StereoRectify with a single moving camera
Hello,
I am looking to make use of the stereoRectify function, but am unsure how I should calculate the rotation matrix R and translation vector T.
I know typically these come out in the wash if you use stereoCalibrate(). However in my situation, I only have a single camera which is moved. At each of the two points I take an image, I know the camera's position & orientation.
Any advice would be most welcome.
if you got a single camera only, but multiple shots from that, the problem must be renamed to "structure from motion" , you simply can't use the stereoCalibrate -> rectify -> blockmatching pipeline, but you will have to use triangulation/ bundling