Granted Patent B2

US 12,382,052 · App. 17/489,793 · Granted Aug 5, 2025

Inventors: Semih Esenlik (Munich, DE); Max Blaeser (Aachen, DE); Zhijie Zhao (Shenzhen, CN); Han Gao (Munich, DE); Anand Meher Kotra (Munich, DE); Biao Wang (Munich, DE); Elena Alexandrovna Alshina (Munich, DE)

Assignee: HUAWEI TECHNOLOGIES CO., LTD.

H04N19/132H04N19/105H04N19/159H04N19/176H04N19/186

Abstract

Claims (62)

1. A method of coding implemented by a decoding device, comprising:

obtaining a value of a parameter for a current block, the value of the parameter indicating a partition mode for the current block;

obtaining a first prediction mode for the current block;

obtaining a second prediction mode for the current block;

generating a first prediction value for a chroma sample in the current block according to the first prediction mode;

generating a second prediction value for a chroma sample in the current block according to the second prediction mode;

obtaining a combined prediction value by combining the first prediction value and the second prediction value;

obtaining a weight value for a luma sample by calculating the weight value for the luma sample; and obtaining a first weight value for a chroma sample by calculating the first weight value for the chroma sample;

obtaining a combined prediction value for the chroma sample in the current block, according to the first prediction value for the chroma sample and the first weight value for the chroma sample; and

wherein the combined prediction value is obtained by combining the first prediction value and the second prediction value according to a blending operation, the blending operation is implemented using a function of a sample distance sample_dist as input and sample Weight1 as output,

wherein said sample_dist function is a division or multiplication with a constant number operation, a right shifting a constant amount operation, a taking the absolute value or a clipping operation or a combination of those, and wherein sample_dist=((x<<1)+1)*Dis[angleIdx1]+ ((y<<1)+1))*Dis[angleIdx2]−offset (distanceIdx), and wherein angleIdx1 and angleIdx2 represent quantized trigonometric parameters of a separation line and angleIdx2 is equal to a remainder of a variable (displacementX) plus 8 and divided by 32, x and y are the −x and −y coordinates of a sample with respect to top-left sample of a coding block, offset (distanceIdx) is an offset value, which is a function of an index value (distanceIdx), and Dis[ ] is a lookup table describing the change in the sample distance with respect to a unit increase in the x or y directions, and wherein sample Weight1=Clip3(0, 8, f (sample_dist)), wherein f( ) 1s a division/addition/multiplication with a constant number operation, a right shifting with a constant amount operation, taking the absolute value operation or a clipping operation or a combination of those,

wherein sample Weight1=Clip3(0, 8, sample_dist*K+Clip3(0, 8, sample_dist+4)), sample_dist represents a distance of the luma sample to a separation line of the current block, and K is an integer with a value greater than 0.

2. The method according to claim 1 , wherein the combined prediction value is obtained by combining the first prediction value and the second prediction value according to the partition mode.

3. The method of claim 1 , wherein the obtaining of the first weight value for the chroma sample, is related to obtaining of the weight value for the luma sample in the current block as,

sampleWeight C 1( x,y )=sampleWeight1(2* x, 2* y );

wherein x and y are coordinates of a chroma sample with respect to the top-left coordinate of the chroma block of the coding block, sampleWeightC1 (a, b) is a first weight value for the chroma sample which is located at a sample coordinate (a,b) in the chroma block of the current block, a and b being coordinate values; and sampleWeight1 (c, d) represents a weight value corresponding to a luma sample which is located at a sample coordinate (c, d) in a luma block of the current block, c and d being coordinate values.

4. The method of claim 1 , wherein the obtaining the first weight value for the chroma sample, is related to obtaining of the weight value for the luma sample in the current block as,

sampleWeight C 1( x,y )=sampleWeight1(2* x− 1,2* y− 1);

or

sampleWeight C 1( x,y )=(sampleWeight1(2* x− 1,2* y− 1)+sampleWeight1(2* x, 2* y )+ K )>>1;

or

sampleWeight C 1( x,y )=(sampleWeight1(2* x− 1,2* y− 1)+sampleWeight1(2* x, 2* y )+sampleWeight1(2* x, 2* y− 1)+sampleWeight1(2* x− 1,2* y )+ K )>>2;

or

sampleWeight C 1( x,y )=(sampleWeight1(2* x− 1,2* y− 1)+sampleWeight1(2* x, 2* y )+sampleWeight1(2* x, 2* y− 1)+sampleWeight1(2* x− 1,2* y )+2)>>2;

or

sampleWeight C 1( x,y )=(sampleWeight1(2* x− 1,2* y− 1)+sampleWeight1(2* x, 2* y )+1)>>1;

wherein x and y are coordinates of a chroma sample with respect to the top-left coordinate of the chroma block of the coding block, sampleWeightC1 (a, b) is a first weight value for the chroma sample which is located at a sample coordinate (a,b) in the chroma block of the current block, a and b being coordinate values; and sampleWeight1 (c, d) represents a weight value corresponding to a luma sample which is located at a sample coordinate (c, d) in a luma block of the current block, c and d being coordinate values; and K is an integer value.

5. The method according to claim 1 , wherein the partition mode is a geometrical model.

6. The method according to claim 1 , wherein the parameter is an angle parameter or a distance parameter.

7. The method according to claim 1 , where the first prediction mode and the second prediction mode are different.

8. The method according to claim 1 , wherein the first prediction mode or the second prediction mode is an inter prediction mode, wherein information for the inter prediction mode comprises a reference picture index or a motion vector.

9. The method according to claim 1 , wherein the first prediction mode or the second prediction mode is an intra prediction mode, wherein information for the intra prediction mode comprises an intra prediction mode index.

10. The method according to claim 1 , wherein K 1s indicated by a value of an indicator in a bitstream decoded by the decoding device.

11. A video decoder, comprising: one or more processors; and a non-transitory computer-readable storage medium coupled to the one or more processors and storing programming instructions for execution by the one or more processors, wherein the one or more processors, upon executing the programming instructions, cause the video decoder to carry out operations of:

obtaining a value of a parameter for a current block, the value of the parameter indicating a partition mode for the current block;

obtaining a first prediction mode for the current block;

obtaining a second prediction mode for the current block;

generating a first prediction value for a chroma sample in the current block according to the first prediction mode;

generating a second prediction value for a chroma sample in the current block according to the second prediction mode;

obtaining a combined prediction value by combining the first prediction value and the second prediction value;

obtaining a weight value for a luma sample by calculating the weight value for the luma sample; and obtaining a first weight value for a chroma sample by calculating the first weight value for the chroma sample;

obtaining a combined prediction value for the chroma sample in the current block, according to the first prediction value for the chroma sample and the first weight value for the chroma sample; and

wherein the combined prediction value is obtained by combining the first prediction value and the second prediction value according to a blending operation, the blending operation is implemented using a function of a sample distance sample_dist as input and sample Weight1 as output,

wherein said sample_dist function is a division or multiplication with a constant number operation, a right shifting a constant amount operation, a taking the absolute value or a clipping operation or a combination of those, and wherein sample_dist=((x<<1)+1)*Dis[angleIdx1]+ ((yv<<1)+1))*Dis[angleIdx2]−offset (distanceIdx), and wherein angleIdx1 and angleIdx2 represent quantized trigonometric parameters of a separation line and angleIdx2 is equal to a remainder of a variable (displacementX) plus 8 and divided by 32, x and y are the −x and −y coordinates of a sample with respect to top-left sample of a coding block, offset (distanceIdx) is an offset value, which is a function of an index value (distanceIdx), and Dis[ ] is a lookup table describing the change in the sample distance with respect to a unit increase in the x or y directions, and

wherein sample Weight1=Clip3(0, 8, f (sample_dist)), wherein f( ) is a division/addition/multiplication with a constant number operation, a right shifting with a constant amount operation, taking the absolute value operation or a clipping operation or a combination of those,

wherein sample Weight1=Clip3(0, 8, sample_dist*K+Clip3(0, 8, sample_dist+4) sample_dist represents a distance of the luma sample to a separation line of the current block, and K is an integer with a value greater than 0.

12. The video decoder according to claim 11 , wherein the combined prediction value is obtained by combining the first prediction value and the second prediction value according to the partition mode.

13. The method according to claim 11 , wherein K 1s indicated by a value of an indicator in a bitstream decoded by the video decoder.

14. A non-transitory computer-readable medium having stored thereon instructions that, when executed by one or more processors, cause the one or more processors to perform operations including:

obtaining a value of a parameter for a current block, the value of the parameter indicating a partition mode for the current block;

obtaining a first prediction mode for the current block;

obtaining a second prediction mode for the current block;

generating a first prediction value for a chroma sample in the current block according to the first prediction mode;

generating a second prediction value for a chroma sample in the current block according to the second prediction mode;

obtaining a combined prediction value by combining the first prediction value and the second prediction value;

obtaining a weight value for a luma sample by calculating the weight value for the luma sample; and obtaining a first weight value for a chroma sample by calculating the first weight value for the chroma sample;

obtaining a combined prediction value for the chroma sample in the current block, according to the first prediction value for the chroma sample and the first weight value for the chroma sample; and

wherein the combined prediction value is obtained by combining the first prediction value and the second prediction value according to a blending operation, the blending operation is implemented using a function of a sample distance sample_dist as input and sample Weight1 as output,

wherein said sample_dist function is a division or multiplication with a constant number operation, a right shifting a constant amount operation, a taking the absolute value or a clipping operation or a combination of those, and wherein sample_dist=((x<<1)+1)*Dis[angleIdx1]+ ((y<<11)+1))*Dis[angleIdx2]−offset (distanceIdx), and wherein angleIdx1 and angleIdx2 represent quantized trigonometric parameters of a separation line and angleIdx2 is equal to a remainder of a variable (displacementX) plus 8 and divided by 32, x and y are the −x and −y coordinates of a sample with respect to top-left sample of a coding block, offset (distanceIdx) is an offset value, which is a function of an index value (distanceIdx), and Dis[ ] is a lookup table describing the change in the sample distance with respect to a unit increase in the x or y directions, and

wherein sampleWeight1=Clip3(0, 8, f (sample_dist)), wherein f( ) is a division/addition/multiplication with a constant number operation, a right shifting with a constant amount operation, taking the absolute value operation or a clipping operation or a combination of those,

wherein sample Weight1=Clip3(0, 8, sample_dist*K+Clip3(0, 8, sample_dist+4) sample_dist represents a distance of the luma sample to a separation line of the current block, and K is an integer with a value greater than 0.

15. The method according to claim 14 , wherein K 1s indicated by a value of an indicator in a bitstream decoded by the one or more processors.

Assignments (1)

Priority Claims (1)

Continuity (2)

References Cited (32)