Method and apparatus for decoding a video block
A co-located block of a video block to be decoded is determined. The co-located block is either (i) a forward reference block or (ii) a backward reference block. A reference motion vector is selected for the video block. When the co-located block is backward reference block and has a forward reference motion vector and a backward reference motion vector, the forward reference motion vector of the co-located block is selected. When the co-located block is forward reference block and has a forward reference motion vector and a backward reference motion vector, the backward reference motion vector of the co-located block is selected. When the co-located block has only one reference motion vector, the one reference motion vector of the co-located block is selected. When the co-located block has no reference motion vector, a zero-reference motion vector for the video block is selected. A motion vector for the video block is derived using the selected reference motion vector, and the video block is decoded using the derived motion vector.
1. A method for decoding a video block, the method comprising:
determining a co-located block of the video block, the co-located block being either (i) a forward reference block or (ii) a backward reference block;
selecting a reference motion vector for the video block, wherein the selection of the reference motion vector comprises:
when the co-located block is backward reference block and has a forward reference motion vector and a backward reference motion vector, selecting the forward reference motion vector of the co-located block,
when the co-located block is forward reference block and has a forward reference motion vector and a backward reference motion vector, selecting the backward reference motion vector of the co-located block,
when the co-located block has only one reference motion vector, selecting the one reference motion vector of the co-located block, and
when the co-located block has no reference motion vector, selecting a zero-reference motion vector for the video block;
deriving a motion vector for the video block using the selected reference motion vector; and
decoding the video block using the derived motion vector.
2. The decoding method according to claim 1 , wherein the selection of the reference motion vector for the video block further comprises:
decoding from a bitstream a position flag indicating whether the co-located block is forward reference block or backward reference block; and
determining whether the co-located block is forward reference block or backward reference block based on the decoded position flag.
3. A decoding apparatus for decoding a video block, the decoding apparatus comprising a processor and memory, the processor, working together with the memory, being configured to:
determine a co-located block of the video block, the co-located block being either (i) a forward reference block or (ii) a backward reference block;
select a reference motion vector for the video block, wherein the selection of the reference motion vector comprises:
when the co-located block is backward reference block and has a forward reference motion vector and a backward reference motion vector, select the forward reference motion vector of the co-located block,
when the co-located block is forward reference block and has a forward reference motion vector and a backward reference motion vector, select the backward reference motion vector of the co-located block,
when the co-located block has only one reference motion vector, select the one reference motion vector of the co-located block, and
when the co-located block has no reference motion vector, select a zero-reference motion vector for the video block;
derive a motion vector for the video block using the selected reference motion vector; and
decode the video block using the derived motion vector.
4. The decoding apparatus according to claim 3 , wherein the processor, working together with the memory, is further configured to, as part of the selection of the reference motion vector for the video block:
decode from a bitstream a position flag indicating whether the co-located block is forward reference block or backward reference block; and
determine whether the co-located block is forward reference block or backward reference block based on the decoded position flag.