LUMA-BASED CHROMA INTRA-PREDICTION METHOD THAT UTILIZES DOWN-SAMPLED LUMA SAMPLES DERIVED FROM WEIGHTING AND ASSOCIATED LUMA-BASED CHROMA INTRA-PREDICTION APPARATUS

• US 20200137394A1
• Filed: 10/21/2019
• Published: 04/30/2020
• Est. Priority Date: 10/26/2018
• Status: Active Grant

• Current Assignee: MediaTek, Inc.
• Original Assignee: MediaTek, Inc.
• Inventors: Shih, Cheng-Hsuan, Lin, Jian-Liang
• Granted Patent: US 10,939,118 B2
• Time in Patent Office: Days
• CPC Class Codes: H04N 19/115 Selection of the code volume for a coding unit prior to coding
• CPC Class Codes: H04N 19/117 Filters, e.g. for pre-processing or post-processing sub-band filter banks H04N19/635
• CPC Class Codes: H04N 19/136 Incoming video signal characteristics or properties
• CPC Class Codes: H04N 19/149 by estimating the code amount by means of a model, e.g. mathematical model or statistical model
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/186 the unit being a colour or a chrominance component
• CPC Class Codes: H04N 19/59 involving spatial sub-sampling or interpolation, e.g. alteration of picture size or resolution
• CPC Class Codes: H04N 19/593 involving spatial prediction techniques
• CPC Class Codes: H04N 19/80 Details of filtering operations specially adapted for video compression, e.g. for pixel interpolation H04N19/635, H04N19/86 take precedence

Abstract

A luma-based chroma intra-prediction method includes: applying, by a filter circuit with a first weighting table, weighting to reconstructed luma samples to generate a first down-sampled luma sample, wherein the reconstructed luma samples are external to a luma block; computing parameters of a linear model, wherein a pair of the first down-sampled luma sample and a reconstructed chroma sample that is external to a chroma block is involved in computing the parameters of the linear model; and determining a predicted value of a chroma sample included in the chroma block according to the linear model and a second down-sampled luma sample that is derived from the luma block.

Claims

1. A luma-based chroma intra-prediction method comprising:

applying, by a filter circuit with a first weighting table, weighting to reconstructed luma samples to generate a first down-sampled luma sample, wherein the reconstructed luma samples are external to a luma block;

computing parameters of a linear model, wherein a pair of the first down-sampled luma sample and a reconstructed chroma sample that is external to a chroma block is involved in computing the parameters of the linear model; and

determining a predicted value of a chroma sample included in the chroma block according to the linear model and a second down-sampled luma sample that is derived from the luma block.

2. The luma-based chroma intra-prediction method of claim 1, wherein a 2×

2 block that consists of the reconstructed luma samples external to the luma block includes a top left reconstructed luma sample, a top right reconstructed luma sample, a bottom left reconstructed luma sample, and a bottom right reconstructed luma sample;

the first weighting table includes a first weighting value, a second weighting value, a third weighting value, and a fourth weighting value; and

applying weighting to the reconstructed luma samples comprises;

generating the first down-sampled luma sample for the 2×

2 block by;

applying the first weighting value to the top left reconstructed luma sample;

applying the second weighting value to the top right reconstructed luma;

applying the third weighting value to the bottom left reconstructed luma sample; and

applying the fourth weighting value to the bottom right reconstructed luma sample.

3. The luma-based chroma intra-prediction method of claim 1, further comprising:

applying, by the filter circuit with the first weighting table, weighting to reconstructed luma samples included in the luma block to generate the second down-sampled luma sample.

4. The luma-based chroma intra-prediction method of claim 3, wherein a 2×

2 block that consists of the reconstructed luma samples included in the luma block includes a top left reconstructed luma sample, a top right reconstructed luma sample, a bottom left reconstructed luma sample, and a bottom right reconstructed luma sample;

the first weighting table includes a first weighting value, a second weighting value, a third weighting value, and a fourth weighting value; and

applying weighting to the reconstructed luma samples included in the luma block comprises;

generating the second down-sampled luma sample for the 2×

2 block by;

applying the first weighting value to the top left reconstructed luma sample;

applying the second weighting value to the top right reconstructed luma;

applying the third weighting value to the bottom left reconstructed luma sample; and

applying the fourth weighting value to the bottom right reconstructed luma sample.

5. The luma-based chroma intra-prediction method of claim 1, further comprising:

applying, by the filter circuit with a second weighting table, to reconstructed luma samples included in the luma block to generate the second down-sampled luma sample, wherein the second weighting table is different from the first weighting table.

6. The luma-based chroma intra-prediction method of claim 5, wherein a 2×

2 block that consists of the reconstructed luma samples included in the luma block includes a top left reconstructed luma sample, a top right reconstructed luma sample, a bottom left reconstructed luma sample, and a bottom right reconstructed luma sample;

the second weighting table includes a first weighting value, a second weighting value, a third weighting value, and a fourth weighting value; and

applying weighting to the reconstructed luma samples included in the luma block comprises;

generating the second down-sampled luma sample for the 2×

2 block by;

applying the first weighting value to the top left reconstructed luma sample;

applying the second weighting value to the top right reconstructed luma;

applying the third weighting value to the bottom left reconstructed luma sample; and

applying the fourth weighting value to the bottom right reconstructed luma sample.

7. The luma-based chroma intra-prediction method of claim 1, wherein a top boundary of the luma block is aligned with a top boundary of a coding tree unit (CTU), the reconstructed luma samples are located above the top boundary of the CTU, the first weighting table includes weighting values, and applying weighting to the reconstructed luma samples comprises:

applying the weighting values to the reconstructed luma samples, respectively, wherein the weighting values include zero values.

8. The luma-based chroma intra-prediction method of claim 7, wherein a 2×

2 block that consists of the reconstructed luma samples is located above the top boundary of the CTU;

the weighting values include a first weighting value, a second weighting value, a third weighting value, and a fourth weighting value; and

applying the weighting values to the reconstructed luma samples comprises;

applying the first weighting value to a top left reconstructed luma sample in the 2×

2 block;

applying the second weighting value to a top right reconstructed luma sample in the 2×

2 block;

applying the third weighting value to a bottom left reconstructed luma sample in the 2×

2 block; and

applying the fourth weighting value to a bottom right reconstructed luma sample in the 2×

2 block;

wherein the first weighting value, the second weighting value, the third weighting value, and the fourth weighting value have only one non-zero value.

9. The luma-based chroma intra-prediction method of claim 8, wherein the first weighting value, the second weighting value, the third weighting value, and the fourth weighting value are set by {0, 0, 0, 1} or {0, 0, 1, 0}.

10. The luma-based chroma intra-prediction method of claim 7, wherein a 2×

2 block that consists of the reconstructed luma samples is located above the top boundary of the CTU;

the weighting values include a first weighting value, a second weighting value, a third weighting value, and a fourth weighting value; and

applying the weighting values to the reconstructed luma samples comprises;

applying the first weighting value to a top left reconstructed luma sample in the 2×

2 block;

applying the second weighting value to a top right reconstructed luma sample in the 2×

2 block;

applying the third weighting value to a bottom left reconstructed luma sample in the 2×

2 block; and

applying the fourth weighting value to a bottom right reconstructed luma sample in the 2×

2 block;

wherein the first weighting value, the second weighting value, the third weighting value, and the fourth weighting value have only two non-zero values.

11. The luma-based chroma intra-prediction method of claim 10, wherein the first weighting value, the second weighting value, the third weighting value, and the fourth weighting value are set by {0, 0, 1,1}.

12. A luma-based chroma intra-prediction method comprising:

applying, by a filter circuit with a weighting table, weighting to reconstructed luma samples included in a luma block to generate a down-sampled luma sample; and

determining a predicted value of a chroma sample included in a chroma block according to a linear model and the down-sampled luma sample.

13. The luma-based chroma intra-prediction method of claim 12, wherein a 2×

2 block that consists of the reconstructed luma samples includes a top left reconstructed luma sample, a top right reconstructed luma sample, a bottom left reconstructed luma sample, and a bottom right reconstructed luma sample;

the weighting table includes a first weighting value, a second weighting value, a third weighting value, and a fourth weighting value; and

applying weighting to the reconstructed luma samples included in the luma block comprises;

generating the down-sampled luma sample for the 2×

2 block by;

applying the first weighting value to the top left reconstructed luma sample;

applying the second weighting value to the top right reconstructed luma;

applying the third weighting value to the bottom left reconstructed luma sample; and

applying the fourth weighting value to the bottom right reconstructed luma sample.

14. A luma-based chroma intra-prediction method comprising:

computing parameters of a linear model by utilizing only N down-sampled luma samples that are derived from reconstructed luma samples external to a 2N×

2N luma block and only N reconstructed chroma samples that are external to an N×

N chroma block, where N is a positive integer;

determining predicted values of chroma samples included in the N×

N chroma block according to the linear model and down-sampled luma samples that are derived from the 2N×

2N luma block.

15. The luma-based chroma intra-prediction method of claim 14, wherein all of the N reconstructed chroma samples are located directly above a top boundary of the N×

N chroma block, and all of the reconstructed luma samples are located directly above a top boundary of the 2N×

2N luma block.

16. The luma-based chroma intra-prediction method of claim 14, wherein all of the N reconstructed chroma samples are located directly left to a left boundary of the N×

N chroma block, and all of the reconstructed luma samples are located directly left to a left boundary of the 2N×

2N luma block.

17. A luma-based chroma intra-prediction apparatus comprising:

a filter circuit, arranged to employ a first weighting table for applying weighting to reconstructed luma samples to generate a first down-sampled luma sample, wherein the reconstructed luma samples are external to a luma block; and

a parameter derivation circuit, arranged to compute parameters of a linear model, wherein a pair of the first down-sampled luma sample and a reconstructed chroma sample that is external to a chroma block is involved in computing the parameters of the linear model; and

a chroma intra-prediction circuit, arranged to determine a predicted value of a chroma sample included in the chroma block according to the linear model and a second down-sampled luma sample that is derived from the luma block.

18. A luma-based chroma intra-prediction apparatus comprising:

a filter circuit, arranged to employ a weighting table for applying weighting to reconstructed luma samples included in a luma block to generate a down-sampled luma sample; and

a chroma intra-prediction circuit, arranged to determine a predicted value of a chroma sample included in a chroma block according to a linear model and the down-sampled luma sample.