Answers are grounded in this paper and cite their sources.
We describe a direct-matching approach to determining motion of a rigid body from two time-sequential perspective views. The method does not require finding feature correspondences. It is valid under the following assumptions: (1) Brightness constancy—the brightness of an image point corresponding to a fixed object point does not change after motion. (2) The mathematical form of the object surface is known to within a finite number of parameters. We also show how the computation can be simplified in the case of small motion.
No takes yet. Share an insight, caveat, or question.
Tinglin Huang (1986) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: