Publication Date: 2003-Dec-18
The IP.com Prior Art Database
ITU-T H.264 or ISO/IEC MPEG-4, Part 10 allows using the same principles of spatial scalability known from the existing standards such as MPEG-2. This means that it is possible to encode a video sequence in two or more layers, each layer using a spatial resolution which is equal to or less than the spatial resolution of the next higher layer. The layers are related to each other in such a way that a higher layer represents the difference between the original sequence and a lower layer after it has been locally decoded and up-scaled to the spatial resolution of the original. H.264 employs elaborate block matching, where not only the different block sizes can be used for matching used but also different reference pictures. This can impose a tremendous computational load, especially in case several encoding passes need to be done to produce the final result. In general it can be observed that the base layer and the enhancement layer exhibit similar dynamics, which is explained by the fact that each layer originates from a same original sequence. This implies that the two layers need not be encoded completely separately, i.e. that some encoding decisions obtained during encoding of the base layer can be re-used during encoding of the enhancement layer. For example, the re-use of motion vectors is already known in MPEG-2-based schemes. It is possible to re-use in the similar way the information about reference picture selection. That is, th...