Coding of a spatial sampling of a two-dimensional information signal using subdivision
Coding schemes for coding a spatially sampled information signal using sub-division and coding schemes for coding a sub-division or a multitree structure are described, wherein representative embodiments relate to picture and/or video coding applications.
1. A decoder for decoding a coded signaling of a multi-tree structure, the multi-tree structure prescribing a spatial multi-tree subdivision of a tree root block according to which the tree root block is recursively multi-partitioned into leaf blocks, the coded signaling comprising an indication of a highest hierarchy level and a sequence of flags associated with nodes of the multi-tree structure unequal to the highest hierarchy level, each flag specifying whether the associated node is an intermediate node or child node and whether an area of the tree root block corresponding to the node with which the respective flag is associated, is multi-partitioned,
wherein the spatial multi-tree subdivision is a quadtree-subdivision, and in accordance with a depth-first traversal order, a zigzag scan is defined among nodes of a same hierarchy level of the quadtree subdivision having a common parent node, the decoder being configured to:
decode the indication of the highest hierarchy level from a data stream, and then sequentially entropy-decode, in the depth-first traversal order, the sequence of flags from the data stream with skipping nodes of the highest hierarchy level and automatically appointing same leaf nodes, and using probability estimation contexts which are the same for flags associated with nodes of the multi-tree structure lying within a same hierarchy level of the multi-tree structure, but different for nodes of the multi-tree structure lying within different hierarchy levels of the multi-tree structure.
2. A method for decoding a coded signaling of a multi-tree structure, the multi-tree structure prescribing a spatial multi-tree subdivision of a tree root block according to which the tree root block is recursively multi-partitioned into leaf blocks, the coded signaling comprising an indication of a highest hierarchy level and a sequence of flags associated with nodes of the multi-tree structure unequal to the highest hierarchy level, each flag specifying whether the associated node is an intermediate node or child node, and whether an area of the tree root block corresponding to the node with which the respective flag is associated, is multi-partitioned,
wherein the spatial multi-tree subdivision is a quadtree-subdivision, and in accordance with a depth-first traversal order, a zigzag scan is defined among nodes of a same hierarchy level of the quadtree subdivision having a common parent node, the method comprising:
decoding the indication of the highest hierarchy level from a data stream, and then sequentially entropy-decoding, in the depth-first traversal order, the sequence of flags from the data stream with skipping nodes of the highest hierarchy level and automatically appointing same leaf nodes, using probability estimation contexts which are the same for the flags associated with nodes of the multi-tree structure lying within a same hierarchy level of the multi-tree structure, but different for nodes of the multi-tree structure lying within different hierarchy levels of the multi-tree structure.
3. An encoder for generating a coded signaling of a multi-tree structure, the multi-tree structure prescribing a spatial multi-tree subdivision of a tree root block according to which the tree root block is recursively multi-partitioned into leaf blocks, the coded signaling comprising an indication of a highest hierarchy level and a sequence of flags associated with nodes of the multi-tree structure unequal to the highest hierarchy level, each flag specifying whether the associated node is an intermediate node or child node, and whether an area of the tree root block corresponding to the node with which the respective flag is associated, is multi-partitioned,
wherein the spatial multi-tree subdivision is a quadtree-subdivision, and in accordance with a depth-first traversal order, a zigzag scan is defined among nodes of a same hierarchy level of the quadtree subdivision having a common parent node, the encoder being configured to:
encode the indication of the highest hierarchy level from a data stream, and then sequentially entropy-encode, in the depth-first traversal order, the sequence of flags from the data stream with skipping nodes of the highest hierarchy level and automatically appointing same leaf nodes, and using probability estimation contexts which are the same for flags associated with nodes of the multi-tree structure lying within a same hierarchy level of the multi-tree structure, but different for nodes of the multi-tree structure lying within different hierarchy levels of the multi-tree structure.
4. A method for generating a coded signaling of a multi-tree structure, the multi-tree structure prescribing a spatial multi-tree subdivision of a tree root block according to which the tree root block is recursively multi-partitioned into leaf blocks, the coded signaling comprising an indication of a highest hierarchy level and a sequence of flags associated with nodes of the multi-tree structure unequal to the highest hierarchy level, each flag specifying whether the associated node is an intermediate node or child node, and whether an area of the tree root block corresponding to the node with which the respective flag is associated, is multi-partitioned,
wherein the spatial multi-tree subdivision is a quadtree-subdivision, and in accordance with a depth-first traversal order, a zigzag scan is defined among nodes of a same hierarchy level of the quadtree subdivision having a common parent node, the method comprising:
encoding the indication of the highest hierarchy level from a data stream, and then sequentially entropy-encoding, in the depth-first traversal order, the sequence of flags from the data stream with skipping nodes of the highest hierarchy level and automatically appointing same leaf nodes, and using probability estimation contexts which are the same for flags associated with nodes of the multi-tree structure lying within a same hierarchy level of the multi-tree structure, but different for nodes of the multi-tree structure lying within different hierarchy levels of the multi-tree structure.