| 摘要 |
An image portion is encoded or decoded (e.g. by HEVC) by prediction using a predictor determined as a function of both a reference image portion and motion information item (e.g. motion vector, MV) values. After obtaining the first item value (e.g. prediction mode, resolution of a MV), the second item value (different, e.g. prediction mode, reference frame index, number of MV predictors, reference lists) is evaluated as a function of the obtained first item value. Predictor is then determined as a function of both the reference image portion and the first and second item values. Third item (e.g. prediction mode) value may be obtained, the second item value being evaluated as a function of the third. Also disclosed is encoding or decoding first and second image portions by prediction using predictors determined as a function of both a reference (different) image and MVs. First and second (different) resolutions of MVs are associated respectively with a prediction unit of the first and second portions. After obtaining MV information, first and second MVs associated with the first and second portions respectively are evaluated as a function of the obtained MV information. First and second predictors are then determined for the first and second portions respectively. |
