Video analytics: IP and analytics without worry

03 Mar 09

When it comes to security, the enterprise network has become a critical component. Now, more than ever, it enables corporations to tie their security solutions together with their network infrastructure to take advantage of most of the benefits a network-based solution has to offer.

That includes the ability to integrate security with day-to-day functions, such as risk management, human resources, asset management and IT operations.

Specific to physical security and surveillance systems, a network-based solution affords access to real-time video from anywhere, the ability to archive images and data in a central location, and keep ongoing tabs on devices to identify potential problems, such as power failure to a camera.

But despite these advancements, networking professionals still worry about the impact of an IP surveillance system, paired with other capabilities, on the network.

These concerns only increase as security professionals look to connect more devices and solutions on the network, including video analytics appliances.

Revolution in monitoring?

Video analytics, or intelligent video management, has brought to the market significant improvements in automating security surveillance and has been touted as a revolutionary monitoring tool.

Instead of reacting to an incident only after minutes, and in some cases hours, after someone breaches a secure area, video analytics enables security guards to receive an immediate pre-warning alert to an event thanks to a technology that can visually detect an intruder and then track that individual as they enter a facility.

The question asked by many network administrators today is: When you have a large-scale deployment of hundreds of cameras, paired with video analytics, how much bandwidth will these cameras require?

Will these IP-based devices choke the system? How secure is a network-based video analytic security system?

IP-based surveillance systems and video analytics can successfully reside on an existing enterprise network with limited liabilities.

Not only do these two solutions provide a surveillance solution to reduce shrinkage and theft, monitor perimeters and deliver proactive alerts to incidents, some systems on the market today require minimal bandwidth on the front end by operating on the edge of the network.

Networks stretched

There's little doubt that for many companies the enterprise network is stretched far beyond the capabilities only imagined possible just a few years ago. Today, it's not uncommon to find that nearly every employee has access to company hosted email, the Internet and important data and files saved on a central company-based network.

There's network access across multiple buildings and, in many instances, across multiple states or countries. With these capabilities so inherent in today's business environment, how would a business function without real-time connectivity or access to its data?

With the increasing demand for connectivity comes risk. How much bandwidth does the network have, how secure is the network and what kind of functions will slow it down? Many IT professionals worry about stretching the network too thin, and rightfully so.

In recent years, video surveillance systems have come out of the box. They've been removed from the 'siloed' environments of yesterday, with their own cabling, network servers and separate dedicated management to IP-based systems that reside on a single unified network, with shared infrastructure and connectivity to the entire IT backbone.

For some IT professionals, adding surveillance systems and other security capabilities onto the network can be a scary proposition. Video often contains large streams of data, especially when security professionals demand quality video and lots of it.

Today, those in the security industry expect high resolution full motion video and audio – the days of fractional frames per second are over. Applications need to clearly show a person's face or other in-demand details in video footage, such as the dollar amount on a cash register.

Video overload

With video surveillance a 24/7 business, how do you ensure that your network is not going to be overloaded by video? A significant amount of video recorded today does not contain any relevant or actionable data – a surveillance camera can literally record video for hours before a person of interest walks into the field of view, or a suspicious car drives into a monitored parking lot late at night. To limit the strain on the network, there are a few methods IT professionals can deploy, such as recording video at the edge, using compression technology such as MPEG-2, MPEG-4 and tweaking resolution and bitrate of the video stream itself.

When security professional seek true incremental value from surveillance, they turn to video analytics.

Many professionals view video analytics as an IT intensive solution in more ways than one. Traditionally, video analytics has required an onsite IT expert for implementation because a number of the systems on the market today are software-based and require a dedicated PC to operate and hours of set up time to make it operational for detection.

The problem is that network-based surveillance and centralised video analytics solutions can choke the system – if not set up properly.

They require that the information be constantly transmitted to a central location or server for processing and storage. Then, to view that data or video, it takes up additional bandwidth to retrieve the information. These solutions can require a tremendous amount of bandwidth, both to view the video and store hundreds of hours of it for back processing.

Analytics at the edge

Today, the market see the migration towards IP-based hardware edge devices with built in video analytics, such as IP cameras and encoders. These devices are now being introduced by several security system providers on the market. There are several dozens of different video analytics companies in the market today and less than a handful offer true edge device capabilities.

Deploying an edge device is one method to reduce the strain on the network in terms of system requirements and bandwidth. When smart cameras and encoders process images at the edge, they record or transmit only important events – for example, only when someone enters a predefined area that is under surveillance, such as a perimeter along a fence.

Other video methods transmit or record all video processed, including stagnant video when no one has entered the field of view.

This distributed approach for video analytics makes surveillance easier for network administrators to adopt. It limits the risks associated with overloading the system by enabling administrators to predetermine what video is important to view and keep and which video is not pertinent.

It also provides an important choice in the market. For some companies, a full software-based approach to video analytics might be the right way to go. Still, most others will find that edge-based video analytics devices provide the full-fledge capabilities required without potentially choking the system or calling for a complete system redesign. Video analytics has been proven to improve the bottom line, helping to reduce loss, operating expenses and labour, while enhancing productivity.

Get the latest stories first with info4security's newsletters: Click to signup

Post and bookmark this story at the following sites:

What is this?

del.icio.us	Digg	Reddit	Google	Newsvine