Effective prediction using partition coding
The way of predicting a current block by assigning constant partition values to the partitions of a bi-partitioning of a block is quite effective, especially in case of coding sample arrays such as depth/disparity maps where the content of these sample arrays is mostly composed of plateaus or simple connected regions of similar value separated from each other by steep edges. The transmission of such constant partition values would, however, still need a considerable amount of side information which should be avoided. This side information rate may be further reduced if mean values of values of neighboring samples associated or adjoining the respective partitions are used as predictors for the constant partition values.
1. A decoder for reconstructing a depth map of a video signal using encoded information from a data stream, the decoder configured to:
derive a bi-partition of a block of the depth map into first and second portions;
predict the block of the depth map by determining a first predicted value for the first portion based on values of a first set of neighboring samples of the depth map, or determining a second predicted value for the second portion based on values of a second set of neighboring samples of the depth map, the first and second sets of neighboring samples adjoining the block of the depth map;
retrieve, from the data stream, one or more refinement syntax elements providing a first or second refinement value for the first or second predicted value, wherein the one or more refinement syntax elements include a syntax element indicating an absolute value of the first or second refinement value; and
refine the prediction of the block by applying the first refinement value to the first predicted value for the first portion or applying the second refinement value to the second predicted value for the second portion.
2. The decoder according to claim 1 , wherein the first predicted value includes a mean of values of the first set of neighboring samples and the second predicted value includes a mean of values of the second set of neighboring samples.
3. The decoder according to claim 1 , configured to, in applying the first or second refinement value, scale the first or second refinement value using a quantization step size depending on a reference quantization step size at which a predetermined spatially sampled component associated with the depth map is transmitted within the data stream.
4. The decoder according to claim 3 , wherein the decoder is configured to use the reference quantization step size in order to reconstruct a texture sample array corresponding to the depth map.
5. The decoder according to claim 1 , wherein the decoder is configured to, in deriving the bi-partition of the block of the depth map,
predict a position of a wedgelet separation line within the block of the depth map depending on a wedgelet separation line of a neighboring block such that the wedgelet separation line at the predicted position forms an extension of the wedgelet separation line of the neighboring block into the block of the depth map, and
refine the predicted position of the wedgelet separation line using line refinement information obtained from the data stream, the wedgelet separation line dividing the block of the depth map into the first and second portions.
6. The decoder according to claim 1 , further configured to use the block of the depth map as a reference in a prediction loop of the decoder.
7. The decoder according to claim 1 , wherein the one or more refinement syntax elements include a first syntax element indicating a sign value of the first refinement value and a second syntax element indicating a sign value of the second refinement value.
8. The decoder according to claim 1 , further configured to, in applying the first or second refinement value, linearly combine the first or second refinement value with the first or second predicted value, respectively.
9. An encoder for encoding into a data stream a depth map of a video signal, the encoder configured to:
derive a bi-partition of a block of the depth map into first and second portions;
predict the block of the depth map by determining a first predicted value for the first portion based on values of a first set of neighboring samples of the depth map, or determining a second predicted value for the second portion based on values of a second set of neighboring samples of the depth map, the first and second sets of neighboring samples adjoining the block of the depth map;
determine a first refinement value associated with the first portion based on a first original value associated with the first portion and the first predicted value, or determining the second refinement value associated with the second portion based on a second original value associated with the second portion and the second predicted value, wherein the first or second refinement value is applied to the first or second predicted value, respectively, for refining the prediction of the block of the depth map, and
encode, into the data stream, one or more refinement syntax elements representing the first or second refinement value, wherein the one or more refinement syntax elements include a syntax element indicating an absolute value of the first or second refinement value.
10. The encoder according to claim 9 , wherein the first predicted value includes a mean of values of the first set of neighboring samples and the second predicted value includes a mean of values of the second set of neighboring samples.
11. The encoder according to claim 9 , wherein, to refine the prediction of the block of the depth map, the first or second refinement value is linearly combined with the first or second predicted value, respectively.
12. The encoder according to claim 9 , further configured to quantize the first or second refinement value using a quantization step size depending on a reference quantization step size at which a predetermined spatially sampled component associated with the depth map is transmitted within the data stream.
13. The encoder according to claim 12 , wherein the encoder is configured to use the reference quantization step size in order to encode a texture sample array corresponding to the depth map.
14. The encoder according to claim 9 , wherein the encoder is configured to, in deriving the bi-partition of the block of the depth map,
predict a position of a wedgelet separation line within the block of the depth map depending on a wedgelet separation line of a neighboring block such that the wedgelet separation line at the predicted position forms an extension of the wedgelet separation line of the neighboring block into the block of the depth map,
refine the predicted position of the wedgelet separation line using line refinement information, the wedgelet separation line dividing the block of the depth map into the first and second portions, and
encode the line refinement information into the data stream.
15. The encoder according to claim 9 , further configured to use the block of the depth map as a reference in a prediction loop of the encoder.
16. The encoder according to claim 9 , wherein the one or more refinement syntax elements include a first syntax element indicating a sign value of the first refinement value, and a second syntax element indicating a sign value of the second refinement value.
17. A non-transitory computer-readable medium for storing data associated with a video, comprising:
a data stream stored in the non-transitory computer-readable medium, the data stream comprising one or more refinement syntax elements indicating a first or second refinement value associated with a respective first or second portion of a block of a depth map associated with the video, wherein the one or more refinement syntax elements are encoded using a plurality of operations including:
deriving a bi-partition of a block of the depth map into the first and second portions;
predicting the block of the depth map by determining a first predicted value for the first portion based on values of a first set of neighboring samples of the depth map, or determining a second predicted value for the second portion based on values of a second set of neighboring samples of the depth map, the first and second sets of neighboring samples adjoining the block of the depth map;
determining a first refinement value associated with the first portion based on a first original value associated with the first portion and the first predicted value, or determining the second refinement value associated with the second portion based on a second original value associated with the second portion and the second predicted value, wherein the first or second refinement value is applied to the first or second predicted value, respectively, for refining the prediction of the block of the depth map, and
encoding, into the data stream, the one or more refinement syntax elements representing the first or second refinement value, wherein the one or more refinement syntax elements include a syntax element indicating an absolute value of the first or second refinement value.
18. The computer-readable medium according to claim 17 , wherein the first predicted value includes a mean of values of the first set of neighboring samples and the second predicted value includes a mean of values of the second set of neighboring samples.
19. The computer-readable medium according to claim 17 , wherein the one or more refinement syntax elements include a first syntax element indicating a sign value of the first refinement value, and a second syntax element indicating a sign value of the second refinement value.
20. The computer-readable medium according to claim 17 , wherein, to refine the prediction of the block of the depth map, the first or second refinement value is linearly combined with the first or second predicted value, respectively.