Home
  Subscribe to eNewsletter
  Online
  Breaking News
  Daily News
  Laura's Blog
  Bill’s Blog
  Classified
  Jobs
  Digital Edition
  Webinars
  Showrooms
  SecurityMag.com
  SmartHome Mag.com
  Print
  Subscribe
  SDM’s Current Issue
  Technology @ Work
  5 Minute Tech Quiz
  Technology Solutions & Skills
  Networkings
  Product Reviews
  Plain Talk
  Security & The Law
  SDM 100
  Top Systems Integrators
  Field Guides & Directories
  Annual Buyers Guide
  Monitoring Field Guide
  Annual Guide to Monitoring
  Annual Guide to Distributors
  Access Control
  Events
  SDM 100 Gala
  Securing New Ground
  Security 500
  SecurityXchange for Integrators
  SecurityXchange Enterprise
  Resources
  Online Reader Service
  Archives
  Career Center
  Event Calendar
  White Papers
  List Rental
  Awards
  Dealer of the Year
  Systems Integrator of the Year
  NBFFA First Line of Defense
  Integrator of the Year Entry Form
  Dealer of the Year Entry Form
  Info
  Contact Us
  Marketing Services
  Media Planner
Search in: EditorialProductsCompanies
TECHNOLOGY EXPLAINED
Distributing Video Analytics Between the Server and the Edge

by Russ Gager
February 1, 2008

ARTICLE TOOLS
EmailEmailPrintPrintReprintsReprintsshareShareshare Use



This is one of hundreds of video surveillance cameras with analytic capabilities that are deployed to protect Israel’s railways.
One trend in video analytics is to distribute the software towards the “edge” of the system, either in the camera or the video encoder, which converts analog video to digital. One example of this type of technology is the analytics system from Agent VI, Fort Myers, Fla., which is distributed between the camera or encoder and the system’s computer server. This differs from other analytics solutions in which all of the processing takes place in one location.

“We do not do the whole algorithm in one location, either in the field or central server, but rather we distribute the algorithm to two separate components,” explains Gadi Talmon, vice president of business development.

“We do some initial processing inside an IP camera or video encoder, whatever you have in the field,” he continues. “Then you send the result of this processing over the network to a central server doing the second part of the algorithm. We do not need to transmit the video from all the cameras in the field to a central server.”

This improves performance and efficiency, conserves bandwidth and makes the system useful for enterprise installations with thousands of cameras, Talmon maintains. It also enables the initial analysis to be performed in or near the camera before the image is compressed.

“In most cases, we can use existing hardware devices in the field and just embed a piece of software inside those IP cameras or other IP devices, and upgrade existing hardware for video analytic capability,” Talmon insists.

The company embeds its analytic software on the camera’s DSP chip. “The reason that we can do this is we do not run the entire algorithm in the field; therefore, we don’t have too much of a processing requirement,” Talmon observes.

“Normally, an IP camera or a video encoder would have maybe 20 to 30 percent of its processor available for other applications,” he calculates. “It’s not enough to run an entire video analytic algorithm, but it is enough to run our software, which is only doing an initial part of the algorithm. Therefore, it is suitable for the limited processing resources of video devices in the field.”

All of the company’s analytic software can fit on one chip, Talmon asserts, but it is not necessary for his market.

“Our target market is enterprise-grade customers, and usually they are using many cameras, and in many cases, those cameras are installed in more than one location, so that requirement to install everything in the camera or near the camera is not relevant for large installations,” he believes.


Sidebar: Bosch Expands Online Training Offerings

Bosch Security Systems Inc., Fairport, N.Y., offers two new eLearning courses: Access Personal Edition and the AutoDome Modular Camera System. These online, self-paced classes can be completed in sections whenever it is most convenient for the participant.

The Access Personal Edition course provides attendees with step-by-step instructions on how to install and use the access control software, configure the Access Modular Controller hardware and use card personalization to design and create badges. This course includes software demonstrations and simulations, and self-checks at the end of each module.

The AutoDome Modular Camera System program walks attendees through the installation of various mounts, including the parts needed for each interface option. The navigation of keyboard controls and on-screen display menus for configuring features, including TCP/IP connectivity, as well as basic troubleshooting are covered. The course incorporates interactive labs and review questions on the material.

Both courses also provide a final assessment and have received Continuing Education Unit (CEU) accreditation from the National Training School (NTS), a member service of the National Burglar and Fire Alarm Association (NBFAA). When a course is successfully completed with a score of 80 percent or higher on the final assessment, a certificate will be filled out with the attendee’s name, course name and NTS CEU information.

For additional information on these or other eLearning or instructor-led courses from Bosch, visit http://training.boschsecurity.us.


Russ Gager
gagerr@bnpmedia.com
Russ Gager is Senior Editor for SDM magazine.


Did you enjoy this article? Click here to subscribe to the magazine.

Resources + Guides

Buyers GuideBuyer's Guide
The industry’s most compre- hensive resource for security products, companies and suppliers.

eNewsletterseNews Signups
Subscribe to our free eNewsletters.

SDM 100 SDM 100























© 2008 BNP Media. All rights reserved. | Privacy Policy