Rigid and Nonrigid Motion Analysis: Robust Recovery of 3-D Structure and Motion

This paper presents an incremental approach to understanding 3D structure and motion of rigid as well as nonrigid objects from image sequences. To be applicable to natural imagery like TV signals, the recovery process should meet two requirements: robustness to noise and ability of coping with deformable objects. These requirements can be met by the use of smoothness-of-motion constraint. The smoothness-of-motion constraint provides a framework for temporal integration of motion information over a longer sequence, rather than a set of image pairs. Based on the nature of motion, i.e, motion smoothness, 3D structure and motion information is successively and smoothly updated so as to agree with the observed transformation in the image. A model of rigid and nonrigid motion is introduced, and the smoothness-of-motion constraint is formulated as a stabilizing function. Some preliminary results are also given.