Smile for the Camera
Security 2006 - The Future of Video Security - Translating technology into serving our needs – cutting through jargon, explaining benefits, making it all work for you.

By Erica D. Harrison, CPP

Two decades ago, this older guy who was the building security director, said not to worry that video cameras were facing into the sun and you couldn’t get a good read on the faces coming through doors. He said that upstairs, where the target was, they had feeds from dozens of cameras. Plus, the security personnel all carried submachine guns. They were trained to shoot first and ask questions later.

Last week, (I’m now the older person), I mentioned to the young guys who carry the submachine guns at another equally attractive target, that their cameras are facing into the sun at loading docks. I suggest they add a‘collar’ to the lenses to cut down on sun glare/flare that interferes with picture clarity. They were receptive to the idea. Yes, they monitor the feeds from dozens of cameras and they are trained to shoot first and ask questions later.

Two decades ago, the 10 feet of critical video tape that showed the actual break-in could be destroyed as it wrapped around the capstan in the VCR. Today, the camera that picks up the break-in may feed directly to a hard-drive and be laser-burned to memory on a digital video recorder (DVR). Or, the information may be sent to a server. Now, all you have to do is ensure that the drive doesn’t get destroyed and that it is backed up elsewhere, avoiding any potential loss of those critical frames.

Video surveillance is a living, breathing part of modern security. It’s gone through changes, refinements and the technology has advanced exponentially. As with most technology, the descriptions of what you can do with video feeds have developed their own language: “behavior recognition software… facial nodal points… frame grabber… intelligent video.”

Our objective is to find out what the terms mean in simple English, which ones are significant for our use in security and to get some help determining just how many of how much, we need.

Our seminar presentation at SECURITY 2006, The Future of Video Technology, could also be titled: Video Security – Translating Technology into Serving Our Needs – Cutting Through Jargon, Explaining Benefits, Making it All Work. The objective: Help attendees answer the question: How does what the salesman and spec sheets say, translate into what I actually want to accomplish?

If you’ve watched the television series about casinos in Las Vegas, you’ve seen intelligent video technology in action, serving the security department’s needs. Facial recognition software, which identified the 80 nodal points constituting a face and compared for the 17 points that give you a “match,” let the security guru check employees’ ID pictures against live faces of people gaming in the casino. At least on one show, the guru picks-up a card-counting employee and his accomplice at the one-armed-bandit. Unfortunately, no one writes our scripts so neatly. At Security 2006, we will have the benefit of expert presenters and expert vendors who put information into easy-to-grasp formats, if not immediately catching the bad guys for us.

Consider using information from SECURITY 2006 as you develop guidelines and determine your organization’s actual requirements. You will have the opportunity to hear from security professionals who use advanced systems as part of their current security plans. It’s a shortcut method for getting an overview of successes and shortfalls. Then you can better develop a game plan for upgrading current video systems and preparing for your next-generation installations.

The basic questions we all ask include: What do you really need? What’s good about the state-of-the-art? How much of it can you really put into action? If you ask Bob Gauvreau, head of corporate security for the City of Ottawa, it is first things first. “And that’s proper planning,” he says. “Plan for what you face and what you need. The second thing is to buy what you need and not what your finances require.” The security executive does not shop by price. (Security Magazine, March 8, 2004.)

Many of us wish we were in his shoes. Indeed, some cost-savings that come with today’s video technology may help us get closer to that goal.

Video Goes To Work

Kevin Anderson* (not his real name), security manager at the corporate headquarters for a major software organization was tasked with developing procedures and choosing the equipment to create the digital heart of the company’s North American area physical security system. In a world where you can get similar products from many vendors and all promise to do anything necessary, it is tough to determine if equipment really can do the tasks at hand. Kevin relied on personnel from Software House to help him integrate the cameras and security systems.

“You get on the right path by choosing systems you are most comfortable with. For me, I have to know them [the systems] inside and out. I really don’t recommend that you rely on the sales organization [for this.] I can describe what I’ve got, see how it is operating and understand the system and the layout. This is one way I can make sure my specs are met by the installations in and around our spaces.”

With the current system, his console monitoring staff was able to detect someone trying to saw through sheetrock in one of their buildings. The control room officer immediately dispatched a security team to control the scene. However, the incident did lead to revamping systems. Now there is additional failsafe security in walls, precluding intrusion without detection. The need to look through the walls to prevent disruptions and theft was added to the security mission. “We cannot afford to have someone steal a server.” It is not that a server is so valuable in itself. More, a missing server could mean disruptions in service to a client, or lost data. These issues represent the risk.

If someone tries to pry through a door at any location within Kevin’s current system, the activity immediately gets flashed on the ‘big’ TV for the console operator’s attention. The
C-Cure software brings details of the location, the door and type of activation into a highlighted entry on another large monitor, so the console operator has the full information to make the next decision in the security process.

Kevin called on his earlier experience, working in a prison, to help identify avoidable video security system pitfalls. “You had a lot of little screens [in the prison console room, with lots of little pictures] and you couldn’t see every little thing. At CA, I need the problem areas in focus or if something is going on at a location, there has to be detail. An open door [that should be secured], glass breaking –I want my console operator to concentrate on that and make it easy for him to see [the detail]. After all, “he [the console operator] can only focus on so many things at once.”

Kevin’s system design takes into account the ergonomics of being a console operator for an entire shift. The console operator is seated at a comfortable viewing angle to the three major screens presenting information from all the North American feeds. The displays are uniformly bright and in a similar color balance so it is easier to spot changes on a screen, rather than always having your eyes go to the brightest or most outstanding feed. Everything from the viewing angle for the console operator and the angle of the three major screens in the control room represents what you can accomplish by carefully integrating security issues with ergonomic design.

The company’s offices in different locations throughout North America face different threat levels with different inherent risks. Of course, not everything that might affect these operations is visible through surveillance cameras. And breaking stories are too old to be useful by the time they make broadcast on the TV feed. So Kevin added services from NC4 to assist in supplying ‘of the minute’ threat advisories for sites with higher risk profiles. Now, he can provide his executives working in lower Manhattan with detailed immediate information about any disturbance in the area that might affect them.

For some security professionals, video surveillance is the tool of choice for monitoring and managing security staff. It is a powerful tool when taking charge of a large staff performing diverse functions in a 44-story veritable New York City landmark. Karla Beth Kudatzky, Director of Security, Fire and Life Safety for Edward J Minskoff Equities, Inc at 590 Madison Avenue, can view her security team provided by Guardian Security in action with the click of a computer mouse. Karla notes that digital video surveillance provides an overview of almost everything going on all over the facility as well as the adjacent public atrium. She utilizes 50 strategically placed cameras to ‘see’ what is going on both inside and outside the building. She is able to view the security team's performance to ensure that the security criteria and objectives are met for this notable black marble building that is IBM’s headquarters in Manhattan.

Karla’s security team provides her with continuous audio/radio reports of their observations within the building and the areas surrounding it. Over time, they’ve gotten to know the building’s personnel, tenants, and they recognize many of the people who work in the vicinity or are ‘regulars’ to the locale. They can alert Karla to unusual situations immediately: a motorcycle parked on the sidewalk across the street with a package strapped to the rear wheel and no rider, or an unusual number of black SUVs coming down East 57th Street, or perhaps an unexplained gathering of people in the public atrium. These are not necessarily things that intelligent video software would alert.

When necessary, Karla also utilizes a video console operator who she instructs to “zoom” specific cameras to give her more detailed views. The zoom feature is not only convenient but enhances effectiveness as she monitors for potential internal or external threats and ensures that the building has a detailed video record of individuals, vehicles and specific incidents.

The digital video surveillance system at 590 is continuously backedup on the building’s network equipment. However, Karla finds it convenient to burn CD’s of video information that can be utilized for building-specific needs or at times, for special requests by the NYPD. As Karla notes, the video record has also been an invaluable training tool for the security team.

Karla’s day usually begins at about 5:30AM when she arrives at 590 Madison. By 6:00AM, as she has breakfast in her office and as she addresses a myriad of early morning responsibilities, she also reviews video from the network and has an opportunity to survey activities from the night before. She can also monitor what is going on in real time. Karla can replay video from numerous camera feeds at once to track someone; to reconstruct an incident; or to see a video account of a security officer's oral and written report that would never have been possible before.

“A supervisor on the ground could never get the full picture of the information the way that 50 camera feeds deliver it to me in an instant. I can see our security team on 56th Street checking vehicle trunks and undercarriages before they allow drivers into elevators for the parking levels. I can check on a line forming at a lobby concierge desk to see if another officer is required there. I can check a mechanical area of the building and dispatch an officer if there is any unauthorized activity.

For me, video surveillance means I can oversee the security team with better perspective and faster than it could ever be accomplished on foot [walking through the building]. In real time, I can track developing situations and contact my security team to give them instructions that mitigate difficulties and prevent potential problems.”

Intelligent Video –
Using the Technology
The general phrase we hear now is: Intelligent Video. It’s more than clear pictures that hypnotize console operators. The software running these systems has a series of ‘rules’ that guide it and security managers add more ‘rules’ to meet their specific requirements. In the end, whatever gets highlighted for console operators and brought to front focus, is the significant video, not just lots of pictures.

One of the rules might be to highlight any activity in the employee parking lot after 2:00AM and before 6:00AM. You arrive at work and find a parking spot in the employee lot
at 3:00AM. Without the console operator’s intervention you and your vehicle, the moving items, are highlighted with a colored rectangle. The console operator is alerted that there’s something worth watching. But, the operator can also zoom in on your license plate and compare it to his databank. Then, and as you get into view of the next camera, the facial recognition software identifies that it really is you, just getting a very early start on the day.

Intelligent videosurveillance systems work with camera patterns that provide overlap or underlap coverage. If you can afford it, choose an overlapping system. You get more of a
“moving picture” that is easier for people to process. It is similar to the video feeds we’ve seen on the news, covering the interior of tunnels as systems track cars traveling through
the passageway. You avoid abrupt cuts and dead spots.

In some systems, the software is programmed to ‘watch’ items that should be stationary. The “rules” may allow that the items can be moved from location “A” but if not replaced to that location at the end of say, 40 seconds, the command center gets a trouble signal and alert. Additionally, there will be video footage of the items, their removal and the location. Example: the keys to open the gate have not been replaced on the hook. They’ve been missing for three times what it should take to open the lock….

Systems also notify the command center when something stays in a place where it’s not supposed to be. Let’s say a control room operator is monitoring cameras covering 40 active loading bays. The exception is going to be: why is the pallet of televisions out on the dock #12, with no one around for more than 30 seconds?

Shunsuke Matsuoka of Sony Corporation said that part of intelligent video is systems that identify what is ‘not’ a problem. For instance, comparing the last 15 frames of a shadow moving back and forth that fits the programmed pattern for wind in a tree, is an environmental incident. This is not something that requires command center alerts.

The above is a video monitor whose images are processed through intelligent software. The display, without any words, gives viewers a presentation of an entire airport terminal,
an overview of a particular area, and then a close-up of the specific section of concern. It integrates how people ‘see’ and how they ‘think’ and relate space and specifics.

If you watch NASCAR broadcasts on TV you see intelligent video in action every week. A small bird’s eye view ‘graphic’ of the track comes up in a bottom corner of the screen. A beam of light is imposed on the graphic, tracing how the cars will traverse the track during the race. Then, the main picture on your TV screen gives you a real-time video of the cars moving around the backstretch. A triangular beam will show say, where the Budweiser #8
car is in the pack. The close-up picture in the lower right is a live interior shot of Junior in the #8, as he planes the oval.

In a demonstration of intelligent video systems for surveillance purposes, a man in a yellow t-shirt walks around a building’s perimeter on camera. The system alerts the operator that there’s something worth watching. The system highlights the man with a colored rectangle.

He stays highlighted; his movements traced as he comes into range for each roof-mounted camera in the system. The operator remains hands-free, ready to contact anyone required depending on the subject’s actions or any problematic situation on-screen.

Those who have old black-and-white camera systems should not be discouraged. Intelligent video can serve them as well. Age Eide, Christer Jahren, Stig Jorgensen and group at Ostfold College in Halden, Sweden, have developed facial recognition software that works on grey scale video output. The systems use two networks. One identifies and recognizes where the eyes are in the video frame and the second one looks for the differences between those eyes and the areas around them, and the eyes of all the individuals in the database… For more information,‘Google’: “Eye Identification For Face Recognition with Neural Networks.”

Advanced video systems bring their own set of questions into focus and these may be more the meat of what we will have to discuss in the next few years as intelligent video becomes the norm:

• The remote viewer – is that just going to be me as the security director? Or is it just the officers I assign to the video room, or all of us? If it’s just the control room officers, what is the threshold [of activity] that requires them to inform me that there is something to look at? Is there some historical data for reference – or is this just learn as you go?
  
  
• What if non-security executives want video feeds over our networks, such as the COO as well as the CEO, do I ever say no? And if I say ‘yes’ how do I best inform them so they don’t determine too low a threshold for response?
  
  
• The moment there’s an incident – does everything we do with the camera and image controllers go “live” to everyone’s screens? Just to my screen and the control room?
  
  
• How long does it take everyone on the security team to respond/call-in/ activate a response plan to an incident on-screen?
  
  
• How do we keep extraneous responses from non-security executives who are monitoring an incident, from impeding our response as the security department?
  
  
• Do we test ourselves frequently so we don’t over-respond and not have the support we need for the next incident that is going to be right on the heels of the event on-screen?
  
  
• What do we do when the video ‘reading’ software feeding the control room, tells us there’s a bad guy at the front desk? What’s the procedure for alerting the lobby officer?
  
  
• Do we have instructions in the security manual for the above – after all he’s presenting proper identification? What type of training exercises work best? We want to video those exercises for critique as well.
  
  
• Does having a picture of the problem affect our policies? Does it affect how we are held accountable because others can review our behavior as well as the incident itself later on?
  
  
• Does it make us lazy?
  
  
• Does one suspicious incident mean everything goes to ALERT? What level of response will be deemed reasonable so not all resources report to the emergency [in case it is just the beginning of a series of events…] Who sets this policy? Have they set it?
  
  
• What’s our procedure when we are wrong? When what we saw, who we saw, was interpreted incorrectly? (motion going wrong direction is because the person has to get back to elevator, box left in lobby belongs to President of company, etc)
  
  
• How do we modify or not modify to accommodate the changes?
  
  
• Can you really secure who is looking at your cameras over the internet? Where is it most ‘hack-able’? Could you decipher a fake feed?
  
  
• Could you decipher if someone picks up your feed and distributes it elsewhere?
  
  
• Who checks the refresh rate to ensure that pictures are not flashing at same rate as fluorescent bulbs since this unnecessarily fatigues the control room officer monitoring the screens.

As we continue to get our systems into overdrive, the video issues list will change. You can add your own bullets to the above.

The other morning, a news announcer said NYPD is installing hundreds of surveillance cameras through the City and some in Brooklyn. The response was part: “beware of Big Brother” and part: “finally, someone will really see what’s going on.” The future of video technology may give us an opportunity to go beyond what Werner Wolf used to say. Just before he described critical action on the playing field, he’d elaborate: “Let’s go to the video tape.” And, in a quality snippet of footage, we’d watch a basket through the hoop against all odds, or a slide tagged out just before the runner’s foot touched home plate. In our real lives, the future of video technology will give us a hard-recorded tool that will simplify some investigations and make vastly more complex the question of “who’s watching the store.”

>> Return to This Issue's Table of Contents