Projective decomposition of planar facets

Since for recognition tasks it is known that planar invariants are more easily obtained than others, decomposing a scene in terms of planar parts becomes very interesting. This paper presents a new approach to finding the projections of planar surfaces in a pair of images. For this task we introduce the facet concept defined by linked edges (chains) and corners. We use collineations as projective information to match and verify their planarity. Collineations are constrained by the fundamental matrix information and a Kalman filter approach is used to refine its computation. Furthermore the Kalman filter is used to enrich the number of detected facets by finding coplanar facets.