Motion Compensated Video Super Resolution

摘要

In this paper we present a variational, spatiotemporal video super resolution scheme that produces not just one but n high resolution video frames from an n frame low resolution video sequence. We use a generic prior and the output is artifact-free, sharp and superior in quality to state of the art home cinema video processors. Unlike many other super resolution schemes, ours does not limit itself to just translational or affine motion, or to certain subclasses of image content to optimize the output quality. We present a link between image reconstruction and super resolution and formulate our super resolution constraint with arbitrary up-scaling factors in space from that.