Estimating Multiple Independent Motions in Segmented Images using Parametric Models with Local Deformations

This paper presents a new model for optical ow based on the motion of planar regions plus local deformations. The approach exploits brightness information to organize and constrain the interpretation of the motion by using segmented regions of piecewise smooth brightness to hypothesize planar regions in the scene. Parametric ow models are tted to these regions in a two step process which rst computes a coarse t and then reenes it using a generalization of the standard area-based regression approaches. Since the assumption of pla-narity is likely to be violated, we allow local deformations from the planar assumption. This para-metric+deformation model exploits the strong constraints of parametric approaches while retaining the adaptive nature of regularization approaches.