USE OF 2-DIMENSIONAL DEFORMABLE MESH STRUCTURES FOR VIDEO CODING .2. THE ANALYSIS PROBLEM AND A REGION-BASED CODER EMPLOYING AN ACTIVE MESHREPRESENTATION

This paper explores the use of the deformable mesh structure for motio n/shape analysis and synthesis in an image sequence, In Part I of this paper, we reviewed theory and techniques developed in the finite elem ent method for function interpolation and mapping using a given mesh s tructure, This constitutes the synthesis problem in a video coder empl oying a mesh-based motion model, Here in Part II, we present algorithm s for the analysis problem, including scene-adaptive mesh generation a nd node tracking over successive frames, We also describe a region-bas ed video coder that integrates the analysis and synthesis algorithms p resented in this paper, The coder describes each region by an ensemble of connected quadrilateral elements embedded in a mesh structure, For each region, its shape and texture are described by the nodal positio ns and image functions of the elements in this region in an initial fr ame, while its motion (including shape deformation) is characterized b y the nodal trajectories in the following frames, which are in turn sp ecified by a few motion parameters, This coder has been applied to a t ypical common intermediate format (CIF) resolution, head-and-shoulder type sequence, The visual quality is significantly better than the H.2 63-TMN4 algorithm at about 50 Kb/s (for the luminance component only, 30 Hz).