I have the following: cameraMatrix (intrinsic parameters), distortionCoefficients, and 4 detected points of a Rect. In world coordinate system these 4 points all have z-coord of 0 (Rect lays on the ground). I know the proportion of the rect in world coordinates (width and height).
I know that homography can be used to relate pixel coordinates of same plane in two different images.
I want to generate depth information of camera in relation to this plane (maybe out of homographyMat ?). So some information like: cameraHeight above ground (resp. relative camera position to the plane) and the angle/angles between optical camera axis and axis perpendicular to the plane.
Is this possible?