Cernium Intelligent Video Management

Video surveillance is radically changing, as technology advances and larger camera counts increasingly expose the flaws in traditional systems. By now, everyone knows all too well that traditional video surveillance does little to improve security effectiveness and operations efficiency. In many cases, the video is not monitored and used only for forensic purposes. Even when monitored, human inefficiency and distraction prevents 24x7 analysis of all the video. In fact, studies from Sandia National Laboratory show that human guards can only effectively focus on even limited numbers of video feeds for as little as 15 minutes at a time.

The recent destruction of the governor's mansion in San Antonio, Texas provides a sobering but realistic example of the failings of traditional video. A video surveillance system was in place, was being actively monitored and the arson incident was captured on video – but the guard was watching a different screen and missed the entire incident! The old approach to video surveillance is simply not an effective, stand-alone answer to real-time security.

The new standard for video-based security lies in intelligent video management, which enables a more proactive approach to video surveillance in a scalable, costefficientway. Intelligent video management changes the security operations mission and the role of the security operator by combining the power of leading video management platforms, like Milestone XProtect, with analytics software.

By analyzing each video sequence, analytics-equipped systems automatically annotate the video with essential information about its content, allowing an information-based approach to video distribution, monitoring, search and storage. The result? More effective security operations and better video management.

For users and systems designers considering this new approach, selecting the right analytics solution is critical. Key factors to be considered before implementing an analytics solution include analytics capability, adaptability, ease of installation/use and cost.

Analytics Capability
True analytics solutions provide analysis, detection, classification and reporting capabilities that extend well beyond advanced motion detection. They are also robust enough to handle different types of security scenarios, not limiting the user to a single alarm-like function on a given camera. Why look for just a moving object when you can look for a 1) person; 2) running away from a building entrance; 3) towards a car that's parked near the back entrance; 4) after hours?

In addition, leading systems carry the intelligence through to the user interface so that data from video can effectively be acted upon and analyzed for trends. The ability to detect people is not useful when the software produces only system messages labeled 'motion' or 'event'. Both real-time crime interdiction and forensic search improve dramatically when the analytics software tags video with, for example, 'Loitering Person, Building 2 back entrance, 09/17/2008, 0235.

With a database of rich information from video, query functions deliver real intelligence that can improve security and operations.

Adaptability
The selected solution should be open standards-based and able to easily integrate with leading video management platforms. Ideally, the solution is also COTS (commercial, off-the shelf) hardware compatible, providing maximum system architecture and design flexibility. This approach prevents hardware/vendor lock-in, keeping the overall security architecture agile, scalable and supportable.

Ease of Implementation, Installation and Configuration The analytics solution should be straightforward and simple to install and expand. More importantly, it should be easy to configure, change preferences and add cameras. Users should expect the majority of their interactions with a video analytics application to be point-and-click.

Cost
Just as the San Antonio example illustrates, it's impossible to predict exactly where the next security violation will occur in any installation. The ideal analytics solution should be cost-efficient enough to enable analytics on every camera, with as many combinations of user-defined scenarios as necessary to improve safety and security. Analytics solutions need not any longer require tons of hardware. A single middle-of-the-road enterprise server with a 2U form factor can now process 64 or more video inputs – at an amortized hardware cost of less than $50 a channel.

While intelligent video may be the up-and-coming capability in video surveillance, it is the integration of video management systems and with analytics that is the key to the growth and further evolution of the technology. Video management systems provide the platform for analytics to be truly effective, performing the intense work of capturing, storing, playing and distributing video data. The analytics information layer in turn enables optimization of each of those key video management functions. It is not just "intelligent video", but rather "intelligent video management," – two technologies closely entwined that will set the standard for the future of video surveillance.

Melissa Blake is the Product Manager for Cernium's flagship product, Perceptrak. Ms. Blake's experience includes over 17 years of service in product development and product management functions for leading communications companies such as INTELSAT and MCI Worldcom.

Full Article click here>>>

More News from the