2013
07.29

[msu1][Camera 2][2013-07-25 21-59-28][Person]

Most cameras have the ability to generate some sort of alert based on “motion detection”. That basically means any pixel change will generate an alert or alarm. Not very effective when trying to monitor a site, as you get an alert with any lighting change, anytime a tree or bush moves in the wind, etc. You can mask out areas within the cameras view, but you will still generate a high level of false alerts.

Video analytics learn the scene within a cameras view, and can tell the difference between people, vehicles, etc., and can disregard smaller insignificant events. Analytic cameras can utilize loitering or dwelling rules, to generate an alert if a person or vehicle is an area beyond a certain time frame. In addition, you can be alerted when an object is placed or removed from within a scene. Many other options are available to include people counting, etc. Video analytics have come a long way, and are becoming very affordable. This makes remote video monitoring and detection much more successful when operators can spend their time looking at important events, as opposed to multiple false alarms. Attached is an example of person detection with analytics. You can see the camera recognizes the person within the scene, and alerts the operator to the event. This is perfect for remote sites, where nobody should be.

No Comment.

Add Your Comment

Comments are closed.