.
Subterranean sensors and fiber can also be used to detect a threat in a critical area of your facility such as on the fence, in the buffer zone, around potential traffic areas leading up to or even in- side your property. Solutions like these are already being used on fence lines in military applications and commercial industries, and as costs continue to drop they are making their way into new com- mercial markets.
Capturing with Cameras
Tri-spectral cameras — cameras able to capture video using thermal imaging, illuminated IR and visible light all in one form factor — can help you achieve multiple goals: early detection, verification, recogni- tion and identification. In the absence of sufficient external lighting, thermal and infra-red illumination can detect a threat using the non- visible light spectrum.
While they can’t provide a lot of forensic detail, thermal imaging can display the heat signature of a threat even in total darkness. With illuminated IR the camera can capture important details to help you verify such things as a vehicle’s make and model or an individual’s gender, clothing, facial hair and distinguishing tattoos in a totally dark scene. In daylight, the camera can use visible light to record other important identifying features such as colors.
With the addition of advanced analytics, thermal cameras can detect a presence at long range with just a few pixels on target. With slightly more pixels on target they can start to recognize the differ- ence between an object or a person. With more pixels on target they can identify characteristics of the target, such as a person carrying a crowbar or a weapon.
This ability to detect at a distance with good accuracy has signifi- cantly decreased the number of false alarms users experienced with earlier, less reliable analytics software. Years ago, along the Texas border the problem of blowing tumbleweeds was triggering false alarms, and it got so bad that operators eventually turned the soft- ware off. Today’s analytics have improved to the point where they can be programmed to filter out such irrelevant distractions, greatly reducing false positives.
There are also a host of behavioral analytics currently on the mar- ket to detect suspicious loitering or objects left behind, but within the next few years we’ll be seeing more sophisticated behavioral analysis software that will be able to make predictions based on facial and body movement.
For instance, if the video captures a person repeatedly eying a particular asset or exhibiting minor behavior out of the ordinary it will trigger an alert that the behavior is suspicious and needs to be investigated.
DENY: withhold admittance to
DELAY: to stop, detain or hinder for a time
If the threat ignores the deterrents and manages to circumvent
detection, electrifying the fence line may stop the threat in its tracks — but liability is a risk with this system. Should the outer perimeter be breached, there are several IP-based tools you can use to prevent entry into the inner perimeter surrounding critical assets.
IP-based access control systems can be integrated with high-res- olution cameras to verify that the person using a key card or fob is actually the person to whom the card was issued. This could be tied to facial recognition software that would withhold releasing the door lock until a match was verified.
Alternative Links
As an alternative, VoIP communications can be linked to the security solution, requiring an operator to verbally screen the individual be- fore unlocking the door. Or the VoIP system could incorporate voice recognition software to verify the individual’s permission level to en- ter that location before releasing the door lock.
Subterranean sensors can be used in restricted areas to trigger loudspeaker warnings and other alerts if motion is detected in areas where it’s not expected. For instance, if rail deliveries are scheduled to arrive through a gate four times a day, a sensor embedded near that gate can be programmed to send an alert if there’s an attempt to open the gate at an unscheduled time. This could also trigger delaying actions like the automatic closing of a secondary inner gate should the first be breached.
Advanced seismic wave sensors, a relatively new technology on the market, take sensor analytics to a new level. Sophisticated algo- rithms analyze wave length, potency and other variables to not only identify where the sound is coming from but also what is generating it. The algorithm can discern such things as a human footsteps, the slamming of car door or even the chambering of a rifle — important details that security officers can use to verify, locate and delay the threat from penetrating further.
DEFEND: secure against attack
From a defensive standpoint, situational awareness is critical both for you and any outside authorities coming to your rescue. Here, too, many of the IP-security tools mentioned above can provide crucial real-time information to help responders strategize tactics while en route to your location and continually fine-tune them once onsite as the fluid situation unfolds.
High-resolution visual, thermal and illuminated IR cameras can track perpetrators as they move through the facility from brightly lit areas to dark corridors and corners. They can provide important details as to the number of perpetrators, if they possess weapons and what kinds. Embedded sensors can be used to detect movement in restricted sites with the most critical assets and automatically lock down the location, isolating the threat to a limited area or they can trigger an automatic lockdown of all the doors in the entire plant or other action commensurate with the level of threat.
The ability to quickly aggregate and synthesize data from multiple surveillance cameras and other security systems will provide defend- ers with a more in-depth assessment of the situation on which to base a timely response. New self-learning analytics, also known as artifi- cial intelligence or deep learning, will have the ability to parse, learn from and make determinations or predictions with increased accu- racy. These tools will further aid defenders in their quest for more comprehensive awareness and faster alerts to potential threats.
How Secure is Your Perimeter?
Perimeter security is both a science and an art. As the threat land- scape continues to evolve, your perimeter security
needs to evolve as well. So while you might not
need every IP tool mentioned in this article right
now, it’s in your best interest to know what’s out there and what’s coming down the pike.
Joe Morgan is the business development manager, critical infrastructure at Axis Communications Inc.
44
0318 | SECURITY TODAY
PERIMETER SECURITY