Video Surveillance System architecture consists of four key components. They are Security Surveillance cameras, Network Infrastructure, Video Storage, Video Management System, and the last but not the least Video Analytics (VA).
Though, Video Analytics is an optional component, which can be induced into the surveillance deployment on a voluntary note; its inclusion is said to make the surveillance deployment fool proof.
VA can be implemented in two different configurations and that are-
Server based Video Analytics- In this approach, Video Analytics is implemented on a dedicated server which pulls the video and analyzes it in order to issues alerts or obtained results from analysis. Since, the whole process is independent of the video cameras; this type of deployment is seen in most of the VA deployments.
Basically, in this approach, the video gets transferred from a video management system to a dedicated server, which can quench the high processing needs of video Analytics and then the video frames are analyzed.
But during the analysis, the video quality gets degraded due to compression and transmission effects and so the performance of the overall server diminishes. So, most surveillance deployments go for a dedicated VA server which can only handle up to 16-18 cameras.
But nowadays, as technology is progressing, Video Management System vendors such as DNF Security are coming up with a 4-in-1 system which can record, manage, archive and do video analytics on recorded videos. DNF Security Falcon Extreme is one such machine which has high processing power and is loaded with the capability of handling the processing of more than 125 HD cameras at a time. Moreover, this machine is ONVIF complaint and so is compatible to work with all leading Camera and Video Management Software brands.
Edge based Recording- In this approach, analytics is implanted into the video camera in the form of a smart video encoder( required for analog cameras, whereas IP cameras are coming up with embedded VA skills) having the capability of running video analytics functionality.
But what most surveillance managers’ report is that this approach doesn’t yield satisfactory results as it imposes limitations on the overall surveillance system design and performance. They report that most devices lack sufficient processing power for high end VA requirements and fail to live up to their expectations.
Now, coming to Video Analytics types, here is a list and has many more to add
- People Counting
- Directional Alert
- Object left behind the scene
- Object taken from a scene
- Noise Cancellation
- Motion Detection
- Thumbnail Searching
- Face Detection
- OCR Detection
- Change of Camera view
- Speed of Vehicle Alert
- License Plate Recognition
- Digital Black Light Compensation
- Loitering Alert
Share your knowledge in the comments section and feel free to add to the list
storageservers.wordpress.com 