Method and apparatus of motion vector prediction for scalable video coding
View Patent ↗Inter-layer motion mapping information may be used to enable temporal motion vector prediction (TMVP) of an enhancement layer of a bitstream. For example, a reference picture and a motion vector (MV) of an inter-layer video block may be determined. The reference picture may be determined based on a collocated base layer video block. For example, the reference picture may be a collocated inter-layer reference picture of the reference picture of the collocated base layer video block. The MV may be determined based on a MV of the collocated base layer video block. For example, the MV may be determined by determining the MV of the collocated base layer video block and scaling the MV of the collocated base layer video block according to a spatial ratio between the base layer and the enhancement layer. TMVP may be performed on the enhancement layer picture using the MV of the inter-layer video block.
1. A method comprising:
receiving, via a video decoder, a bitstream, the bitstream comprising a base layer and an enhancement layer;
determining, based on an indication in the bitstream, whether to use an inter-layer reference picture or a temporal enhancement layer picture as a collocated picture (ColPic) for temporal motion vector prediction (TMVP) of an enhancement layer picture;
upon determining to use the inter-layer reference picture as the collocated picture (ColPic) for TMVP of the enhancement layer picture, adding, via the video decoder, the inter-layer reference picture into a reference picture list for the enhancement layer picture, wherein the inter-layer reference picture comprises texture determined from texture of a base layer picture, motion vectors determined from motion vectors of the base layer picture, and reference picture indices determined from reference picture indices of the base layer picture; and
decoding, via the video decoder, the enhancement layer picture using the inter-layer reference picture as the collocated picture (ColPic) for TMVP of the enhancement layer picture.
2. The method of claim 1 , further comprising:
upsampling, via the video decoder, the texture of the base layer picture to determine the texture of the inter-layer reference picture; and
scaling, via the video decoder, the motion vectors of the base layer picture to determine the motion vectors of the inter-layer reference picture.
3. The method of claim 1 , wherein the inter-layer reference picture is at a different time instance as the enhancement layer picture.
4. The method of claim 1 , wherein the inter-layer reference picture is at the same time instance as the enhancement layer picture.
5. The method of claim 1 , further comprising:
marking an area of the inter-layer reference picture to indicate that the area is not available for use in TMVP.
6. The method of claim 1 , further comprising:
determining, via the video decoder, compressed motion vectors of the base layer picture; and
determining, via the video decoder, the motion vectors of the inter-layer reference picture based on the compressed motion vectors the base layer picture.
7. The method of claim 6 , wherein determining the motion vectors of the inter-layer reference picture comprises scaling the compressed motion vectors of the base layer picture according to a spatial ratio between the base layer and the enhancement layer and determining the motion vectors of the inter-layer reference picture based on the scaled motion vectors.
8. The method of claim 1 , further comprising: determining, via the video decoder, a reference picture list for the inter-layer reference picture by copying a reference picture list of the base layer picture.
9. The method of claim 8 , further comprising:
determining, via the video decoder, a picture order count (POC) of each reference picture in the reference picture list of the inter-layer reference picture by copying a POC of a corresponding reference picture in the reference picture list of the base layer picture.
10. The method of claim 5 , wherein marking the area of the inter-layer reference picture comprises setting a reference picture index corresponding to the area to a value of −1.
11. The method of claim 5 , wherein marking the area of the inter-layer reference picture is performed in response to determining that a corresponding area in the base layer picture is intra coded.
12. A decoder comprising:
a processor configured to:
receive a bitstream, the bitstream comprising a base layer and an enhancement layer;
determine, based on an indication in the bitstream, whether to use an inter-layer reference picture or a temporal enhancement layer picture as a collocated picture (ColPic) for temporal motion vector prediction (TMVP) of an enhancement layer picture;
upon determining to use the inter-layer reference picture as the collocated picture (ColPic) for TMVP of the enhancement layer picture, adding the inter-layer reference picture into a reference picture list for the enhancement layer picture, wherein the inter-layer reference picture comprises texture determined from texture of a base layer picture, motion vectors determined from motion vectors of the base layer picture, and reference picture indices determined from reference picture indices of the base layer picture; and
decode the enhancement layer picture using TMVP using the inter-layer reference picture as the collocated picture (ColPic) for TMVP of the enhancement layer picture.
13. The decoder of claim 12 , wherein the texture is determined by upsampling the texture of the base layer picture, and the motion vectors are determined by scaling the motion vectors of the base layer picture.
14. The decoder of claim 12 , wherein the inter-layer reference picture is at a different time instance as the enhancement layer picture.
15. The decoder of claim 12 , wherein the inter-layer picture is at the same time instance as the enhancement layer picture.
16. The decoder of claim 12 , wherein an area of the inter-layer reference picture is marked to indicate that the area is not available for use in TMVP.
17. The decoder of claim 12 , wherein the processor is further configured to:
determine compressed motion vectors of the base layer picture; and
determine the motion vectors of the inter-layer reference picture based on the compressed motion vectors of the base layer picture.
18. The decoder of claim 17 , wherein the processor configured to determine the motion vectors of the inter-layer reference picture comprises the processor configured to scale the compressed motion vectors of the base layer picture according to a spatial ratio between the base layer and the enhancement layer and determine the motion vectors of the inter-layer reference picture based on the scaled motion vectors.
19. The decoder of claim 12 , wherein the processor is further configured to:
determine a reference picture for the inter-layer reference picture by copying a reference picture list of the base layer picture.
20. The decoder of claim 19 , wherein the processor is further configured to determine a picture order count (POC) of each reference picture in the reference picture list of the inter-layer reference picture by copying a POC of a corresponding reference picture in the reference picture list of the base layer picture.
21. The decoder of claim 16 , wherein a reference picture index corresponding to the area is set to a value of −1 to indicate that the area is not available for use in TMVP.
22. The decoder of claim 16 , wherein the area of the inter-layer reference picture is marked in response to determining that a corresponding area in the base layer picture is intra coded.
23. The method of claim 1 , wherein decoding comprises predicting a motion vector of the enhancement layer picture using a motion vector of the inter-layer reference picture.
24. The decoder of claim 12 , wherein the processor is configured to decode the enhancement layer picture using TMVP by predicting a motion vector of the enhancement layer picture using a motion vector of the inter-layer reference picture.