Let’s face it. CrimeGuard is a video surveillance facial recognition system currently used by 3VR customers. It works with low-fidelity video images from multiple camera sources and conventional surveillance networks.

When RCA introduced the tube camera way back when, it was the introduction of hardware-based security video systems.

Today the floor of the International Security Conference early next month in Las Vegas will still be full chock-a-block with cameras, monitors, transmission hardware, digital video and network video recorders.

But software is now the king of video.

Driving the shift: software built into the cameras, DVRs, NVRs and computer-dependent monitoring and command centers. Whether intelligence will move to the edge or not, it is apparent from what is being introduced at ISC that more software-based intelligence and software will be in cameras, software-based compression and metadata schemes carrying more intelligently-annotated images more quickly and – not surprisingly – video analytics is the buzz phrase.


The charm is not just the decision-making a video system can do but, with software more prominent, security video more effortlessly integrates with other security and building systems and expands into enterprise operation, business and financial systems.

Security Magazine has invited two experts to talk about analytics advances and next generation compression/decompression. Michael Rubinov of Nice Systems sees the intelligent security cycle as moving from detect to verify to resolve to investigate to improvements. Then Mark Oliver at Stretch will fill us in about the future compression developments, also driven by software. 

The first step of this cycle is to DETECT, in other words acquire specific knowledge of an incident as it occurs. Consider the following. A network of 3,000 city surveillance cameras, recording 24/7 might capture a half-million hours of video footage each week, more than 2 million hours a month. But in terms of detecting a specific threat, only two minutes of that video would matter – not the other 1,999,999 hours and 58 minutes when the cameras were watching while nothing happened. The problem is evident, human beings aren’t wired to sit and stare at video screens for hours on end. Even the most diligent operators will succumb to fatigue, boredom, distraction and sensory overload.

Go middleware. The IntelliView IVA is a middleware device that combines digital video recording, video analytics, automated alarm and event notification, flexible networking capabilities, and configuration of multiple cameras into one compact package.


An intelligent video solution solves this problem by using video analytics (pre-set algorithms built into the video surveillance system software) to identify and detect specific threats, conditions or behaviors. Special video analytics software monitors live video feed from cameras to detect potential threats as they occur.

There are many forms of analytics. Some city centers are using gunshot detection software coupled with video surveillance to identify crimes in progress. For public utilities, intelligent video might show intruders approaching a reservoir or power plant, allowing for the prevention of security breaches before they happen. In the case of public safety and transportation networks, intelligent video might be used to identify accidents or traffic congestion so that emergency vehicles can be rapidly dispatched, investigations initiated or traffic rerouted. Another application is license plate recognition, which scans license plates to ferret out suspects with felony or misdemeanor warrants, vehicles that are stolen or involved in Amber Alerts, and even possible terrorist targets.

Intelligent video surveillance is also used for intrusion detection, crowd control and perimeter protection and abandoned bag or package detection and for many public buildings, office complexes, and school districts. Advanced analytics may detect someone loitering outside an exit or a car parked for too long.

The second step in the intelligent video security cycle is to VERIFY.

The nature and seriousness of an incident needs to be assessed instantly. In other words, how real is the threat? Is it an impending disaster, or simply a false alarm? The clock is ticking. If the control center operator waits too long to act, a dangerous situation could escalate out of control. But making the wrong decision based on misinformation could lead to serious consequences as well, such as misallocated resources, a costly evacuation, widespread panic, a too little too late response, or worse the loss of life.

Fortunately, intelligent video surveillance systems have built in capabilities that can help the operator rapidly and accurately evaluate a potential threat. When the system detects a threat and signals an alarm it automatically displays relevant video footage. The operator can instantly replay the video to review what just happened, while simultaneously surveying live video from the same scene. Together, these two pieces of visual information help the operator assess the nature and severity of the threat.


Intelligent video surveillance technology accelerates and improves the resolution of incidents in several ways. First, in addition to alerting the control room operator of a threat, the system can immediately send out an automatic notification up the chain of command or down through the response organization.

Second, the system can incorporate relevant task lists based on pre-defined response scenarios, so the control room operator immediately knows what to do next. When a threat is detected, instructions specific to that type of incident automatically pop up on the operator’s screen, directing what steps to take, who to contact, and providing other essential data.

Third, the system ensures that first responders have access to vital video information. The same on-scene video seen by the control room operator can be streamed to mobile units or incident commanders in the field, for viewing on a laptop, PDA or other handheld device.


Now that the incident is over, the next step in intelligent security cycle methodology is to INVESTIGATE. Today’s technology offers a way. The same content analytics available for real-time threat detection can be applied to the investigative process as well. Now investigators can perform content-based video searches to quickly zero in on video that’s likely to be of interest in an investigation. For example, an investigator might search for video based on suspicious objects, movements, behaviors or events. Some systems can detect faces in video recordings and then build a visual database, just like a photo-album, to help police cull out suspects.

The intelligent video system also records every click of the mouse and every camera viewed, providing a full investigation and review of how security officers responded to an incident.

The final and arguably most important step in the video security cycle is the ability to IMPROVE systems and operations in order to reduce and/or prevent future incidents. Isn’t prevention the first step in the cycle, you may ask? Yes, the various uses of video surveillance and other visual technology by public and private entities are usually to prevent and discourage crime by its very presence. It is also true that invaluable information can be obtained and used to improve various situations for the prevention of future crimes.

An overview taken from a recorded situation of any kind becomes a tool to provide a bird’s eye view of overall security operations. For example, by showing all events in time and space on a map, in other words depicting the threat map (with the help of coloring in red areas with high incident rate, in orange areas with medium incident rate, etc.) one can get an idea of where and under what circumstances situations are most likely to occur. Looking at it another way, it becomes a guide to discovering the weakest link of the security operation and/or system that is in place.

Now with video data available to analyze each and every situation or scenario, whether  related to criminal activities or poor operations, the proper preventative measures can be instituted that will improve both systems and operations and ultimately deter even the attempt of a criminal act, which brings the process full circle.


IT advances are sweeping away coaxial wiring and VCR recorders in favor of Ethernet connectivity and hard disk drive arrays. One fundamental problem remains, however:  Video is BIG.  A single frame of standard definition (720x480) video has over 345,000 pixels per frame. Each pixel requires Red, Green, and Blue values to describe its color, so serializing the data for broadcast over an Ethernet network requires over 200 megabits per second of bandwidth.  Using one of the newer-generation high-resolution sensors can easily exceed the capabilities of Gigabit Ethernet.

Clearly, some form of compression needs to be applied to the video before transport to make it more manageable.  This is the role of a video coder/decoder (CODEC).

The increased compression efficiency can make bidirectionally inter-coded schemes a viable option in digital video recorders at a monitoring station.  Examples of CODECs that use this bidirectionally predicted scheme include MPEG2, MPEG4 Part 2 Advanced Simple Profile, and H.264 Main Profile.

As designers plan for future generations of more intelligent and high definition video surveillance systems, they face a difficult set of challenges.

Advances made in video compression technologies will continue to find a home in surveillance and they must be quickly embraced in an increasingly competitive market environment if the product line is to remain competitive.  These new generations must also retain backward compatibility with existing installations.  A fragmented market rich in standards and innovative approaches makes this a particularly difficult problem.  Designers are increasingly turning to software-defined approaches to retain the flexibility they need.  Software configurable processors are leading the way in supplying the compute power needed to fuel these approaches and give designers the freedom to innovate.  The result of these advances will be a quantum leap in the feature sets and performance available to chief security officers.    

About the Sources
Security Magazine thanks Michael Rubinov of Nice Systems and Mark Oliver at Stretch for their insights into the impact of software, intelligence, analytics and improving compression/decompression in security video systems.

Bridging the gap to intelligence. Fusion firmware from Honeywell includes enhancements such as fully integrated video and data analytics and expanded support for digital IP video sources.

SIDEBAR: A Sunoco Solution

Bob Moraca, security director for Sunoco, is advancing to more intelligence built into his security video system. Sunoco, a Fortune 50 company, has locations spanning 28 states across the United States. Along with several refineries and chemical plants, as well as thousands of miles of pipeline, Sunoco boasts over 5,000 convenience stores and service stations and employs over 10,000 people. Integrator Interface, working with Sunoco, specified and installed a new digital video solution for their central monitoring facility using a Rapid Eye DVR. The technology allows Sunoco to get information more quickly, almost real-time, within minutes, so that they could get information to the proper authorities. Operators can even communicate with their employees and their customers at convenience stores, bringing control of security to a new level. “We have the ability to look into the convenience stores remotely from this central monitoring facility, see what’s going on and over a two-way speaker system, actually interact with either customers or employees,” stated Moraca. DVR makers are adding more software to bridge the gap between analog, IP and intelligent video. The emphasis for security managers now is on moving from legacy gear.

A vehicle is equipped with a camera, a hardened computer and license recognition software. Tampa Parking Division personnel can then search a daily-updated database.

SIDEBAR: License Plate Recognition at the City of Tampa Parking Division

As the third most populous city in Florida, the City of Tampa holds immense responsibility to protect and serve the thousands of motorists who travel and park throughout the city streets daily. Specifically, it is the Parking Division within the Department of Public Works that is primarily responsible for overseeing the approximately 17,000 parking spaces. Parking officers would spot check license plates by comparing information to a printed manual of vehicles which held unpaid fines. Not only was this process tedious and time consuming, the City of Tampa was not satisfied with the results. It was at this point that the Parking Division of the Department of Public Works turned to Genetec and its mobile License Plate Recognition (LPR) solution, AutoVu, to fulfill their various objectives.

The solution required equipping one of the Parking Division’s vehicles with ruggedized cameras and a hardened PC running the application. This provided staff with an automatic scofflaw detector. As the vehicle is driven within parking facilities or on the streets of Tampa, the software detects and reads the license plates of parked vehicles. Once plates are identified, the system checks for a match within an onboard motorist database, which is automatically updated by Tampa Parking Division’s central server. When a match is detected, the AutoVu interface generates an alarm notifying the parking officer that a bootable vehicle has been detected.