Description
The exact boundary of the camera's vision rectangle is sometimes difficult for non-technical users to intuit, and for an application I'm working on I'd like to make it easy for non-technical users to re-define the boundaries of different regions of the monitored space.
It would be convenient, therefore, to have some type of artificial marker the user can place in the scene as the method of input: i.e. to define a particular polygon where some effect occurs, they just need to set up the camera to point generally in the area, then place the markers at the corners of the polygon and press a button to tell the application to memorize their locations; and the application can then compare later person bounding rectangles to those stored points.
I'm sure there'd be myriad other use cases for being able to track markers with the same sensor and in the same coordinate space as the person tracking info.