RESIDUAL TREE STRUCTURE OF TRANSFORM UNIT PARTITIONING

1 . A method comprising:

receiving a prediction unit (PU) for a coding unit (CU) of video content, wherein the PU is partitionable into a plurality of PU partition types;

determining, by a computing device, a PU partition type for the PU;

determining, by the computing device, a residual tree structure based on the PU partition type for partitioning of the CU into transform units (TUs), the residual tree including a binary partition of a node into two TUs;

determining, by the computing device, a TU partition for the PU partition type based on the residual tree structure and a desired level of partitioning in the residual tree structure; and

using the TU partition in a transform operation.

2 . The method of claim 1 , wherein:

for a 2N×N PU partition type, a level in the residual tree structure partitions a 2N×2N node into two nodes of 2N×N, and

for a N×2N PU partition type, a level in the residual tree structure partitions a 2N×2N node into two nodes of N×2N.

3 . The method of claim 1 , wherein:

for a 2N×0.5N PU partition type, a level in the residual tree structure partitions a 2N×N node into two nodes of 2N×0.5N, and

for a 0.5N×2N PU partition type, a level in the residual tree structure partitions a N×2N node into two nodes of 0.5N×2N.

4 . The method of claim 1 , wherein:

for a 2N×0.5N PU partition type, a level in the residual tree structure partitions a 2N×2N node into a first node of 2N×0.5N and a second node of 2N×1.5N, and

for a 0.5N×2N PU partition type, a level in the residual tree structure partitions a 2N×2N node into a first node of 0.5N×2N and a second node of 1.5N×0.5N.

5 . The method of claim 1 , further comprising signaling the desired level from an encoder to a decoder.

6 . The method of claim 1 , wherein an encoder and a decoder implicitly determine the desired level independently.

7 . The method of claim 1 , wherein when a default depth of the residual tree is used, the desired level of the residual tree for each PU partition type is fixed.

8 . The method of claim 1 , wherein the residual tree structure includes nodes that branch out with the binary partition of TUs and a quadtree partition of TUs.

9 . An apparatus comprising:

one or more computer processors; and

a non-transitory computer-readable storage medium comprising instructions, that when executed by the one or more computer processors, control the one or more computer processors to be configured for:

receiving a prediction unit (PU) for a coding unit (CU) of video content, wherein the PU is partitionable into a plurality of PU partition types;

determining a PU partition type for the PU;

determining a residual tree structure based on the PU partition type for partitioning of the CU into transform units (TUs), the residual tree including a binary partition of a node into two TUs;

determining a TU partition for the PU partition type based on the residual tree structure and a desired level of partitioning in the residual tree structure; and

using the TU partition in a transform operation.

10 . The apparatus of claim 9 , wherein:

for a 2N×N PU partition type, a level in the residual tree structure partitions a 2N×2N node into two nodes of 2N×N, and

for a N×2N PU partition type, a level in the residual tree structure partitions a 2N×2N node into two nodes of N×2N.

11 . The apparatus of claim 9 , wherein:

for a 2N×0.5N PU partition type, a level in the residual tree structure partitions a 2N×N node into two nodes of 2N×0.5N, and

for a 0.5N×2N PU partition type, a level in the residual tree structure partitions a N×2N node into two nodes of 0.5N×2N.

12 . The apparatus of claim 9 , wherein:

for a 2N×0.5N PU partition type, a level in the residual tree structure partitions a 2N×2N node into a first node of 2N×0.5N and a second node of 2N×1.5N, and

for a 0.5N×2N PU partition type, a level in the residual tree structure partitions a 2N×2N node into a first node of 0.5N×2N and a second node of 1.5N×0.5N.

13 . The apparatus of claim 9 , wherein when a default depth of the residual tree is used, the desired level of the residual tree for each PU partition type is fixed.

14 . The apparatus of claim 9 , wherein the residual tree structure includes nodes that branch out with the binary partition of TUs and a quadtree partition of TUs.

15 . A method for decoding video content, the method comprising:

receiving a bitstream of encoded video content;

determining, by a computing device, a prediction unit (PU) partition type for a PU for a coding unit (CU) of the video content, wherein the PU is partitionable into a plurality of PU partition types;

determining, by the computing device, a residual tree structure based on the PU partition type for partitioning of the CU into transform units (TUs), the residual tree including a binary partition of a node into two TUs;

determining, by the computing device, a TU partition for the PU partition type based on the residual tree structure and a desired level of partitioning in the residual tree structure; and

using the TU partition in a transform operation in decoding the video content.

16 . The method of claim 15 , wherein determining the TU partition comprises receiving the desired level from an encoder.

17 . The method of claim 15 , wherein determining the TU partition comprises implicitly determining the desired level independently from an encoder.

18 . The method of claim 15 , wherein:

for a 2N×N PU partition type, a level in the residual tree structure partitions a 2N×2N node into two nodes of 2N×N, and

for a N×2N PU partition type, a level in the residual tree structure partitions a 2N×2N node into two nodes of N×2N.

19 . The method of claim 15 , wherein:

for a 2N×0.5N PU partition type, a level in the residual tree structure partitions a 2N×N node into two nodes of 2N×0.5N, and

for a 0.5N×2N PU partition type, a level in the residual tree structure partitions a N×2N node into two nodes of 0.5N×2N.

20 . The method of claim 15 , wherein:

for a 2N×0.5N PU partition type, a level in the residual tree structure partitions a 2N×2N node into a first node of 2N×0.5N and a second node of 2N×1.5N, and

for a 0.5N×2N PU partition type, a level in the residual tree structure partitions a 2N×2N node into a first node of 0.5N×2N and a second node of 1.5N×0.5N.

21 . An apparatus configured to decode video content comprising:

one or more computer processors; and

a non-transitory computer-readable storage medium comprising instructions, that when executed by the one or more computer processors, control the one or more computer processors to be configured for:

receiving a bitstream of encoded video content;

determining a prediction unit (PU) partition type for a PU for a coding unit (CU) of the video content, wherein the PU is partitionable into a plurality of PU partition types;

determining a residual tree structure based on the PU partition type for partitioning of the CU into transform units (TUs), the residual tree including a binary partition of a node into two TUs;

determining a TU partition for the PU partition type based on the residual tree structure and a desired level of partitioning in the residual tree structure; and

using the TU partition in a transform operation in decoding the video content.