Parallel decoder with inter-prediction of video pictures

1. A parallel decoder for decoding compressed video picture data including inter-coded picture item data with motion vector data, the decoder comprising:

a temporary storage for storing a plurality of structures of picture data to be decoded including structures to organize inter-prediction items;

at least one decoding module for decoding the stored picture data, wherein the at least one decoding module includes at least one inter-prediction module that uses inter-prediction item data to decode an inter-coded picture item by referring to already decoded reference picture item data; and

a control module that controls the structure of inter-prediction item data in the temporary storage as a function of positions in a decoding order of corresponding reference picture items, and prioritizes a decoding order of stored inter-prediction item data by the at least one inter-prediction module as a function of the decoding order of the reference picture item data,

wherein the structures to organize inter-prediction items to be decoded comprise respective queues of inter-prediction item data entities,

wherein each inter-prediction item data entity contains an identification of the relevant reference picture item data that the at least one inter-prediction module uses to decode the respective inter-coded picture item,

wherein the structures of picture item data to be decoded comprise respective trees, each tree having linked nodes that contain inter-prediction item data entities and references to any child nodes,

wherein the inter-prediction module extracts the inter-prediction item data entities by tree traversal and uses the inter-prediction item data entities in decoding.

2. The parallel decoder of claim 1 , wherein the identification of the relevant reference picture item data includes motion vector data, an index to the relevant reference picture and a block size.

3. The parallel decoder of claim 1 , wherein the queues are prioritized by the positions of the reference picture item data in decoding order.

4. The parallel decoder of claim 3 , further comprising a syntax parser, in communication with the control module, that distributes the inter-prediction item entities to the queues according to the position in decoding order of the relevant reference picture item.

5. The parallel decoder of claim 4 , wherein the syntax parser uses the position of the bottom right pixel as the reference picture item position.

6. The parallel decoder of claim 1 , wherein each queue contains picture inter-prediction item data for a respective region of the reference picture according to the decoding order.

7. The parallel decoder of claim 1 , wherein each queue contains picture inter-prediction item data for a respective region of the reference picture according to raster scan order.

8. The parallel decoder of claim 1 , wherein the picture data relates to one of slices and tiles of the picture, and the queues contain inter-prediction item data for respective slices or tiles of the picture.

9. The parallel decoder of claim 1 , wherein the decoding module comprises parallel decoding outputs, and the inter-prediction item data for respective parallel decoding outputs are interleaved in the same queues.

10. The parallel decoder of claim 1 , wherein the identification of the relevant reference picture item data includes motion vector data, an index to the relevant reference picture and a block size.

11. The parallel decoder of claim 1 , wherein the control module prioritizes the structure of inter-prediction item data in the temporary storage and the decoding order of stored inter-prediction item data by the order of traversal of the nodes.

12. The parallel decoder of claim 1 , wherein the control module inserts a first inter-prediction item data entity in a tree as a root node, and subsequent inter-prediction item data entities as child nodes in order of prioritization for decoding.