Using epipolar geometry for a pair of cameras, E = t^R, where t can be any multiple of the t found by SVD. Is there any way to obtain the metrically correct value of t if we have performed loop closure using a sequence of monocular images?
Using epipolar geometry for a pair of cameras, E = t^R, where t can be any multiple of the t found by SVD. Is there any way to obtain the metrically correct value of t if we have performed loop closure using a sequence of monocular images?