I imagine placing a green treadmill in a green room with red, blue, yellow, and orange dots demarking the corners of the treadmill. The top left of the treadmill will be a red dot that will augment the 3 dimensional object over the treadmill perfectly. The treadmill will be represented by a Fortnite 2 by 4 that a actor runs up by using the treadmill. I have the detection done with Tensorflow and OpenCV to detect the dots from a camera. I now need a way to represent a 3 dimensional object model augmented over the real life data points of the treadmill's four corners for a virtual camera to match the perspective of the real camera.