Method and apparatus for Intra prediction using non-square blocks

Patent metadata

Field	Value
Publication number	US-9769472-B2
Application number	US-201214234152-A
Country	US
Kind code	B2
Filing date	Jul 26, 2012
Priority date	Aug 17, 2011
Publication date	Sep 19, 2017
Grant date	Sep 19, 2017

Abstract

A method and apparatus for Intra prediction of a block based on neighboring pixels around the block are disclosed. Embodiments according to the present invention use square blocks as well as non-square blocks for Intra prediction. For a 2N×2N Luma CU (coding unit), the CU can be partitioned into 2N×N, N×2N, 2N×2N or N×N PUs. The 2N×N and N×2N PUs can be further processed by either square transforms only or both non-square and square transforms. In one embodiment, the 2N×N PU or the N×2N PU is processed as two N×N TUs (transform units) and each of the N×N TU is further split into smaller N×N TUs based on quad-tree split. In another embodiment, the 2N×N and N×2N PUs are processed as two 2N×0.5N and 0.5N×2N TUs respectively.

First claim

The invention claimed is:

1. A method for Intra prediction of a block based on neighboring pixels around the block, the method comprising:
utilizing a processor to perform:
receiving a 2N×2N Luma block corresponding to Luma pixels of the block;
if a size of the Luma block is larger than a smallest coding unit (SCU), setting a prediction block size of the Luma block to be one of: a square shape 2N×2N block size, a non-square shape 2N×N block size, or a non-square shape N×2N block size;
if the size of the Luma block is equal to the SCU, setting the prediction block size of the Luma block to be one of: a square shape 2N×2N block size, a non-square shape 2N×N block size, a non-square shape N×2N block size, or a square shape N×N block size; and
performing Intra prediction for the Luma block based on neighboring Luma pixels, wherein the Luma block is partitioned into one or more prediction units (PUs) according to partition mode.

2. The method of claim 1 , wherein the non-square shape PU is processed using only square transforms or using both non-square and square transforms, and wherein each transform unit (TU) is further split into smaller TUs based on a quad-tree split.

3. The method of claim 1 , the partition mode associated with said one or more PUs is indicated using a codeword table for the Luma block in Intra and Inter slices, wherein the codeword table includes entries for the 2N×N and N×2N PUs.

4. The method of claim 2 , wherein the 2N×N PU or the N×2N PU is processed either using square transforms only or using both non-square and the square transforms.

5. The method of claim 4 , wherein, when the 2N×N PU or the N×2N PU is processed as two N×N TUs (transform units), each of the N×N TU is further split into smaller N×N TUs based on quad-tree split if the N×N TU is larger than a smallest size, and wherein the smaller N×N TUs associated with the 2N×N PU or the N×2N PU use a same Intra prediction mode as the 2N×N PU or the N×2N PU respectively.

6. The method of claim 5 , wherein the smallest size corresponds to 4×4.

7. The method of claim 4 , wherein the 2N×N PU is processed as two 2N×0.5N TUs (transform units) and the 2N×0.5N TU is further split into smaller 2N×0.5N TUs based on quad-tree split if the 2N×0.5N TU is larger than a smallest size, and wherein the smaller 2N×0.5N TUs associated with the 2N×N PU use a same Intra prediction mode as the 2N×N PU.

8. The method of claim 7 , wherein the smallest size corresponds to 8×2.

9. The method of claim 4 , wherein the N×2N PU is processed as two 0.5N×2N TUs (transform units) and the 0.5N×2NTU is further split into smaller 0.5N×2NTUs based on quad-tree split if the 0.5N×2NTU is larger than a smallest size, and wherein the smaller 0.5N×2NTUs associated with the N×2N PU use a same Intra prediction mode as the N×2N PU.

10. The method of claim 9 , wherein the smallest size corresponds to 2×8.

11. The method of claim 4 , wherein the 2N×N PU is processed as two 2N×0.5N TUs (transform units) if the 2N×2N Luma CU is larger than 8×8, wherein the 2N×0.5N TU is further split into smaller 2N×0.5N TUs based on quad-tree split if the 2N×0.5N TU is larger than 16×4; and wherein the 2NxN PU is processed as two 4×4 TUs if the 2N×2N Luma CU has a size of 8×8.

12. The method of claim 4 , wherein the N×2N PU is processed as two 0.5N×2NTUs (transform units) if the 2N×2N Luma CU is larger than 8×8, wherein the 0.5N×2N TU is further split into smaller 0.5N×2NTUs based on quad-tree split if the 0.5N×2NTU is larger than 4×16; and wherein the N×2N PU is processed as two 4×4 TUs if the 2N×2N Luma CU has a size of 8×8.

13. The method of claim 1 , further comprising: receiving a Chroma block corresponding to Chroma pixels of the block; and performing Intra prediction for the Chroma block, wherein the Chroma block is partitioned into one or more Chroma PUs having the square shape only.

14. The method of claim 13 , wherein the Chroma block corresponding to a 2N×2N CU is Intra predicted using N×N block size, wherein a candidate set of Chroma Intra prediction modes is derived for the Chroma block.

15. The method of claim 13 , wherein an ordered mode set of Chroma Intra prediction modes consisting of {LM, DM 0 , DM 1 , Planar, Vertical, Horizontal, DC and Vertical+8} are used to derive a candidate set for the Chroma Intra prediction modes, wherein DM 0 and DM 1 correspond to two DMs (direct modes) associated with two non-square PUs of a 2N×2N Luma CU respectively, and wherein first six different modes of the ordered mode set are selected as candidate modes of the candidate set for Chroma Intra prediction.

16. An apparatus for Intra prediction of a block based on neighboring pixels around the block, the apparatus comprising at least one circuit configured for:
receiving a 2N×2N Luma block corresponding to Luma pixels of the block;
if a size of the Luma block is larger than a smallest coding unit (SCU), setting a prediction block size of the Luma block to be one of: a square shape 2N×2N block size, a non-square shape 2N×N block size, or a non-square shape N×2N block size;
if the size of the Luma block is equal to the SCU, setting the prediction block size of the Luma block to be one of: a square shape 2N×2N block size, a non-square shape 2N×N block size, a non-square shape N×2N block size, or a square shape N×N block size; and
performing Intra prediction for the Luma block based on neighboring Luma pixels, wherein the Luma block is partitioned into one or more prediction units (PUs) according to a partition mode.

17. The apparatus of claim 16 , wherein the non-square shape PU is processed using only square transforms or using both non-square and square transforms, and wherein each transform unit (TU) is further split into smaller TUs based on a quad-tree split.

18. The apparatus of claim 16 , the partition mode associated with said one or more PUs is indicated using a codeword table for the Luma block in Intra and Inter slices, wherein the codeword table includes entries for the 2N×N and N×2N PUs.

19. The apparatus of claim 17 , wherein the 2N×N PU or the N×2N PU is processed either using square transforms only or using both non-square and the square transforms.

20. The apparatus of claim 19 , wherein, when the 2N×N PU or the N×2N PU is processed as two N×N TUs (transform units), each of the N×N TU is further split into smaller N×N TUs based on quad-tree split if the N×N TU is larger than a smallest size, and wherein the smaller N×N TUs associated with the 2N×N PU or the N×2N PU use a same Intra prediction mode as the 2N×N PU or the N×2N PU respectively.

21. The apparatus of claim 19 , wherein the 2N×N PU is processed as two 2N×0.5N TUs (transform units) and the 2N×0.5N TU is further split into smaller 2N×0.5N TUs based on quad-tree split if the 2N×0.5N TU is larger than a smallest size, and wherein the smaller 2N×0.5N TUs associated with the 2N×N PU use a same Intra prediction mode as the 2N×N PU.

22. The apparatus of claim 19 , wherein the N×2N PU is processed as two 0.5N×2NTUs (transform units) and the 0.5N×2NTU is further split into smaller 0.5N×2NTUs based on quad-tree split if the 0.5N×2NTU is larger than a smallest size, and wherein the smaller 0.5N×2NTUs associated with the N×2N PU use a same Intra prediction mode as the N×2N PU.

23. The apparatus of claim 19 , wherein the 2N×N PU is processed as two 2N×0.5N TUs (transform units) if the 2N×2N Luma CU is larger than 8×8, wherein the 2N×0.5N TU is further split into smaller 2N×0.5N TUs based o

Assignees

• Liu Shan
• Zhang Ximin
• Lei Shaw-Min
• Mediatek Singapore Pte Ltd

Inventors

• Liu Shan
• Zhang Ximin
• Lei Shaw-Min

Classifications

• H04N19/136

  Incoming video signal characteristics or properties · CPC title

• H04N19/00024 Primary

  Electricity · mapped topic

• H04N19/70

  characterised by syntax aspects related to video coding, e.g. related to compression standards · CPC title

• H04N19/593

  involving spatial prediction techniques · CPC title

• H04N19/105 Primary

  Selection of the reference unit for prediction within a chosen coding or prediction mode, e.g. adaptive choice of position and number of pixels used for prediction · CPC title