Simplification of LM Mode

• US 20130188705A1
• Filed: 01/24/2013
• Published: 07/25/2013
• Est. Priority Date: 01/24/2012
• Status: Active Grant

• Current Assignee: Futurewei Technologies Incorporated (Huawei Investment & Holding Co., Ltd.)
• Original Assignee: Futurewei Technologies Incorporated (Huawei Investment & Holding Co., Ltd.)
• Inventors: Liu, Lingzhi
• Granted Patent: US 9,225,984 B2
• Time in Patent Office: Days
• US Class Current: 375/240.12
• CPC Class Codes: H04N 19/186 the unit being a colour or a chrominance component
• CPC Class Codes: H04N 19/42 characterised by implementation details or hardware specially adapted for video compression or decompression, e.g. dedicated software implementation H04N19/635 takes precedence
• CPC Class Codes: H04N 19/50 using predictive coding H04N19/61 takes precedence
• CPC Class Codes: H04N 19/593 involving spatial prediction techniques

Abstract

A video codec comprising a processor configured to compute an average of a plurality of reconstructed chroma samples located in neighboring blocks of a chroma block and adjacent to the chroma block, compute an intermediate variable by applying an integer function on the average of the reconstructed chroma samples, and generate a predicted chroma sample for the chroma block based on the intermediate variable.

Claims

1. A video codec comprising:

a processor configured to;

compute an average of a plurality of reconstructed chroma samples located in neighboring blocks of a chroma block and adjacent to the chroma block;

compute an intermediate variable by applying an integer function on the average of the reconstructed chroma samples; and

generate a predicted chroma sample for the chroma block based on the intermediate variable.

2. The video codec of claim 1, wherein the integer function is selected from the group consisting of a rounding function, a ceiling function, and a flooring function.

3. The video codec of claim 2, wherein the intermediate variable is a first intermediate variable denoted as avgC, and wherein the processor is further configured to compute a second intermediate variable representing a difference between a sum of the plurality of reconstructed chroma samples and 2*nS*avgC, wherein nS denotes a width of the chroma block, wherein a number of the reconstructed chroma samples equals 2*nS, and wherein generating the predicted chroma sample is further based on the second intermediate variable.

4. The video codec of claim 1, wherein the processor is further configured to:

generate a first plurality of filtered luma samples by filtering a first plurality of reconstructed luma samples located in a corresponding reconstructed luma block;

generate a second plurality of filtered luma samples by filtering a second plurality of reconstructed luma samples located in neighboring blocks of the corresponding reconstructed luma block and adjacent to the corresponding reconstructed luma block;

compute an average of the second plurality of filtered luma samples; and

compute a third intermediate variable by applying the integer function on the average of the second plurality of filtered luma samples,

wherein generating the predicted chroma sample is further based on the third intermediate variable.

5. The video codec of claim 3, wherein the processor is further configured to:

generate a first plurality of filtered luma samples by filtering a first plurality of reconstructed luma samples located in a corresponding reconstructed luma block;

generate a second plurality of filtered luma samples by filtering a second plurality of reconstructed luma samples located in neighboring blocks of the corresponding reconstructed luma block and adjacent to the corresponding reconstructed luma block;

compute an average of the second plurality of filtered luma samples; and

compute a third intermediate variable by applying the integer function on the average of the second plurality of filtered luma samples,

wherein generating the predicted chroma sample is further based on the third intermediate variable.

6. The video codec of claim 5, wherein the processor is further configured to compute a fourth intermediate variable representing a difference between a sum of the second plurality of filtered luma samples and 2*nS*avgY′

, wherein both a width of the corresponding reconstructed luma block and a number of the second plurality of filtered luma samples equal 2*nS, wherein avgY′

denotes the third intermediate variable, and wherein generating the predicted chroma sample is further based on the fourth intermediate variable.

7. The video codec of claim 6, wherein computing the average of the reconstructed chroma samples comprises selectively right shifting the sum of the reconstructed chroma samples by a number of bits, wherein computing the average of the second plurality of filtered luma samples comprises selectively right shifting the sum of the second plurality of filtered luma samples by the number of bits, and wherein the number of bits is determined by nS and an input bit depth of the chroma block.

8. The video codec of claim 6, wherein the processor is further configured to:

compute an intermediate variables denoted as a1 based on the first, second, third, and fourth intermediate variables;

compute an intermediate variables denoted as a2 based on the third and fourth intermediate variables, wherein computing a1 and a2 uses no multiplier that has a bit width greater than an input bit depth of the chroma block; and

compute an integer division of a1 and a2 to generate an intermediate variable denoted as alpha,

wherein generating the predicted chroma sample is further based on alpha.

9. The video codec of claim 6, wherein the second and fourth intermediate variables are denoted as RErrC and RErrY′

respectively, wherein each of the first plurality of filtered luma samples is denoted as p_Y′

[x, y] with x and y being position indexes, wherein an input bit depth of the chroma block is denoted as BitDepth_C, wherein the processor is further configured to compute additional intermediate variables denoted as a1, a2, k, k1, alpha, and beta according to the following equations;

a1=LC−

(2*nS*avgY′

*avgC+avgY′

*RErrC+avgC*RErrY′

);

a2=LL−

(2*nS*avgY′

²+2*avgY′

*RErrY′

);

k1=Max(0, log₂(abs(a2))−

5)−

Max(0, log₂(abs(a1))−

14)+2;

k=13−

Max(0, log₂(abs(alpha))−

6); and

beta=avgC−

((alpha*avgY′

)>

>

k1),

wherein LC and LL are determined by the plurality of reconstructed chroma samples and the second plurality of filtered luma samples, wherein alpha is an integer division of a1 and a2, and wherein generating the predicted chroma sample, denoted as predSamples[x, y], uses equation;

predSamples[x, y]=Clip1_C(((p_Y′

[x, y]*alpha)>

>

k)+beta), with x, y=0 . . . nS−

1,

where function Clip1C(x) is defined as;

Clip1_C(x)=Clip3(0,(1<

<

BitDepth_C)−

1,x),

where function Clip3(x, y, z) is defined as;

10. The video codec of claim 9, wherein the processor is further configured to determine LL and LC via

either equations:

11. The video codec of claim 6, wherein the processor is further configured to set a parameter denoted as LM_REDUCED_BIT, and wherein avgY′

and avgC are computed according to the following equations;

12. The video codec of claim 6, wherein an input bit depth of the chroma block is between 8 and 10, and wherein the width of the chroma block is no less than 4 and no greater than 32.

13. The video codec of claim 3, wherein the processor is further configured to:

compute an average of a plurality of reconstructed luma samples located in neighboring blocks of a corresponding reconstructed luma block and adjacent to the corresponding reconstructed luma block;

compute a fifth intermediate variable, denoted as avgY, by applying the integer function on the average of the plurality of reconstructed luma samples; and

compute a sixth intermediate variable representing a difference between a sum of the plurality of reconstructed luma samples and 2*nS*avgY, wherein a width of the corresponding reconstructed luma block equals nS, and

wherein generating the predicted chroma sample is further based on the sixth intermediate variable.

14. The video codec of claim 13, wherein the processor is further configured to:

compute an intermediate variables denoted as a1 based on the first, second, fifth, and sixth intermediate variables;

compute an intermediate variables denoted as a2 based on the fifth and sixth intermediate variables, wherein computing a1 and a2 uses no multiplier that has a bit width greater than an input bit depth of the chroma block; and

compute an integer division of a1 and a2 to generate an intermediate variable denoted as alpha,

wherein generating the predicted chroma sample is further based on alpha.

15. The video codec of claim 2, wherein the processor is further configured to compute a difference, denoted as Δ

p, between each of the plurality of reconstructed chroma samples and the intermediate variable, and wherein generating the predicted chroma sample is further based on the difference.

16. The video codec of claim 15, wherein the processor is further configured to:

compute an average of a plurality of luma samples located in neighboring blocks of a corresponding reconstructed luma block and adjacent to the corresponding reconstructed luma block;

compute a second intermediate variable by applying the integer function on the average of the plurality of luma samples; and

compute a third intermediate variable to represent a difference between the second intermediate and each of the plurality of luma samples, and

wherein generating the predicted chroma sample is further based on the third intermediate variable.

17. The video codec of claim 2, wherein the processor is further configured to:

generate a residual block indicating a difference between the chroma block and the prediction block; and

entropy encode the residual block to generate an encoded residual block.

18. The video codec of claim 2, wherein the processor is further configured to:

decode an encoded residual block to generate a decoded residual block; and

combine the decoded residual block and the prediction block to generate the chroma block.

19. A method of video coding comprising:

computing an average of a plurality of reconstructed chroma samples located in neighboring blocks of a chroma block and adjacent to the chroma block;

computing an intermediate variable by applying an integer function on the average of the reconstructed chroma samples; and

generating a predicted chroma sample for the chroma block based on the intermediate variable.

20. The method of claim 19, wherein the integer function is selected from the group consisting of a rounding function and a flooring function.

21. The method of claim 20, wherein the intermediate variable is a first intermediate variable denoted as avgC, the method further comprising computing a second intermediate variable representing a difference between a sum of the plurality of reconstructed chroma samples and 2*nS*avgC, wherein nS denotes a width of the chroma block, wherein a number of the reconstructed chroma samples equals 2*nS, and wherein generating the predicted chroma sample is further based on the second intermediate variable.

22. The method of claim 21, further comprising:

generating a first plurality of filtered luma samples by filtering a first plurality of reconstructed luma samples located in a corresponding reconstructed luma block;

generating a second plurality of filtered luma samples by filtering a second plurality of reconstructed luma samples located in neighboring blocks of the corresponding reconstructed luma block and adjacent to the corresponding reconstructed luma block;

computing an average of the second plurality of filtered luma samples; and

computing a third intermediate variable by applying the integer function on the average of the second plurality of filtered luma samples,

wherein generating the predicted chroma sample is further based on the third intermediate variable.

23. The method of claim 22, further comprising computing a fourth intermediate variable representing a difference between a sum of the second plurality of filtered luma samples and 2*nS*avgY′

, wherein both a width of the corresponding reconstructed luma block and a number of the second plurality of filtered luma samples equal 2*nS, wherein avgY′

denotes the third intermediate variable, and wherein generating the predicted chroma sample is further based on the fourth intermediate variable.

24. The method of claim 23, wherein computing the average of the reconstructed chroma samples comprises selectively right shifting the sum of the reconstructed chroma samples by a number of bits, wherein computing the average of the second plurality of filtered luma samples comprises selectively right shifting the sum of the second plurality of filtered luma samples by the number of bits, and wherein the number of bits is determined by nS and an input bit depth of the chroma block.

25. The method of claim 23, further comprising:

computing an intermediate variables denoted as a1 based on the first, second, third, and fourth intermediate variables;

computing an intermediate variables denoted as a2 based on the third and fourth intermediate variables, wherein computing a1 and a2 uses no multiplier that has a bit width greater than an input bit depth of the chroma block; and

computing an integer division of a1 and a2 to generate an intermediate variable denoted as alpha,

wherein generating the predicted chroma sample is further based on alpha.

26. The method of claim 23, further comprising setting a parameter denoted as LM_REDUCED_BIT, wherein avgY′

and avgC are computed according to the following equations;

27. The method of claim 21, further comprising:

computing an average of a plurality of reconstructed luma samples located in neighboring blocks of a corresponding reconstructed luma block and adjacent to the corresponding reconstructed luma block;

computing a fifth intermediate variable, denoted as avgY, by applying the integer function on the average of the plurality of reconstructed luma samples; and

computing a sixth intermediate variable representing a difference between a sum of the plurality of reconstructed luma samples and 2*nS*avgY, wherein a width of the corresponding reconstructed luma block equals nS,

wherein generating the predicted chroma sample is further based on the sixth intermediate variable.

28. The method of claim 20, further comprising:

computing a difference, denoted as Δ

p, between each of the plurality of reconstructed chroma samples and the intermediate variable;

computing an average of a plurality of luma samples located in neighboring blocks of a corresponding reconstructed luma block and adjacent to the corresponding reconstructed luma block;

computing a second intermediate variable by applying the integer function on the average of the plurality of luma samples; and

computing a third intermediate variable to represent a difference between the second intermediate and each of the plurality of luma samples,

wherein generating the predicted chroma sample is further based on Δ

p and the third intermediate variable.

29. An apparatus comprising:

a processor configured to;

generate a prediction block for a chroma block, wherein the prediction block comprises a predicted chroma sample, and wherein generating the prediction block comprises;

computing a sum of a plurality of reconstructed chroma samples located in neighboring blocks of the chroma block and adjacent to the chroma block;

computing an intermediate variable based on the sum and using an integer function selectable from the group consisting of a rounding function and a flooring function; and

generating a predicted chroma sample in the prediction block using the intermediate variable.

30. The apparatus of claim 29, wherein computing the intermediate variable comprises:

if a second intermediate variable depending on a size and an input bit depth of the chroma block is zero,

computing an average of the plurality of reconstructed chroma samples from the sum;

applying the integer function on the average;

otherwise,

right shifting the sum to generate a shifted sum;

computing an average value from the shifted sum; and

applying the integer function on the average value.