“Geo-referencing” a camera allows it to become a very powerful detection and tracking device. For example, you may have dog in a video which manifests itself as 500 pixels, a very substantial object. But let’s assume we really don’t care about detecting dogs, we are more interested in detecting a human, which is farther away and only accounts for 50 pixels. Since we have now enabled our camera with “geo-intelligence” it has the added information of knowing where these objects occur in real space. As a result, the system can compute lots of additional information about the target, including its real height, real width, real speed, real acceleration, etc. All these parameters can then be used with other videos cues to place the object in a classification category; person, animal, car, truck, boat, etc. So a geospatial system can now be smart enough to ignore the dog and alert on the human. This video shows a geospatially enabled camera in action.
You’ll see the targets appear on the video, in this case a car and person.
You can see the detection happen in the video.
The alarm location is then indicated on the map with triangle
While this is happening you’ll see the system classifying the targets as vehicle, human or “not sure”
All the time dynamically updating their tracks on the map.