TRANSFORM UNIT PARTITIONING FOR CHROMA COMPONENTS IN VIDEO CODING

• US 20130107970A1
• Filed: 10/31/2012
• Published: 05/02/2013
• Est. Priority Date: 11/01/2011
• Status: Active Grant

• Current Assignee: Qualcomm, Inc.
• Original Assignee: Qualcomm, Inc.
• Inventors: Wang, Xianglin, Guo, Liwei, Zheng, Yunfei, Karczewicz, Marta
• Granted Patent: US 9,807,401 B2
• Time in Patent Office: Days
• US Class Current: 375/240.18
• CPC Class Codes: H04N 19/122 Selection of transform size, e.g. 8x8 or 2x4x8 DCT; Selection of sub-band transforms of varying structure or type
• 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 19/61 in combination with predictive coding
• CPC Class Codes: H04N 19/96 Tree coding, e.g. quad-tree coding

Abstract

A video encoding device is configured to obtain an N by N array of residual values for a luma component and a corresponding N/2 by N array of residual values for a chroma component. The video encoding device may partition the N/2 by N array of residual values for the chroma component into two N/2 by N/2 sub-arrays of chroma residual values. The video encoding device may further partition the sub-arrays of chroma residual values based on the partitioning of the array of residual values for the luma component. Video encoding device may perform a transform on each of the sub-arrays of chroma residual values to generate transform coefficients. A video decoding device may use data defining sub-arrays of transform coefficients to perform a reciprocal process to generate residual values.

Claims

1. A method for decoding video data comprising:

obtaining an N by N array of transform coefficients for a luma component of video data;

obtaining a plurality of corresponding N/2 by N/2 sub-arrays of transform coefficients for a chroma component of video data;

performing an inverse transform on each of the N/2 by N/2 sub-arrays of chroma transform coefficients to generate a plurality of N/2 by N/2 sub-arrays of residual sample values; and

combining each of the N/2 by N/2 sub-arrays of residual sample values to form an N/2 by N array of residual sample values for the chroma component.

2. A method for decoding video data comprising:

obtaining a plurality of N/2 by N/2 sub-arrays of transform coefficients for a luma component of video data;

obtaining a plurality of corresponding N/4 by N/4 sub-arrays of transform coefficients for a chroma component of video data;

performing an inverse transform on each of the plurality of N/4 by N/4 sub-arrays of chroma transform coefficients to generate a plurality of N/4 by N/4 sub-arrays of residual sample values; and

combining each of the N/4 by N/4 sub-arrays of residual sample values to form an N/2 by N array of residual sample values for the chroma component.

3. The method of claim 2, wherein N equals 16 and wherein the plurality of N/2 by N/2 sub-arrays of transform coefficients for the luma component of video data includes four sub-arrays and the plurality of corresponding N/4 by N/4 sub-arrays of transform coefficients of the chroma component includes eight sub-arrays.

4. A method for decoding video data comprising:

obtaining four N by N/4 sub-arrays of transform coefficients for a luma component of video data;

obtaining eight corresponding N/2 by N/8 sub-arrays of transform coefficients for a chroma component of video data;

performing an inverse transform on each of the N/2 by N/8 sub-arrays of chroma transform coefficients to generate eight N/2 by N/8 sub-arrays of residual sample values; and

combining each of the N/2 by N/8 sub-arrays of residual sample values to form an N/2 by N array of residual sample values for the chroma component.

5. The method of claim 4, wherein N equals 32.

6. A video decoding device comprising one or more processors configured to:

obtain an N by N array of transform coefficients for a luma component of video data;

obtain a plurality of corresponding N/2 by N/2 sub-arrays of transform coefficients for a chroma component of video data;

perform an inverse transform on each of the N/2 by N/2 sub-arrays of chroma transform coefficients to generate a plurality of N/2 by N/2 sub-arrays of residual sample values; and

combine each of the N/2 by N/2 sub-arrays of residual sample values to form an N/2 by N array of residual sample values for the chroma component.

7. A video decoding device comprising one or more processors configured to:

obtain a plurality of N/2 by N/2 sub-arrays of transform coefficients for a luma component of video data;

obtain a plurality of corresponding N/4 by N/4 sub-arrays of transform coefficients for a chroma component of video data;

perform an inverse transform on each of the plurality of N/4 by N/4 sub-arrays of chroma transform coefficients to generate a plurality of N/4 by N/4 sub-arrays of residual sample values; and

combine each of the N/4 by N/4 sub-arrays of residual sample values to form an N/2 by N array of residual sample values for the chroma component.

8. The video decoding device of claim 7, wherein N equals 16 and wherein the plurality of N/2 by N/2 sub-arrays of transform coefficients for the luma component of video data includes four sub-arrays and the plurality of corresponding N/4 by N/4 sub-arrays of transform coefficients of the chroma component includes eight sub-arrays.

9. A video decoding device comprising one or more processors configured to:

obtain four N by N/4 sub-arrays of transform coefficients for a luma component of video data;

obtain eight corresponding N/2 by N/8 sub-arrays of transform coefficients for a chroma component of video data;

perform an inverse transform on each of the N/2 by N/8 sub-arrays of chroma transform coefficients to generate eight of N/2 by N/8 sub-arrays of residual sample values; and

combine each of the N/2 by N/8 sub-arrays of residual sample values to form an N/2 by N array of residual sample values for the chroma component.

10. The video decoding device of claim 9, wherein N equals 32.

11. A method for encoding video data comprising:

obtaining an N by N array of residual values for a luma component of video data;

obtaining a corresponding N/2 by N array of residual values for a chroma component of video data;

partitioning the N/2 by N array of residual values for the chroma component into two N/2 by N/2 sub-arrays of chroma residual values; and

performing a transform on each of the sub-arrays of chroma residual values to generate transform coefficients for the chroma component.

12. The method of claim 11, further comprising partitioning the N by N array of residual values for the luma component into four sub-arrays of luma residual values and further partitioning each of the two N/2 by N/2 sub-arrays of chroma residual values into four sub-arrays of chroma residual values based at least in part on the partitioning applied to the N by N array of residual values for the luma component.

13. The method of claim 12, wherein partitioning the N by N array of residual values for the luma component into four sub-arrays of luma residual values results in four 8 by 8 sub-arrays of luma residual values and wherein further partitioning each of the two N/2 by N/2 sub-arrays of chroma residual values into four sub-arrays of chroma residual values results in eight 4 by 4 sub-arrays of chroma residual values.

14. The method of claim 12, wherein partitioning the N by N array of residual values for the luma component into four sub-arrays of luma residual values results in four sub-arrays of luma residual values with aspect ratios of 4 to 1 and wherein further partitioning each of the two N/2 by N/2 sub-arrays of chroma residual values into four sub-arrays of chroma residual values based at least in part on the partitioning applied to the N by N array of residual values for the luma component results in eight sub-arrays of chroma residual values with aspect ratios of 4 to 1.

15. The method of claim 11, wherein partitioning the N/2 by N array of residual values for the chroma component into two N/2 by N/2 sub-arrays of chroma residual values includes determining not to partition the N/2 by N array of residual values into two N/2 by N/2 sub-arrays of chroma residual values, if the two N/2 by N/2 sub-arrays are below a minimum transform unit size.

16. A video encoding device comprising one or more processors configured to:

obtain an N by N array of residual values for a luma component of video data;

obtain a corresponding N/2 by N array of residual values for a chroma component of video data;

partition the N/2 by N array of residual values for the chroma component into two N/2 by N/2 sub-arrays of chroma residual values; and

perform a transform on each of the sub-arrays of chroma residual values to generate transform coefficients for the chroma component.

17. The device of claim 16, wherein the one or more processors are further configured to partition the N by N array of residual values for the luma component into four sub-arrays of luma residual values and further partition each of the two N/2 by N/2 sub-arrays of chroma residual values into four sub-arrays of chroma residual values based at least in part on the partitioning applied to the N by N array of residual values for the luma component.

18. The device of claim 17, wherein partitioning the N by N array of residual values for the luma component into four sub-arrays of luma residual values results in four 8 by 8 sub-arrays of luma residual values and wherein further partitioning each of the two N/2 by N/2 sub-arrays of chroma residual values into four sub-arrays of chroma residual values results in eight 4 by 4 sub-arrays of chroma residual values.

19. The device of claim 17, wherein partitioning the N by N array of residual values for the luma component into four sub-arrays of luma residual values results in four sub-arrays of luma residual values with aspect ratios of 4 to 1 and wherein further partitioning each of the two N/2 by N/2 sub-arrays of chroma residual values into four sub-arrays of chroma residual values based at least in part on the partitioning applied to the N by N array of residual values for the luma component results in eight sub-arrays of chroma residual values with aspect ratios of 4 to 1.

20. The device of claim 16, wherein partitioning the N/2 by N array of residual values for the chroma component into two N/2 by N/2 sub-arrays of chroma residual values includes determining not to partition the N/2 by N array of residual values into two N/2 by N/2 sub-arrays of chroma residual values, if the two N/2 by N/2 sub-arrays are below a minimum transform unit size.

21. A non-transitory computer-readable medium comprising instructions stored thereon that when executed cause one or more processors to:

obtain an N by N array of residual values for a luma component of video data;

obtain a corresponding N/2 by N array of residual values for a chroma component of video data;

partition the N/2 by N array of residual values for the chroma component into two N/2 by N/2 sub-arrays of chroma residual values; and

performing a transform on each of the sub-arrays of chroma residual values to generate transform coefficients for the chroma component.

22. The non-transitory computer-readable medium of claim 21, wherein the instructions further cause one or more processors to partition the N by N array of residual values for the luma component into four sub-arrays of luma residual values and further partition each of the two N/2 by N/2 sub-arrays of chroma residual values into four sub-arrays of chroma residual values based at least in part on the partitioning applied to the N by N array of residual values for the luma component.

23. The non-transitory computer-readable medium of claim 22, wherein partitioning the N by N array of residual values for the luma component into four sub-arrays of luma residual values results in four 8 by 8 sub-arrays of luma residual values and wherein further partitioning each of the two N/2 by N/2 sub-arrays of chroma residual values into four sub-arrays of chroma residual values results in eight 4 by 4 sub-arrays of chroma residual values.

24. The non-transitory computer-readable medium of claim 22, wherein partitioning the N by N array of residual values for the luma component into four sub-arrays of luma residual values results in four sub-arrays of luma residual values with aspect ratios of 4 to 1 and wherein further partitioning each of the two N/2 by N/2 sub-arrays of chroma residual values into four sub-arrays of chroma residual values based at least in part on the partitioning applied to the N by N array of residual values for the luma component results in eight sub-arrays of chroma residual values with aspect ratios of 4 to 1.

25. The non-transitory computer-readable medium of claim 21, wherein partitioning the N/2 by N array of residual values for the chroma component into two N/2 by N/2 sub-arrays of chroma residual values includes determining not to partition the N/2 by N array of residual values into two N/2 by N/2 sub-arrays of chroma residual values, if the two N/2 by N/2 sub-arrays are below a minimum transform unit size.

26. A method of encoding video data comprising:

determining whether a first transform unit associated with a chroma component includes at least one nonzero transform coefficients;

determining whether a second transform unit associated with the chroma component includes at least one nonzero transform coefficients; and

generating a joint coded block flag for the first transform unit and the second transform unit, wherein the joint coded block flag indicates whether either of the first transform unit or the second transform unit includes at least one nonzero transform coefficients.

27. The method of claim 26, further comprising, when the joint coded block flag indicates that either of the first transform unit or the second transform unit includes at least one nonzero transform coefficient, generating a coded block flag that indicates whether the first transform unit includes at least one nonzero coefficient.

28. The method of claim 27, further comprising when the coded block flag indicates that the first transform unit includes at least one nonzero coefficient, generating a coded block flag for the second transform unit that indicates whether the second transform unit includes at least one nonzero transform coefficient.