General block partitioning method
A method of partitioning in video coding for JVET, comprising representing a JVET coding tree unit as a root node in a quadtree plus binary tree (QTBT) structure that can have quadtree or binary partitioning of the root node and quadtree or binary trees branching from each of the leaf nodes. The partitioning at any depth can use asymmetric binary partitioning to split a child node represented by a leaf node into two child coding units of unequal size, representing the two child coding units as leaf nodes in a binary tree branching from the parent leaf node and coding the child coding units represented by final leaf nodes of the binary tree with JVET. Disclosed is a generalized method of partitioning a block, either square or rectangular, which leads to more flexible block sizes with possible higher coding efficiency.
1. A method for coding video data using one or more processors, the method comprising:
receiving a bitstream indicating how a coding tree unit was partitioned into coding units according to a partitioning structure that allows root nodes to be split with quadtree partitioning, symmetric binary partitioning, or asymmetric binary partitioning;
parse said bitstream to determine a syntax associated with each coding unit, wherein said syntax conforms to the following:
asymmetric partitioning of the coding tree unit is allowed;
if a node uses quadtree partitioning, a child of the node is partitionable via quadtree partitioning, binary partitioning, or asymmetric partitioning;
if a node uses asymmetric partitioning, no further partitions are allowed;
if a node uses binary partitioning, a child of the node is only partitionable via binary partitioning
decoding the identified coding units using JVET in accordance with said syntax.
2. The method of claim 1 , wherein a partition of a coding tree unit having dimensions 32×32 produces at least one of:
a square block of 32×32,
two rectangular blocks of 32×24 and 32×8 with a vertical split ratio equal to 3:1,
two rectangular blocks of 32×16 and 32×16 with a vertical split ratio equal to 2:2,
two rectangular blocks of 32×8 and 32×24 with a vertical split ratio equal to 1:3,
two rectangular blocks of 24×32 and 8×32 with a horizontal split ratio equal to 3:1,
two rectangular blocks of 16×32 and 16×32 with a horizontal split ratio equal to 2:2, and/or
two rectangular blocks of 8×32 and 24×32 with a horizontal split ratio equal to 1:3.
3. The method of claim 1 , wherein a partition of a coding tree unit having dimensions 32×24 produces at least one of:
a square block of 32×24,
two rectangular blocks of 32×8 and 32×16 with a vertical split ratio equal to 1:2,
two rectangular blocks of 32×16 and 32×8 with a vertical split ratio equal to 2:1,
two rectangular blocks of 24×24 and 8×2 with a horizontal split ratio equal to 3:1,
two rectangular blocks of 16×24 and 16×24 with a horizontal split ratio equal to 2:2, and/or
two rectangular blocks of 8×24 and 24×24 with a horizontal split ratio equal to 1:3.
4. The method of claim 1 , wherein a partition of a coding tree unit having dimensions 24×24 produces at least one of:
a square block of 24×24,
two rectangular blocks of 24×16 and 24×8 with a vertical split ratio equal to 2:1,
two rectangular blocks of 24×8 and 24×16 with a vertical split ratio equal to 1:2,
two rectangular blocks of 16×24 and 8×24 with a horizontal split ratio equal to 2:1, and/or
two rectangular blocks of 8×24 and 16×24 with a horizontal split ratio equal to 1:2.
5. The method of claim 1 , wherein partition is signaled via bin assignments, where a bin represents at least one of the following:
whether a child node is split or not;
whether a child node is split horizontally or vertically;
whether a child node is split symmetrically or asymmetrically;
if asymmetrically, whether a left child unit is larger or a right child unit is larger in a binary partition.
6. The method of claim 1 , wherein at least four bins are assigned to each child node of the partitioning structure to define partitioning within the partitioning structure.