Affine motion compensation using a content-based mesh

Abstract

A content-based approach to the design of a triangular mesh is presented, and its application to affine motion compensation is investigated. An image is first segmented into moving objects, which are then approximated with polygons. Following this, a triangular mesh is generated within each polygon, thus ensuring that no triangle straddles multiple regions. Translation and affine motion parameters are determined for each triangle, using bi-directional motion estimation. Results for three test sequences demonstrate the advantages offered by the proposed mesh design method, and by the use of affine motion compensation.

Improving Piecewise Linear Registration of High-Resolution Satellite Images Through Mesh Optimization

“…2) Only the geometrical realization is modified by refining the vertex coordinates (approach typically employed in image registration) [17]- [20]. 3) Both the topological and geometrical realizations are simultaneously refined by splitting/collapsing/swapping edges and refining the vertex coordinates [21]- [24].…”

“…One example of this is the work in [17], which relocates the mesh vertices (the mesh topology remains fixed) in order to compensate for the affine motion in video streaming. Vertex coordinate refinement, although being suitable for coping with smooth image distortions, does not provide enough correction to accommodate the possibly important geometric differences between the satellite images when they are acquired from offnadir angles.…”