Analytics as Resource Management

One thing we always stress about the use of analytics in video security is resource management.  In essence, video analytics are the ultimate resource management tool.   Used correctly, video analytics are “purpose based” meaning that they allow the camera to become a proactive sensor that selectively chooses when video is sent over a network and how it is stored.

Five years ago the big buzz was “anti-terrorism” with analytics.  Everything was about catching the backpack left behind on the train platform.  That was the major reason why (the conventional wisdom went) analytics would take over the video security space.  This reached a fever pitch around 2008 when you could not throw a rock at ISC West without hitting somebody peddling “Intelligent Video” in some fashion or other.    We now have come to realize that a secondary (and arguably equal) benefit to video analytics is in its ability to fight the ever-growing amount of data that needs to be processed, stored, and transmitted by video management systems.

There are two major contributing factors to the increase in data: growing camera counts and development of megapixel cameras.  On one hand we are reaching new levels of ubiquity in the United States market when it comes to security cameras.  People are just more comfortable and accepting to the benefits of security cameras than they ever have been before.  This is driving up the number of cameras that the customers are purchasing.  High Schools that may have purchases 20 cameras 15 years ago are now in the 100+ camera range.  Not only are customers dramatically increasing camera counts, the IP camera segment is now engaged in full-on Megapixel Race to see who can put out the most robust image-quality camera at the least expensive price.  It is perfect when you think about it: what has always been the number one complaint about CCTV?  BAD VIDEO.  We saw this with the old VHS tapes that were recorded over and over.  We continued to see this with digital video due to low quality analog cameras and capture cards putting out degraded images.  Video quality is the one attribute to video security that all people, regardless of experience, can easily grasp and judge.  Given a choice between a CIF analog image and a 5MP IP image in a side-by-side quality comparison is the quintessential NO BRAINER.  

The problem, and what many security professionals either ignore or (yikes!) don’t understand, is that the 500 lbs gorilla sitting in the corner of the room is named Bandwidth.  Does not matter if you are talking about processing power, storage capability, or network transmission; more cameras with bigger images consume exponentially more resources.  Using Analytics allows you to manage the data that needs to be processed, moved and stored on the system by allowing you to control the what, where, when, and why aspects of this data.   What types of events are you interested in seeing?  Where do you want events to be sent?  When should resolutions and frame-rates be accelerated or reduced?  Why are specific video events more important than the majority of the video I am recording?  Why should I take up valuable network bandwidth when nothing of importance is taking place?  All of this begins with understanding the purpose of the camera that you have installed.  If you understand the purpose there is no reason why video analytics cannot play a fundamental role in resource management.