Adaptive Implicit Transform Setting

• US 20190320204A1
• Filed: 04/11/2019
• Published: 10/17/2019
• Est. Priority Date: 04/13/2018
• Status: Active Grant

• Current Assignee: HFI Innovation, Inc. (MediaTek, Inc.)
• Original Assignee: MediaTek, Inc.
• Inventors: Chiang, Man-Shu, Hsu, Chih-Wei
• Granted Patent: US 10,999,604 B2
• Time in Patent Office: Days
• CPC Class Codes: H04N 19/12 Selection from among a plurality of transforms or standards, e.g. selection between discrete cosine transform [DCT] and sub-band transform or selection between H.263 and H.264
• CPC Class Codes: H04N 19/157 Assigned coding mode, i.e. the coding mode being predefined or preselected to be further used for selection of another element or parameter
• CPC Class Codes: H04N 19/159 Prediction type, e.g. intra-frame, inter-frame or bidirectional frame prediction
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/52 by predictive encoding
• CPC Class Codes: H04N 19/61 in combination with predictive coding

Abstract

A video coder that implicitly signals or decides transform settings based on intra prediction settings is provided. The video coder receives input data associated with a current block. The video coder determines a current intra prediction mode for the current block and maps the current intra prediction mode to a target transform mode. In one example, the video coder encodes the current block by performing a transform operation according to the target transform mode on the prediction residuals of the current block generated according to the current intra prediction mode. In another example, the video coder decodes the current block by performing an inverse transform operation according to the target transform mode on the coded prediction residuals of the current block, in which the coded prediction residuals is derived from the bitstream and is generated according to the current intra prediction mode.

Claims

1. An encoding method comprising:

receiving a block of pixels of a video picture to be coded as a current block;

performing an intra prediction operation for the current block according to an intra prediction setting and generating prediction residuals of the current block, the intra prediction setting selecting a current intra prediction mode from a plurality of intra prediction modes for the current block;

mapping the current intra prediction mode to a target transform mode; and

encoding the current block in a bitstream by performing a transform operation on the prediction residuals of the current block according to the target transform mode.

2. The method of claim 1, wherein a collection of intra prediction modes are assigned with a particular transform mode, wherein the target transform mode is the particular transform mode when the current intra prediction mode is in the collection of intra prediction modes.

3. The method of claim 2, wherein the collection of intra prediction modes is based on a set of intra prediction modes in a bottom left-direction, a left direction, a top-left direction, a top direction, and a top-right direction.

4. The method of claim 2, wherein the collection of intra prediction modes is defined by a fixed expression, wherein different blocks with different block sizes have different fixed expressions for defining different collections of intra prediction modes.

5. The method of claim 2, wherein the collection of intra prediction modes is defined by a fixed expression, wherein different fixed expressions are used to assign intra prediction modes in different numerical ranges with transform modes.

6. The method of claim 1, wherein intra prediction modes in a most probable mode (MPM) list of the current block are assigned with a particular transform mode, wherein intra prediction modes in the MPM list of the current block are defined based on intra prediction modes of blocks neighboring the current block.

7. The method of claim 1, wherein an intra prediction mode indicative of uniform block texture is mapped to a particular transform mode.

8. The method of claim 1, wherein the target transform mode is not encoded in the bitstream as a syntax element, and wherein the intra prediction setting is encoded in the bitstream as a syntax element.

9. The method of claim 1, wherein the target transform mode specifies a horizontal transform type and a vertical transform type.

10. An electronic apparatus comprising:

an encoder circuit capable of;

receiving a block of pixels of a video picture to be coded as a current block;

performing an intra prediction operation for the current block according to an intra prediction setting and generating prediction residuals of the current block, the intra prediction setting selecting a current intra prediction mode from a plurality of intra prediction modes for the current block;

mapping the current intra prediction mode to a target transform mode; and

encoding the current block in a bitstream by performing a transform operation on the prediction residuals of the current block according to the target transform mode.

11. A decoding method, comprising:

receiving a bitstream comprising coded prediction residuals of a current block of a video picture;

determining a current intra prediction mode from a plurality of intra prediction modes for the current block according to an intra prediction setting;

mapping the current intra prediction mode to a target transform mode; and

decoding the current block by performing an inverse transform operation on the coded prediction residuals of the current block according to the target transform mode to generate prediction residuals of the current block and by reconstructing the current block based on the prediction residuals and a set of prediction pixels of the current block generated by performing an intra prediction operation for the current block according to the current intra prediction mode.

12. The method of claim 11, wherein a collection of intra prediction modes are assigned with a particular transform mode, wherein the target transform mode is the particular transform mode when the current intra prediction mode is in the collection of intra prediction modes.

13. The method of claim 12, wherein the collection of intra prediction modes is based on a set of intra prediction modes in a bottom left-direction, a left direction, a top-left direction, a top direction, and a top-right direction.

14. The method of claim 12, wherein the collection of intra prediction modes is defined by a fixed expression, wherein different blocks with different block sizes have different fixed expressions for defining different collections of intra prediction modes.

15. The method of claim 12, wherein the collection of intra prediction modes is defined by a fixed expression, wherein different fixed expressions are used to assign intra prediction modes in different numerical ranges with transform modes.

16. The method of claim 11, wherein intra prediction modes in a most probable mode (MPM) list of the current block are assigned with a particular transform mode, wherein intra prediction modes in the MPM list of the current block are defined based on intra prediction modes of blocks neighboring the current block.

17. The method of claim 11, wherein an intra prediction mode indicative of uniform block texture is mapped to a particular transform mode.

18. The method of claim 11, wherein the target transform mode is not encoded in the bitstream as a syntax element.

19. The method of claim 11, wherein the target transform mode specifies a horizontal transform type and a vertical transform type.

20. An electronic apparatus comprising:

a decoder circuit capable of;

receiving a bitstream comprising coded prediction residuals of a current block of a video picture;

determining a current intra prediction mode from a plurality of intra prediction modes for the current block according to an intra prediction setting;

mapping the current intra prediction mode to a target transform mode; and

decoding the current block by performing an inverse transform operation on the coded prediction residuals of the current block according to the target transform mode to generate prediction residuals of the current block and by reconstructing the current block based on the prediction residuals and a set of prediction pixels of the current block generated by performing an intra prediction operation for the current block according to the current intra prediction mode.