People always ask: “Where should I use analytics?” Our answer: anywhere you hang a camera. Listen, every camera that is ever installed should have a reason it was selected with that specific field of view. We refer to this as the camera’s “purpose.” The goal of analytics is to detect when the purpose happens while ignoring everything else.
Roughly 85-90% of analytics applications used today fall under what we refer to as “Area Analytics” which can be defined loosely as “Something, some place, longer than you want it to be there.” This means there is not only a size and shape of the object but also a temporal value to that object’s occupation of the space. This could be 0.1 seconds or it could by 3 minutes for example and defines the type of event that the user is interested in. We could not care about people walking up and down a path all day long but when they stop in a specific area for a significant time interval, we want to know. Using built-in filters to the software allow you to shape and define what will create an event and what will be ignored.
The key to understanding when, where or how to use analytics is rooted in understanding the purpose of the camera you have installed. For the vast majority of real-world cameras, the purpose is to detect some form or trespassing. This is a match made in heaven for a true video analytics-based solution. Why do you ask? Well I got into all sorts of trouble on another blog site (any guesses) a couple years ago when I stated that area analytics are the easiest to set-up, require the least amount of time for tuning, and give the most accurate detection of any analytics application on the market. Now I was not referring strictly to Arteco, but to all true machine-vision based video analytics platforms on the market. My research came from speaking with various manufacturers as well as integrator technicians who have installed and set-up leading applications. To anyone who has experience setting up applications, there is a tangible and significant difference between the time/complication to set up a violated area and speed threshold detection (for example).
The reason I say this is a perfect match for analytics now is because it is an application that is commercially viable in the industry. It is a marriage of what is needed in the market (a way to streamline video resources for trespassing) and an application that gives high rates of accuracy with the least amount of set-up pain. There are all sorts of examples of in recent years of applications and sales models that do not work well when it comes to video analytics. By focusing on understanding the purpose of the camera and what the customer’s expectation of what they are trying to achieve with that specific field of view , we easily understand how to use video analytics applications to help increase the efficiency and effectives of the security profile.
