VIDEO COMPRESSION METHOD ALLOWING SELECTIVE VIDEO OBJECT EDITING

摘要

In a MPEG or other video system, editing is performed using "independently coded regions" (ICRs) embedded into an input video signal. These regions facilitate quick editing and formatting of a compressed output signal; for example, logo insertion, color correction, blue-matting an various types of image sequence mixing and manipulation may be performed by decoding and processing individual regions rather than entire frames. Preferably, each independently coded region is recognizable directly from a compressed video bitstream and is retrieved by decoding select image slices. Once editing or processing is complete, new compressed bitstream data is inserted into the place of the original compressed data. Each independently coded region features the attribute that motion vector and residual data for compressed frames are limited to point only to corresponding regions of anchor frames.