|
The iSentry system is an automated video analysis system. It uses Artificial Intelligence (AI) to learn the normal behavioural patterns of objects observed by a video camera. When an object displays activity that does not fit the model of normal behaviour, the system highlights it as unusual activity and can perform a number of automated responses. iSentry transforms a security environment from passive to active.
iSentry uses Computer Vision, developed for autonomous robots, to track every moving object in the camera’s view. It measures the size, shape, speed and direction of motion of every object it sees. It builds up a model of this behaviour, which it considers normal. When an object moves in a way that does not fit the normal model, iSentry identifies it as unusual.
iSentry’s vision system works excellently at night as well as during the day. Flashing lights do not distract iSentry. Even in low lighting, it can track its targets accurately. iSentry also excels in outdoor environments. Its vision is not tricked by clouds or lighting changes. It even learns the normal behaviour of the wind blowing in the trees and ignores it. Regular traffic, whether vehicular or pedestrian, is learned over time and will not raise an alert.
The unique ability of the iSentry system is the ability to self-learn any give environment, detecting unusual activity and ignoring normal activity. This makes the system far superior to Video Motion Detection (VMD) systems, particularly in busy environments. In busy environments, iSentry is able to filter out around 95% of the “normal” motion observed and only react to the remaining 5% of unusual activity.
|