| 摘要 |
Video encoding and decoding comprising determining a block vector prediction candidate and a block vector difference for a prediction unit of an image of a media presentation and, if the block vector candidate represents an absolute position of a source prediction block, determining a block vector predictor by calculating a difference between the position of the source prediction block and the position of the prediction unit, and determining a block vector for the prediction unit by summing the determined block vector difference and the determined block vector predictor. A first list of block motion vector prediction candidates and a second list of prediction block candidates may be maintained using rules and updated if changes are made. The block vector prediction unit may represent a block in same picture frame than the prediction unit, or if it represents a different picture frame then an indication of a reference image where MV is pointing to is coded. The media presentation may be multi-view, so the block vector prediction unit may represent a block in a different view than prediction unit with an indication of a reference view where MV is pointing to being coded. |
