Really intertesting in depth explanation! I rack my brain for days to month about what happens if you have one camera position only but several frames. you only tilt or yaw it up or sideways. There would be no motion, no parallax whatsoever. Is it still possible to reconstruct a 3D point cloud with SfM Method then? If we start with the Equation of the Observation Matrix W, Camera Motion M and the 3D Point Cloud S W = M * S This should somehow have an effect on M, the Matrix for the Camera Position. Possibly this is even more complex than this and I don't understand how the effect on M could make the solution above unsolvable. Do you have an answer?
Really intertesting in depth explanation!
I rack my brain for days to month about what happens if you have one camera position only but several frames. you only tilt or yaw it up or sideways. There would be no motion, no parallax whatsoever.
Is it still possible to reconstruct a 3D point cloud with SfM Method then?
If we start with the Equation of the Observation Matrix W, Camera Motion M and the 3D Point Cloud S
W = M * S
This should somehow have an effect on M, the Matrix for the Camera Position. Possibly this is even more complex than this and I don't understand how the effect on M could make the solution above unsolvable.
Do you have an answer?