Seeing the wider picture

The events of 11 September 2001 and 7 July 2005 have shown us that the threats to national security posed by terrorism are not only real but growing. Consequently, such threats are accelerating the need for new, integrated security technologies. Nowhere is this trend more in evidence than in our airports and other transportation hubs, where the convergence of physical security with new information technologies is providing real-time situational awareness of incidents as they unfold. Integrated security systems are fast becoming a necessity to not only help security professionals potentially preempt attacks, but allow emergency workers to act faster and more effectively during times of crisis.

A networked future

Next-generation, integrated security systems combine network-centric video surveillance systems along with other data sources such as biometrics and RFID. These networked video surveillance systems apply powerful analytics to the video data. Through this intelligent video, specific threatening behaviours can be detected and those video clips, combined with other sensor data, can be sent to the relevant staff in a variety of ways, such as email, handhelds, phones or computers, enabling the authorities to react to possible threatening situations in real time.

Modern digital video surveillance systems employ a network infrastructure that should be familiar to any IT professional, and offer a number of advantages over their analogue predecessors, including lower total cost of ownership, greater flexibility and scalability, better image quality and built-in intelligence. As importantly, they open the door to applying computer-based analytical tools to the video data they gather. The result is intelligent, proactive security systems capable of recognising threatening behaviours and patterns as they unfold and issuing escalating, actionable alerts and alarms to the appropriate people to respond.

The IP revolution

While movies and television provide images of uniformed guards intently watching banks of monitors for suspicious behaviour, the reality is that it would almost be impossible for guards to actually spot a problem due to the camera rotation sequence and the short time a view actually appears on a monitor before cutting to the next camera. Until recently, the principal benefit of video surveillance was to provide evidence of an event after it had occurred. Investigators might use the recorded images to identify suspects or vehicles, recreate a timeline of events or gather other useful information, but historically, there has really only been one basic tool for video surveillance systems – the rewind button.

But the IP revolution is set to change all that. With their potential for embedded intelligence, the IP cameras of tomorrow will be able to run intelligent video algorithms to automate a number of surveillance tasks. These may include detecting unusual or threatening human behaviour, processing external sensor data, reading vehicle number plates, people-counting, sounding alarms, opening or locking doors, turning lights on or off and sending e-mail alerts with embedded images.

IP cameras operate as autonomous network nodes, receiving commands and transmitting images and other data over the same low-cost Cat-5 UTP cable used in modern IT environments. At the same time, videocassette recorders and many DVRs have been replaced by commodity network-attached storage devices and servers.

Situational awareness

While the information from a single, intelligent camera can be useful, combining the data from several networked cameras offers the opportunity to recognise patterns of behaviour and anticipate a threat before it is carried out. A simple example is a security guard watching a perimeter at an airport or high-rise office building. A car is parked along the fence on a Monday morning, but unless the guard investigates, it's unknown whether this is an employee on a break,someone just watching airplanes take off, or a suspicious vehicle. If the same car appears at a different place along the fence on a Wednesday, the same guard may not be on duty, or he or she may not remember it. Even more significant, if the same car appears at the perimeter at other airports in the region, there is little probability that humans would be able to connect the incidents. Software is ideal at identifying such a connection as a pattern emerges.

Pattern recognition detects repeated elements in two ways: either the software looks at a number of data points and begins to create its own pattern (as in the case of a car parked multiple times along a fence), or the pattern is identified by an input such as a wiretap, which directs the software to look for a particular type of pattern (for example, repeated visits to a facility by the same group of people). In both approaches, multiple inputs are key. The more data points that can be observed, the clearer the pattern becomes. In addition, results from standalone sensors, such as smoke detectors, optical trip wires, perimeter radar systems, biometrics and card readers can be stitched into the framework of video to provide 'video plus' awareness. For example, access denials followed by a tailgating violation in a common zone may indicate that an individual was intentionally trying to evade detection and was not just an employee leading a guest into a secure zone. In this way, integration of many inputs into a common data file helps responders quickly ascertain what is occurring, analyse it, and take appropriate action.

Video analytics will play a larger role in analysing the volumes of data generated by IP cameras, as well as intelligence from sensors and other security systems...

Sophisticated behaviour recognition software available today can recognise a human being, distinguish it from an inanimate object and accurately determine the number of people in the camera view, where they are going and where they have been. Video algorithms can identify specific types or sizes of vehicles, packages, or pieces of luggage, and how long they have been stationary or removed from a location. Algorithms can also detect a wide range of human behaviours and actions, such as loitering, access control tailgating, perimeter intrusion or detecting people who violate posted policy by walking the wrong direction in a checkpoint exit lane. Libraries of algorithms are continuously being expanded and new ones created to serve highly specialised surveillance requirements.

When a pre-defined behaviour occurs, it is detected automatically and alarms and alerts are sent to responders. If a responder does not take action in a specific time period, then the alert or alarm can be sent to the next responder. Sophisticated software is capable of co-ordinating a comprehensive, enterprise-wide response, which is exceptionally important when several different groups of responders are needed to take action during a specific event. Intelligent video can also act as a training and improvement tool. It allows security staff to go back and review individual responses to an incident, analyse the incident logs and response times, and implement improvements where necessary.

The state of integrated systems

It is becoming more and more common to see sensor-based systems being integrated with the communication systems of those given the task of responding to incidents, says Bill Adams, chief executive officer of G5 Technologies, a systems integrator in the USA that works closely with many of the country's regional airports.

A case in point, Tallahassee Airport is employing a new system that is designed specifically to intercept and combat threatening intruders – whether individuals on foot or in vehicles – via behaviour recognition technology. This new security system integrates many of the airport's security subsystems, from access control to an airport-wide network of surveillance cameras equipped with pattern recognition software. The system uses intelligent video software to analyse real-time video data and feed it into a geomatic data information system – which integrates technologies such as mapping, Global Positioning System, remote sensing etc – to significantly improve situational awareness. Real-time alerts with video clips can then be sent to designated airport security and local law enforcement staff for rapid co-ordination and appropriate action.

As well as reporting a concerted effort to integrate legacy systems with geospatial information technology, Adams is also seeing organisations embark on upgrading their standalone security systems to make them interoperable with other sensor systems.

Inevitably, video analytics will play a larger role in analysing the huge volumes of data generated by IP cameras, as well as intelligence from sensors and other security systems such as perimeter and access control systems.

Summary

As new technologies transform video surveillance into IT-based systems, traditional security companies will need to reinvent themselves to become IT-centric for continued success. IT network security, physical security, and intelligent video surveillance are joining forces, and security analytics is the bond that will unite them effectively.

IP-based systems are gaining popularity, and integrated security systems increasingly resemble the networked environments IT professionals are accustomed to designing, deploying and enhancing. This convergence will accelerate as IT and security professionals combine efforts to go beyond integrating disparate systems toward providing a new level of actionable intelligence through the emerging technologies of video and security analytics.