Context-Based Video CodingRichard Vigars, Andrew Calway, David Bull, Context-Based Video Coding. International Conference on Image Processing. September 2013. PDF, 1759 Kbytes.
We present a video CODEC framework which exploits extrin- sic scene knowledge to condition a perspective motion model. An approximate textural-geometric model of the scene is pre- pared prior to coding. During coding, the locations of planar surfaces in the scene are tracked, facilitating the computation of accurate perspective motion warp parameters. These algo- rithms are integrated with H.264 into a hybrid CODEC frame- work, achieving savings of up to 48% for equivalent visual quality.