Virtual decoded reference picture marking and reference picture list
An improved system and method for implementing efficient decoding of scalable video bitstreams is provided. A virtual decoded picture buffer is provided for each lower layer of the scalable video bitstream. The virtual decoded picture buffer stores decoded lower layer pictures for reference. The decoded lower layer pictures used for reference are compiled to create a reference picture list for each layer. The reference picture list generated by the virtual decoded picture buffer is used during a direct prediction process instead of a target reference list to correctly decode a current macroblock.
1. A method comprising:
decoding an enhancement layer of a scalable bitstream;
constructing a reference picture list for the decoded enhancement layer;
constructing a reference picture list for each layer lower than the decoded enhancement layer independently from the reference picture list for the decoded enhancement layer, wherein sample values for the pictures of each lower layer are not decoded and the reference picture list representing the pictures of each lower layer is not associated with decoded sample values;
performing a reference picture marking for the decoded enhancement layer; and
performing a reference picture marking for each layer lower than the decoded enhancement layer independently from the reference picture marking for the decoded enhancement layer.
2. A method according to claim 1 , the method further comprising: performing an inter-layer motion prediction for a block of the enhancement layer from motion vectors of a co-located block of the lower layer, wherein the colocated block of the lower layer has a direct mode prediction and the motion vectors of the colocated lower layer are derived based on elements of the reference picture list for the corresponding lower layer.
3. A method according to claim 2 , wherein the direct mode is a spatial direct mode.
4. A method according to claim 2 , wherein the direct mode is a temporal direct mode.
5. A method comprising:
encoding an enhancement layer of a scalable bitstream;
constructing a reference picture list for the encoded enhancement layer;
constructing a reference picture list for each layer lower than the encoded enhancement layer independently from the reference picture list for the encoded enhancement layer, wherein sample values for the pictures of each lower layer are not decoded and the reference picture list representing the pictures of each lower layer is not associated with encoded sample values;
performing a reference picture marking for the encoded enhancement layer; and
performing a reference picture marking for each layer lower than the encoded enhancement layer independently from the reference picture marking for the encoded enhancement layer.
6. A method according to claim 5 , the method further comprising:
performing an inter-layer motion prediction for a block of the enhancement layer from motion vectors of a co-located block of the lower layer, wherein the colocated block of the lower layer has a direct mode prediction and the motion vectors of the colocated lower layer are derived based on elements of the reference picture list for the corresponding lower layer.
7. A method according to claim 6 , wherein the direct mode is a spatial direct mode.
8. A method according to claim 6 , wherein the direct mode is a temporal direct mode.
9. A decoder configured to decode an enhancement layer of a scalable bitstream, comprising:
a reference picture list construction module for
constructing a reference picture list for the decoded enhancement layer;
constructing a reference picture list for each layer lower than the decoded enhancement layer independently from the reference picture list for the decoded enhancement layer, wherein sample values for the pictures of each lower layer are not decoded and the reference picture list representing the pictures of each lower layer is not associated with decoded sample values;
performing a reference picture marking for the decoded enhancement layer; and
performing a reference picture marking for each layer lower than the decoded enhancement layer independently from the reference picture marking for the decoded enhancement layer.
10. A decoder according to claim 9 , the decoder further comprises:
a module for performing an inter-layer motion prediction for a block of the enhancement layer from motion vectors of a co-located block of the lower layer, wherein the co-located block of the lower layer has a direct mode prediction and the motion vectors of the co-located lower layer are derived based on elements of the reference picture list for the corresponding lower layer.
11. A decoder according to claim 10 , wherein the direct mode is a spatial direct mode.
12. A decoder according to claim 10 , wherein the direct mode is a temporal direct mode.
13. An encoder configured to encode an enhancement layer of a scalable bitstream, comprising:
a reference picture list construction module for
constructing a reference picture list for the encoded enhancement layer;
constructing a reference picture list for each layer lower than the encoded enhancement layer independently from the reference picture list for the encoded enhancement layer, wherein sample values for the pictures of each lower layer are not decoded and the reference picture list representing the pictures of each lower layer is not associated with encoded sample values;
performing a reference picture marking for the encoded enhancement layer; and
performing a reference picture marking for each layer lower than the encoded enhancement layer independently from the reference picture marking for the encoded enhancement layer.
14. An encoder according to claim 13 , the encoder further comprises:
a module for performing an inter-layer motion prediction for a block of the enhancement layer from motion vectors of a co-located block of the lower layer, wherein the co-located block of the lower layer has a direct mode prediction and the motion vectors of the co-located lower layer are derived based on elements of the reference picture list for the corresponding lower layer.
15. An encoder according to claim 14 , wherein the direct mode is a spatial direct mode.
16. An encoder according to claim 14 , wherein the direct mode is a temporal direct mode.
17. A computer program embodied in a non-transitory computer-readable medium, the computer program executable by one or more processors, the computer program comprising instructions to cause one or more processors to perform a method comprising:
decoding an enhancement layer of a scalable bitstream;
constructing a reference picture list for the decoded enhancement layer;
constructing a reference picture list for each layer lower than the decoded enhancement layer independently from the reference picture list for the decoded enhancement layer, wherein sample values for the pictures of each lower layer are not decoded and the reference picture list representing the pictures of each lower layer is not associated with decoded sample values;
performing a reference picture marking for the decoded enhancement layer; and
performing a reference picture marking for each layer lower than the decoded enhancement layer independently from the reference picture marking for the decoded enhancement layer.
18. A computer program product according to claim 17 , wherein the method further comprises:
performing an inter-layer motion prediction for a block of the enhancement layer from motion vectors of a co-located block of the lower layer, wherein the co-located block of the lower layer has a direct mode prediction and the motion vectors of the co-located lower layer are derived based on elements of the reference picture list for the corresponding lower layer.
19. A computer program product according to claim 18 , wherein the direct mode is a spatial direct mode.
20. A computer program product according to claim 18 , wherein the direct mode is a temporal direct mode.