Trends in surveillance: full motion video and video analytics
With increased security concerns in recent years, surveillance has been intensified at airports, seaports, in cities, at borders, etc. To improve the information gathered from these systems, the big trends are the move from still imagery to full motion video and then better video analytics to automatically evaluate the images for unusual events. A ReportsNReports analysis estimated the size of the smart surveillance and video analytics global market at $13.5 billion in 2012, and it's expected to reach $39 billion by 2020.
In comparison to still images, full motion video allows for monitoring activity over time, observing behaviors, and also reduces the chance of missing critical details through continuous viewing.
Two of the challenges for full motion video have been how to capture high quality images regardless of vibration, changing lighting conditions, environmental conditions such as fog, etc. and how process and store all of the data that is obtained.
The quality imagery issue has been solved, as there are rugged video cameras available that provide daylight HDTV for full motion video. They include optimized functionality to adjust for changing lighting conditions and other atmospheric conditions. There are true rugged options that can provide non-stop operation on airborne platforms.
Demonstration of Adimec's Video Contrast Enhancement Function:
Now with excellent source material there needs to be improvements with how to efficiently process the information. The amount of full motion video collected is astronomical. For instance, it has been reported that the data feed from various US drones amassed 327,384 hours (37 years) in 2011 of video footage.
There is a drive towards better video analytics, which involves computer/software processing of the video so that a human does not have to view images in which there is nothing significant occurring.
The goal is to automate the analytics processes using technologies such as pattern recognition and change detection. Ideally this would be done right on the imaging platform so that only the portions of video where something is detected are then transmitted as there is limited bandwidth and storage availability.
While there have been video cameras in use for many years, it was difficult to keep up with the data and view all of the information. Advances in camera technology and video analytics are supporting the need for smart intelligence from full motion video.
Full motion video camera requirements for surveillance and situational awareness
In-camera video contrast enhancement to see through the fog