Motion-based priority for constructing candidate lists in video coding

文档序号:1440084 发布日期:2020-02-14 浏览:13次 中文

阅读说明:本技术 用于在视频译码中构造候选列表的基于运动的优先级 (Motion-based priority for constructing candidate lists in video coding ) 是由 张凯 陈建乐 M·卡切维奇 于 2018-06-22 设计创作,主要内容包括:实施例包含用于产生用于帧间预测的候选运动向量列表的技术。举例来说,根据一些实施例,将候选运动向量列表产生为具有基于候选运动向量中的每一者的运动信息的次序。可在所述列表产生时应用所述次序作为在部分地产生所述列表之后或在产生所述列表之后的重分类。(Embodiments include techniques for generating a list of candidate motion vectors for inter-prediction. For example, according to some embodiments, the list of candidate motion vectors is generated to have an order based on motion information for each of the candidate motion vectors. The order may be applied at the time of the list generation as a re-sort after the list is partially generated or after the list is generated.)

1. A method of constructing a candidate list for inter-prediction of a block of a current picture, comprising:

generating a list of candidate motion vectors for inter-predicting a block of a picture, wherein the list is generated to have an order based on motion information of each of the candidate motion vectors;

selecting a candidate motion vector from the list; and

performing inter prediction of the block based on the selected candidate motion vector.

2. The method of claim 1, wherein each of the candidate motion vectors comprises a motion vector, a reference index, and a motion prediction direction.

3. The method of claim 1, wherein performing inter-prediction comprises performing inter-prediction according to a merge mode.

4. The method of claim 1, wherein each of the candidate motion vectors comprises a motion vector, a reference index, and a motion prediction direction, and wherein performing inter prediction comprises performing inter prediction according to an advanced motion vector prediction mode.

5. The method of claim 1, further comprising encoding the block to a video bitstream based on the inter-prediction of the block, wherein encoding the block comprises:

generating a residual based on the block of the picture and a prediction block generated as part of performing the inter-prediction; and

encoding data for performing the inter prediction into the video bitstream, and encoding the residual.

6. The method of claim 1, further comprising decoding the block from an encoded video bitstream based on the inter-prediction of the block, wherein decoding the block comprises:

decoding a residual from the encoded video bitstream; and

decoding the block of the picture based on a prediction block generated as part of performing the inter-prediction and the residual, and wherein selecting the candidate motion vector from the list is based on information from the encoded video bitstream.

7. The method of claim 1, wherein generating the candidate motion list in the order based on motion information for each candidate motion vector comprises:

constructing the entire candidate list according to a first order; and

reordering the candidate list according to the order based on motion information for each candidate motion vector.

8. The method of claim 1, wherein generating the candidate motion list in the order based on motion information for each candidate motion vector comprises:

constructing a first portion of the candidate list according to a first order in which at least one candidate motion vector is included in the candidate list;

reordering the first portion of the candidate list according to the order based on motion information for each of the candidate motion vectors in the first portion of the candidate list;

constructing a remaining portion of the candidate list, including adding at least one additional candidate motion vector in the candidate list.

9. The method of any of claim 9, wherein the reordering is performed such that candidates having a same value in a partial ordering associated with the order of the candidate list are not reordered.

10. The method of claim 1, wherein the order is based on the prediction direction of at least one of the candidate motion vectors.

11. The method of claim 1, wherein the order is based on information associated with a reference picture of at least one of the candidate motion vectors.

12. The method of claim 1, wherein the order is based on information associated with a reference block of at least one of the candidate motion vectors.

13. The method of claim 11, wherein the information associated with the reference block of at least one of the candidate motion vectors comprises a quantization parameter associated with one of the reference picture or the reference block of the at least one of the candidate motion vectors.

14. The method of claim 11, wherein the order is based on a temporal distance between the reference picture and the current picture of at least one of candidate blocks.

15. The method of claim 1, wherein the order is based on information associated with a coding type associated with at least one of the candidate motion vectors.

16. The method of claim 15, wherein the coding type associated with at least one of the candidate motion vectors is the coding type of the reference picture associated with the at least one of the candidate motion vectors.

17. The method of claim 15, wherein the coding type associated with at least one of the candidate motion vectors is the coding type of the reference block associated with the at least one of the candidate motion vectors.

18. The method of claim 1, further comprising calculating a motion vector similarity metric associated with at least one spatial neighboring candidate in the list, wherein the order is based on the calculated motion vector similarity metric.

19. The method of claim 18, wherein the motion vector similarity metric is calculated based on a co-located reference block of a current block to be predicted and a co-located reference block of at least one neighboring block.

20. A device for constructing a candidate list for inter-prediction of a block of a current picture, comprising:

a memory configured to store motion vectors associated with a plurality of blocks of a picture; and

a processor configured to:

generating a list of candidate motion vectors for inter-predicting a block of the picture, wherein the list is generated to have an order based on motion information of each of the candidate motion vectors;

selecting a candidate motion vector from the list; and

performing inter prediction of the block based on the selected candidate motion vector.

21. The device of claim 20, wherein each of the candidate motion vectors comprises a motion vector, a reference index, and a motion prediction direction.

22. The device of claim 20, wherein the processor is configured to perform inter-prediction according to a merge mode.

23. The device of claim 20, wherein each of the candidate motion vectors comprises a motion vector, a reference index, and a motion prediction direction, and wherein the processors are configured to perform inter-prediction according to an advanced motion vector prediction mode.

24. The device of claim 20, wherein the processor is configured to encode the block to a video bitstream based on the inter-prediction of the block, wherein the processor is further configured to:

generating a residual based on the block of the picture and a prediction block generated as part of performing the inter-prediction; and

encoding data indicative of how to perform the inter prediction and indicative of the residual to the video bitstream.

25. The device of claim 20, wherein the processor is configured to decode the block from an encoded video bitstream based on the inter-prediction of the block, wherein the processor is further configured to:

decoding a residual from the encoded video bitstream; and

decoding the block of the picture based on a prediction block generated as part of performing the inter-prediction and the residual, and wherein the processor is configured to select the candidate motion vector from the list based on information from the encoded video bitstream.

26. The device of claim 20, wherein to generate the candidate motion list in the order based on motion information for each candidate motion vector, the processor is configured to:

constructing the entire candidate list according to a first order; and

reordering the candidate list according to the order based on motion information for each candidate motion vector.

27. The device of claim 20, wherein to generate the candidate motion list in the order based on motion information for each candidate motion vector, the processor is configured to:

constructing a first portion of the candidate list according to a first order in which at least one candidate motion vector is included in the candidate list;

reordering the first portion of the candidate list according to the order based on motion information for each of the candidate motion vectors in the first portion of the candidate list; and

constructing a remaining portion of the candidate list, including adding at least one additional candidate motion vector in the candidate list.

28. The device of claim 27, wherein the reordering is performed such that candidates having a same value in a partial ordering associated with the order of the candidate list are not reordered.

29. The device of claim 20, wherein the order is based on the prediction direction of at least one of the candidate motion vectors.

30. The device of claim 20, wherein the order is based on information associated with a reference picture of at least one of the candidate motion vectors.

31. The device of claim 20, wherein the order is based on information associated with a reference block of at least one of the candidate motion vectors.

32. The device of claim 31, wherein the information associated with the reference block of at least one of the candidate motion vectors comprises a quantization parameter associated with one of the reference picture or the reference block of the at least one of the candidate motion vectors.

33. The device of claim 20, wherein the order is based on a temporal distance between the reference picture and the current picture of at least one of candidate blocks.

34. The device of claim 20, wherein the order is based on information associated with a coding type associated with at least one of the candidate motion vectors.

35. The device of claim 34, wherein the coding type associated with at least one of the candidate motion vectors is the coding type of the reference picture associated with the at least one of the candidate motion vectors.

36. The device of claim 34, wherein the coding type associated with at least one of the candidate motion vectors is the coding type of the reference block associated with the at least one of the candidate motion vectors.

37. The device of claim 20, wherein the processor is further configured to calculate a motion vector similarity metric associated with at least one spatial neighbor candidate in the list, wherein the order is based on the calculated motion vector similarity metric.

38. The device of claim 37, wherein the motion vector similarity metric is calculated based on a co-located reference block of a current block to be predicted and a co-located reference block of at least one neighboring block.

Technical Field

The application relates to motion vector prediction in video codecs. More specifically, the present application relates to the construction of a candidate list from which reference blocks are selected for prediction.

Background

Video coding standards include ITU-T H.261, ISO/IEC MPEG-1 Visual, ITU-T H.262, or ISO/IEC MPEG-2 Visual, ITU-T H.263, ISO/IEC MPEG-4 Visual, and ITU-T H.264 (also known as ISO/IEC MPEG-4AVC), including Scalable Video Coding (SVC) and Multiview Video Coding (MVC) extensions thereof. The latest joint draft of MVC is described in "Advanced video coding for general audio visual service" (ITU-T standard h.264) at 3 months 2010. In addition, there is a newly developed video coding standard, namely High Efficiency Video Coding (HEVC) developed by the video coding joint collaboration group (JCT-VC) of the ITU-T Video Coding Experts Group (VCEG) and the ISO/IEC Moving Picture Experts Group (MPEG). The latest HEVC draft is available from http:// phenix. int-evry. fr/jct/doc _ end _ user/documents/12_ Geneva/wg11/JCTVC-L1003-v34. zip. HEVC g.j. sullivan; J. -r.ohm; han, w; wiegand (12 months 2012). "Overview of the High Efficiency Video Coding (HEVC) Standard" (PDF). IEEE Video Technology Circuit and System Association (IEEE)22 (12). 2012-09-14. There is a need for video codecs that support higher resolutions and higher bit depths.

Disclosure of Invention

Drawings

FIG. 1 is a block diagram showing an example video encoding and decoding system that may utilize the techniques described in this disclosure.

FIG. 2 is a block diagram showing an example video encoder that may implement the techniques described in this disclosure.

FIG. 3 is a block diagram showing an example video decoder that may implement the techniques described in this disclosure.

Fig. 4a and 4b show examples of two prediction units and neighboring blocks at positions marked according to the derivation order of the motion vector candidate list.

Fig. 5 is a diagram showing two example prediction units generated with respect to a Temporal Motion Vector Predictor (TMVP) candidate.

Fig. 6 is a diagram showing the relationship between motion vector 606 and TMVP candidate 616.

Fig. 7 shows a block and an adjacent block.

FIG. 8 is a flow diagram depicting an example technique for generating a merge candidate list.

FIG. 9 is a diagram showing an example merge list before and after reordering.

FIG. 10 is a flow chart showing another embodiment of a method of generating a candidate list.

FIG. 11 depicts an example candidate list according to the process of FIG. 10.

Fig. 12 shows an example of reordering the first four candidates according to the inter prediction direction of each candidate.

Fig. 13 shows an example of how MV similarity of neighboring blocks of another block in the current picture is calculated.

Embodiments include techniques for generating a list of candidate motion vectors for inter-prediction. For example, according to some embodiments, the list of candidate motion vectors is generated to have an order based on motion information for each of the candidate motion vectors.

One embodiment includes a method of constructing a candidate list for inter-prediction of a block of a current picture in a video encoder or decoder. The method comprises the following steps: generating a list of candidate motion vectors for inter-predicting a block of a picture, wherein the list is generated to have an order based on motion information of each of the candidate motion vectors; selecting a candidate motion vector from the list; and performing inter prediction of the block based on the selected candidate motion vector.

One embodiment includes a method for constructing a candidate list for inter-prediction of a block of a current picture. The device may comprise a video encoder or a video decoder. The device includes a memory configured to store motion vectors associated with a plurality of blocks of a picture. The device further includes a processor configured to: generating a list of candidate motion vectors for inter-predicting a block of the picture, wherein the list is generated to have an order based on motion information of each of the candidate motion vectors; selecting a candidate motion vector from the list; and performing inter prediction of the block based on the selected candidate motion vector.

One embodiment includes a method for constructing a candidate list for inter-prediction of a block of a current picture. The device may comprise a video encoder or a video decoder. The device includes means for storing motion vectors associated with a plurality of blocks of a picture. The device further includes means for processing video data configured to: generating a list of candidate motion vectors for inter-predicting a block of the picture, wherein the list is generated to have an order based on motion information of each of the candidate motion vectors; selecting a candidate motion vector from the list; and performing inter prediction of the block based on the selected candidate motion vector.

One embodiment includes a non-transitory computer-readable medium having stored thereon instructions that, when executed by a processor, cause a process to: generating a list of candidate motion vectors for inter-predicting a block of a picture, wherein the list is generated to have an order based on motion information of each of the candidate motion vectors; selecting a candidate motion vector from the list; and performing inter prediction of the block based on the selected candidate motion vector.

32页详细技术资料下载
上一篇:一种医用注射器针头装配设备
下一篇:用于鱼眼虚拟实境视频的增强型高阶信号发送

网友询问留言

已有0条留言

还没有人留言评论。精彩留言会获得点赞!

精彩留言,会给你点赞!

技术分类