Color residual prediction for video coding

Patent metadata

Field	Value
Publication number	US-10397607-B2
Application number	US-201414528794-A
Country	US
Kind code	B2
Filing date	Oct 30, 2014
Priority date	Nov 1, 2013
Publication date	Aug 27, 2019
Grant date	Aug 27, 2019

Abstract

A method of decoding video data includes decoding a first block of video data to produce a block of reconstructed luma residual values and a block of predicted chroma residual values, wherein the block of video data has one of a 4:2:0 or a 4:2:2 chroma sub-sampling format. The method further includes performing a color residual prediction process to reconstruct a block of chroma residual values for the first block of video data using a subset of the reconstructed luma residual values as luma predictors for the block of predicted chroma residual values.

First claim

What is claimed is:

1. A method of decoding encoded video data, the method comprising:
based on each respective transform unit size of respective transform unit sizes of a first set of blocks of a picture of the encoded video data being greater than 4×4, decoding each respective block of the first set of blocks of the picture to produce a respective block of reconstructed luma residual values and a respective block of predicted chroma residual values, wherein each respective block of the first set of blocks of the picture has one of a 4:2:0 or a 4:2:2 chroma sub-sampling format;
based on each of the respective transform unit sizes of the first set of blocks being greater than 4×4, performing a color residual prediction process to reconstruct a respective block of chroma residual values for each respective block of the first set of blocks of the picture using a subset of the reconstructed luma residual values for the respective block as luma predictors for the respective block of predicted chroma residual values;
based on each respective transform unit size of respective transform unit sizes of a second set of blocks of the picture being less than or equal to 4×4, decoding each respective block of the second set of blocks of the picture without performing color residual prediction;
decoding a luma coded block flag for a first block of the second set of blocks of the picture of the encoded video data, wherein the first block of the second set of blocks has one of a 4:2:0 or a 4:2:2 chroma sub-sampling format and
decoding the first block of the second set of blocks of the picture without performing color residual prediction based on the luma coded block flag indicating that the first block of the picture has no non-zero luma transform coefficients.

2. The method of claim 1 , wherein performing the color residual prediction process further comprises scaling the luma predictors to form scaled luma predictors, and reconstructing the chroma residual values based on the scaled luma predictors.

3. The method of claim 2 , wherein performing the color residual prediction process further comprises adding the scaled luma predictors to the predicted chroma residual values.

4. The method of claim 2 , wherein a first block of the first set of blocks of the picture has the 4:2:2 chroma sub-sampling format,
wherein the scaled luma predictors corresponding to the first block comprise scaled reconstructed luma residual values having a same position as the predicted chroma residual values corresponding to the first block, and
wherein performing the color residual prediction process with respect to the first block further comprises adding a corresponding scaled luma predictor to each corresponding predicted chroma residual value.

5. The method of claim 2 , wherein a first block of the first set of blocks of the picture has the 4:2:2 chroma sub-sampling format,
wherein the subset of the reconstructed luma residual values corresponding to the first block comprises reconstructed luma residual values having a same position as the predicted chroma residual values and at least one other reconstructed luma residual value having a position immediately adjacent the predicted chroma residual values, and
wherein performing the color residual prediction process with respect to the first block further comprises:
for each position of a respective predicted chroma residual value, averaging a scaled reconstructed luma value having the same position as a corresponding predicted chroma residual value and the at least one other scaled reconstructed luma residual value having the position immediately adjacent the predicted chroma residual value to form an averaged luma predictor; and
adding a corresponding averaged luma predictor to each corresponding predicted chroma residual value.

6. The method of claim 5 , wherein the averaging is a simple averaging.

7. The method of claim 5 , wherein the average is a weighted averaging.

8. The method of claim 5 , wherein a first block of the first set of blocks of the picture has the 4:2:2 chroma sub-sampling format, and
wherein the at least one other scaled reconstructed luma residual value for the first block comprises scaled reconstructed luma residual values having a position immediately to the right of each of the predicted chroma residual values for the first block.

9. The method of claim 5 , wherein a first block of the first set of blocks of the picture has the 4:2:2 chroma sub-sampling format, and
wherein the at least one other scaled reconstructed luma residual value for the first block comprises eight scaled reconstructed luma residual values having positions immediately adjacent each of the predicted chroma residual values for the first block.

10. The method of claim 2 , wherein a first block of the first set of blocks of the picture has the 4:2:0 chroma sub-sampling format,
wherein the scaled luma predictors for the first block comprise at least one scaled reconstructed luma residual value having a position adjacent each of the predicted chroma residual values, and
wherein performing the color residual prediction process with respect to the first block further comprises adding a corresponding scaled luma predictor to each corresponding predicted chroma residual value.

11. The method of claim 10 , wherein the scaled luma predictors for each predicted chroma residual value comprise a single scaled reconstructed luma value adjacent each chroma residual value.

12. The method of claim 10 , wherein the scaled luma predictors for each predicted chroma residual value comprise two or more scaled reconstructed luma value adjacent each chroma residual value, the method further comprising:
averaging the two or more scaled reconstructed luma values.

13. An apparatus configured to decode encoded video data, the apparatus comprising:
a video data memory configured to store encoded video data representative of a picture; and
a video decoder in communication with the video data memory, the video decoder being configured to:
based on each respective transform unit size of respective transform unit sizes of a first set of blocks of a picture of the encoded video data being greater than 4×4, decode each respective block of the first set of blocks of the picture to produce a respective block of reconstructed luma residual values and a respective block of predicted chroma residual values, wherein each respective block of the first set of blocks of the picture has one of a 4:2:0 or a 4:2:2 chroma sub-sampling format;
based on each of the respective transform unit sizes of the first set of blocks being greater than 4×4, perform a color residual prediction process to reconstruct a respective block of chroma residual values for each respective block of the first set of blocks of the picture using a subset of the reconstructed luma residual values for the respective block as luma predictors for the respective block of predicted chroma residual values;
based on each respective transform unit size of respective transform unit sizes of a second set of blocks of the picture being less than or equal to 4×4, decode each respective block of the second set of blocks of the picture without performing color residual prediction;
decode a luma coded block flag for a first block of the second set of blocks of the picture of the encoded video data, wherein the first block of the second set of blocks has one of a 4:2:0 or a 4:2:2 chroma sub-sampling format and
decode the first block of the second set of blocks of the picture without performing color residual prediction based on the luma coded block flag indicating that the first block of the picture has no non-zero luma

Assignees

• Qualcomm Inc

Inventors

• Kim Woo-Shik
• Pu Wei
• Chen Jianle
• Sole Rojals Joel
• Guo Liwei
• Karczewicz Marta

Classifications

• H04N19/186

  the unit being a colour or a chrominance component · CPC title

• H04N19/50

  using predictive coding (H04N19/61 takes precedence) · CPC title

• H04N19/61

  in combination with predictive coding · CPC title

• H04N19/105 Primary

  Selection of the reference unit for prediction within a chosen coding or prediction mode, e.g. adaptive choice of position and number of pixels used for prediction · CPC title

• H04N19/103 Primary

  Selection of coding mode or of prediction mode · CPC title