Morphological Segmentation of Motion Field in Image Sequence Coding

We introduce a novel region-based video compression framework that uses a morphological operation to eeciently capture motion correspondences between consecutive frames in an image sequence. Our coder is built on the observation that the motion eld associated with typical image sequences can be segmented into component motion sub-eld \clusters" associated with distinct objects or regions in the scene, and further that these clusters can be ee-ciently captured using morphological operators in a \back-ward" framework that avoids needing to explicitly send object boundaries. Cluster matching is used in lieu of the conventional block matching methods of standard video codecs to deene a cluster motion representation paradigm. Preliminary experimental results reveal a 10-20% reduction in motion compensated prediction error energy using our proposed motion compensation framework on the Football sequences over standard block motion compensation methods like MPEG, while also requiring less computational complexity and allowing for better noise rejection.