Video analytics use scientific algorithms to automatically detect people, vehicles or other objects viewed via your CCTV video images. This sophisticated technology can also detect advanced events such as object left or removed, moving across a line, direction of travel and loitering. The system can alert Remote Video Response (RVR) to trigger a response dependent on pre-programmed instructions.
ANPR solutions vary from simple number plate image recognition, up to an intelligent event driven control e.g. barrier control, car park management, site traffic management etc. the more intelligent systems can include foreign number plate recognition and capturing the number plates of multiple vehicles travelling at speed. Once the number plate is captured it can trigger predefined events e.g. black listed vehicles.
Video analytics offer a cost effective alternative to traditional detection methods. This is done by using algorithms allowing the automatic detection of people, vehicles and other objects from video images. They can also detect movement, direction of travel and the appearance or disappearance of items. By classifying the objects it allows the system to include or exclude people, vehicles or animals from the analysis. It is also possible to completely ignore areas of no interest. False alarms can be vastly reduced by using this technology as it can filter out things like trees, clouds, reflections, rain etc. Actual applications can include, loitering, object left, object removed, marketing functions i.e. shopping behaviors, checkout opening in a busy store. Different behavioral patterns can be ignored or activate an alarm dependent on the time of day i.e. loitering in a car sales forecourt which could offer a sales opportunity or identify a security threat.
Intelligent processing at the camera can detect a number of events i.e. intrusion into a predefined area, crossing a predefined line, variable audio from ambient level, camera defocussing and scene change. Any of these, or a combination of one or more, can then trigger an alarm event.
Temporal Events that did not last very long
Spatial Events that have happened in specified areas within the image
This can be achieved by using:
Edge processing e.g. contained within the camera itself
Server based analysis e.g. contained within a separate device
To learn more, please click on the arrow.