Video analytics at center stage as a video management solution

With more images being captured than ever before, an intelligent video system can make it manageable
 - 
Saturday, January 1, 2005

With cameras every where - and more being installed every day - the collection of video for surveillance purposes has skyrocketed.

But with this explosion in image capture and storage has come the issue of sorting through all this video, which has given rise to the need for analyzing images, so security personnel can go quickly to those images and events that are outside the norm or are of interest based on pre-determined policies.

Fortunately, say those in the business of providing video analytics, better technical and hardware capabilities such as improved algorithms and increased computing power have made intelligent video more affordable, more accurate and more accessible to end users.

“We are on the cusp of an explosion in video analytics,” commented Paul Cataldo, vice president-marketing at IntelliVid. He said a variety of factors have converged “to create this rich environment for video analytics,” including more cameras sold, the transition to the DVR, cheap storage and a host of other technological advances.

Sean Patty, chief executive officer of Cernium, likened the video analytics market to that of the DVR - “it starts slow and then it really runs.”

“I really think its catching on now,” he said, citing a “real general application demand for intelligence to be applied to CCTV networks.”

While companies have been offering intelligent video systems as an adjunct to existing CCTV set-ups for several years now, the post-Sept. 11 environment created opportunities with high-profile, government-based projects.

“This year was a validation that video analytics can work,” said Craig Chambers, president and chief operating officer at Pyramid Vision. He said the military is employing it for key installations, while the Transportation Safety Administration is using intelligent video to enhance security for airports.

As a result of success within the government sector, Chambers said commercial enterprises are looking to use it as well, ranging from pharmaceutical facilities to power plants to car dealerships.

“There is video everywhere,” he said, “and people want to make better use of it and interpret it.”

But to gain greater access, said Alan Lipton, chief technology officer at ObjectVideo, systems need to be simple to use. “People want us to go further to add high-tech science to make their lives easier,” he said.

Added Chris Taylor, director of product marketing for ObjectVideo, “In order to gain the benefit (of video analytics) you can’t have them (the end user) have to deploy a Ph.D. staff. The customer wants a smart box, not having to be smart to use it,” he said.

The key is to get the technology from the research and development stage all the way down to the point where a person in a guard station can easily use it.

Lipton said Object Video recently partnered with Texas Instruments to import its video analytic algorithms into a TI DSP device.

As a result, said Yvonne Cager, world manager for DSP at Texas Instruments, TI is now working with several vendors to employ this product in their cameras. “We got their (Object Video’s) software and our hardware together to employ at the end point for cameras,” she said.

Cager concurred with Lipton’s comment that ease of use is important. While end users need to be able to understand what they are seeing from intelligent video, accessing it is made possible through a Windows-based, point-and-click system with a simple tutorial, she said.

Although intelligent systems require some decision-making from the guard or monitoring station, Brooks McChesney, president and chief executive officer of Vidient, said systems are designed to make them user-friendly. “We generate an alert when something happens,” he explained, “and can generate a message to first responders.”

In the past, said Jumbi Edulbehram, vice president-business development at IntelliVid, people were more focused on compressing video and improving the frame rate to store video. “But it just made things worse,” he said. “You were just storing more video, but taking people out (of the surveillance process), so there was no one to look at the video.”

He said companies employing video analytics “are eager to put in new cameras because there is something they now can do with (the video).”

What has also eased the transition to video analytics is the ability to use them with existing CCTV systems, whether they rely on analog, digital or the newest IP addressable cameras.

“We’re indifferent to the type of camera used,” said Patty of Cernium.

“Anywhere there are CCTV systems,” he said, the intelligence can be wrapped into the video management system.

Vidient’s McChesney said while his company has integrated its product with Sony IP cameras, “we also integrate with existing CCTV systems.”

Pyramid Vision’s Chambers agreed that customers are looking to maximize their existing CCTV systems. “All of our systems (VisionAlert, Hawk and Video Flashlight) are designed to work with conventional CCTV,” he said. “You don’t have to design a system around us.”

However, said Lipton of Object Video, the issue is more than going to digital, but it’s about making a case for intelligence.

With more than six million cameras and DVRs sold globally each year, he said, “if OEMs are right and each will have some form of intelligence, it’s a huge market.”