SYNTAX AND SEMANTICS FOR ADAPTIVE LOOP FILTER AND SAMPLE ADAPTIVE OFFSET

摘要

The present invention provides an improved video encoding and decoding method, which maintains the advantages of LCU-based filter parameter signaling as compared to frame-based filter parameter signaling, but considerably reduces signaling overhead. Therefore, signaling syntax is modified by grouping LCUs (Largest Coding Units) together for signaling employing a mapping function. Consequently, filter parameters no longer need to be signaled for each single LCU, but for a group of several LCUs. The syntax structure of the invention avoids redundancies present in the state of the art as far as possible and thus increases the information content of the syntax elements. At the decoder side, the mapping function is applied to infer information about the filter parameters to be applied to a current LCU from information encoded in different syntax structures.

主权项

1. A decoding method for decoding a picture from a bitstream on the basis of Largest Coding Units, the decoding method comprising:
obtaining group information specifying the number of groups in the picture, each of the groups including one or more Largest Coding Units, by parsing a first syntax structure of the bitstream; obtaining unit information specifying the number of the Largest Coding Units in the picture, by parsing a second syntax structure of the bitstream; and deriving filter parameters for a current group to be applied in decoding a current Largest Coding Unit in the current group using the group information and the unit information.