What’s In A Video Analytic

By Chris Farlow

If we ask Wikipedia about video analytics we get redirected to video content analysis (VCA) and information about the capability of automatically analyzing video to detect and determine temporal and spatial events.
Many different functionalities can be implemented in VCA. Video Motion Detection is one of the simpler forms where motion is detected with regard to a fixed background scene. More advanced functionalities include video tracking and egomotion estimation.

Now, I don’t really know what egomotion is but I do understand video tracking. And those video analytics offer a wide variety of capabilities for the system to alarm and alert the ‘operator’ that something has happened.

If we look at a popular video analytic tool, Briefcam, they describe themselves as an “artificial intelligence” (AI) tool that allows for “tracking and identifying suspects” within a video feed and even “finding all the people with red shirts.”

As video analytic tools such as Briefcam improve on accuracy and capabilities, these tools are becoming extremely valuable to law enforcement and loss prevention teams. What would take individuals a multitude of hours to review (and hopefully not miss), video analytic tools can find “the needle in the haystack” in mere minutes.

Another machine learning tool, although not necessarily a specific video analytic, is Snap Surveillance’s Force Multiplier tool. This AI-based add-on tool for leading video management system (VMS) systems is an operator assisted tracking tool that significantly aids the operator of hundreds (even thousands) of cameras in tracking an object or person from one camera field of view to another camera field of view.

A modern Security Operation Center resembles those of a Hollywood movie or TV crime show, with banks of monitors displaying many —many— camera feeds What is not mentioned is when those camera views on the bank of monitors is —not— set up so that each camera is “aside” it’s associated camera.

Meaning, that if an object or a person is walking from left to right on one camera feed, when they get to the “edge” of that camera’s view, what is the “next” camera the object or person going to appear in?

Now, if this is a location with 12 cameras, I —as the operator— probably know which camera video feed is “next in line” and I can easily click on the list of cameras and choose the right camera feed.

If my location has hundreds —or even thousands— of cameras, picking out that “next” camera feed quickly enough so as I don’t lose track of that object or person could be problematic.

Snap’s Force Multiplier is uniquely built on an AI “camera learning” system that uses algorithms to learn the relationship of all the cameras in a location and where one camera “feeds into” another from various angles.
Thus, if in our example above, our operator was tracking a person that is walking from left to right on one camera feed, Force Multiplier would display the camera feed that is associated to the right of the current feed.

If the person we were tracking abruptly stops and turns left (up on our video wall) Force Multiplier would already have the camera feed that would be associated with moving “up” in our display and the operator would only need to click on that camera view to make it the new center ‘focus’ video feed.

And as we continue to track the person -as they zigzag around our field of view- I as the operator am simply clicking on the next “associated” camera feed that Force Multiplier has already determined is the “adjacent” camera feeds.

No more hunting through the long list of cameras to hopefully find the right “next” camera feed. No more switching display views to keep the right cameras up in the display. No more guessing as to what is the camera feed “to the left” of the current view.

With Snap’s Force Multiplier AI-based solution, it is not picking out the “person in the red shirt” for you, but it is assisting each operator with all of the possible camera feeds as they track an object or person around campus.

Snap’s Force Multiplier also allows you to fast-forward the recorded video to catch up to ‘current time’ to see where your suspect has been and where they are located right now. And playing recorded video back in fast-rewind mode will allow you to quickly and easily see where your suspect came from and how they got on the property.

Video Analytics come in all shapes and sizes. And although Force Multiplier is an AI-based system, it is a must-have tool for environments with a large number of cameras installed and operators that are constantly tracking objects and people around campus.

Stop wasting time trying to find where a suspect is “now” or how they “got onto campus.” Utilize Snap’s Force Multiplier tool to help your operators be the most efficient and effective operators they can be.

About The Author
Chris Farlow is North American Solutions Engineer for Snap Surveillance. Snap provides a low-cost software solution that helps turn your existing fleet of cameras into a powerful interconnected network of camera relationships. Snap enables surveillance system users of any experience to easily track live events and quickly recreate steps for forensic purposes. Designed for the 100+ enterprise camera environment, Snap Surveillance’s Force Multiplier fully integrates with a range of video management software packages.
Learn more about Snap Surveillance at SnapSurveillance.com or call 760-429-3444.

Source: snapsurveillance.com