Binarization of prediction residuals for lossless video coding

• US 9,503,750 B2
• Filed: 11/02/2012
• Issued: 11/22/2016
• Est. Priority Date: 11/04/2011
• Status: Active Grant

• Current Assignee: Futurewei Technologies Incorporated (Huawei Investment & Holding Co., Ltd.)
• Original Assignee: Futurewei Technologies Incorporated (Huawei Investment & Holding Co., Ltd.)
• Inventors: Gao, Wen, Jiang, Minqiang, Yu, Haoping
• Primary Examiner(s): Elahi, Shan
• Application Number:
  US13/668,112
• Publication Number:
  US 20130114693A1
• Time in Patent Office: 1,481 Days
• Field of Search: 375/240.03
• US Class Current: 1/1
• CPC Class Codes: H04N 19/11 among a plurality of spatial predictive coding modes
• CPC Class Codes: H04N 19/137 Motion inside a coding unit, e.g. average field, frame or block difference
• CPC Class Codes: H04N 19/147 according to rate distortion criteria rate-distortion as a criterion for motion estimation H04N19/567
• CPC Class Codes: H04N 19/174 the region being a slice, e.g. a line of blocks or a group of blocks
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/182 the unit being a pixel
• CPC Class Codes: H04N 19/593 involving spatial prediction techniques
• CPC Class Codes: H04N 19/91 Entropy coding, e.g. variable length coding [VLC] or arithmetic coding

Abstract

A video codec comprising a processor configured to compute a difference between an original pixel and a prediction pixel to generate a prediction residual, and binarize an absolute value of the prediction residual. A method for video coding comprising computing a difference between an original pixel and a prediction pixel to generate a prediction residual, and binarizing an absolute value of the prediction residual. A video encoder comprising a processor configured to set a number of syntax elements to indicate lossless encoding of some or all coding units (CU) in a video frame, a transmitter configured to transmit a bit stream comprising the syntax elements. A method for video encoding comprising setting a number of syntax elements to indicate lossless encoding of some or all coding units (CU) in a video frame, and transmitting a bit stream comprising the syntax elements.

Claims

1. A video codec comprising:

a memory; and

a processor operably coupled to the memory and configured to;

compute a difference between an original pixel and a prediction pixel to generate a prediction residual;

binarize an absolute value of the prediction residual, wherein binarizing |R| comprises using a truncated Golomb-Rice (TR) code with a Golomb-Rice parameter when the absolute value of the prediction residual, denoted as |R|, is greater than 2, and wherein the Golomb-Rice parameter is denoted as cRiceParam, is an integer, and has a valid range from 0 to n, where n is greater than 3;

compute a difference between a second original pixel and a second prediction pixel to generate a second prediction residual; and

binarize an absolute value of the second prediction residual, wherein binarizing the absolute value of the second prediction residual comprises using an updated cRiceParam when the absolute value of the second prediction residual is greater than 2, and wherein the updated cRiceParam is adaptively based on |R| and the cRiceParam,

wherein the prediction residual and the second prediction residual are binarized without any prior or subsequent transform and quantization steps being performed, and

wherein the prediction residual and the second prediction residual are used for coding of a video frame.

2. The video codec of claim 1, wherein cRiceParam determines another parameter denoted as cTRMax, wherein when |R|-3>

cTRMax, binarizing cRiceParam comprises using a k-th order exponential Golomb (EG) code, and wherein k is a positive integer.

3. The video codec of claim 2, wherein cRiceParam and cTRMax are specified according to following table:

4. The video codec of claim 1, wherein the processor is further configured to determine a minimum value between |R|-3 and a threshold, wherein the threshold is a preconfigured integer greater than 14, and wherein the updated cRiceParam is no less than the cRiceParam when the minimum value equals |R|-3.

5. The video codec of claim 1, wherein the cRiceParam is denoted as c_t, wherein the updated cRiceParam is denoted as c_t+1, wherein the processor is further configured to determine a minimum value between |R|-3 and 15, and wherein c_t+1 is determined by the minimum value and C_t via the following table:

6. A video codec comprising:

a memory; and

a processor operably coupled to the memory and configured to;

compute a difference between an original pixel and a prediction pixel to generate a prediction residual; and

binarize an absolute value of the prediction residual, wherein binarizing |R| comprises using a truncated Golomb-Rice (TR) code with a Golomb-Rice parameter when the absolute value of the prediction residual, denoted as |R|, is greater than 2, and wherein the Golomb-Rice parameter is denoted as cRiceParam, is an integer, and has a valid range from 0 to n, where n is greater than 3;

generate a plurality of prediction residuals including the prediction residual and a second prediction residual, wherein the plurality of prediction residuals are located in one block and each has a non-zero absolute value;

update the cRiceParam to generate an updated cRiceParam after binarizing |R|, wherein updating the cRiceParam is based on a counter and an accumulator, wherein the counter specifies a number of prediction residuals that have been binarized using the TR code, wherein the accumulator specifies a summation of each absolute value of the prediction residuals that have been binarized using the TR code minus 3; and

binarize the absolute value of the second prediction residual using the TR code with the updated cRiceParam when an absolute value of the second prediction residual is greater than 2,

wherein the prediction residual and the second prediction residual are binarized without any prior or subsequent transform and quantization steps being performed, and

wherein the prediction residual and the second prediction residual are used for coding of a video frame.

7. The video codec of claim 6, wherein the counter is denoted as N, wherein the accumulator is denoted as A, and wherein N increases by one and A increases by |R|-3 when 2<

|R|≦

cTRMax+3, otherwise N increases by one and A increases by cTRMax+3 when |R|>

cTRMax+3.

8. The video codec of claim 7, wherein the updated cRiceParam is computed as:

for(i=0;

(N<

<

i)<

A;

i++);

where cRiceParam=min(i, 5).

9. A method for video coding in a video coder, comprising:

computing, by the video coder, a difference between an original pixel and a prediction pixel to generate a prediction residual;

binarizing, by the video coder, an absolute value of the prediction residual;

computing, by the video coder, a difference between a second original pixel and a second prediction pixel, to generate a second prediction residual; and

binarizing, by the video coder, an absolute value of the second prediction residual, wherein binarizing the absolute value of the second prediction residual comprises using an updated cRiceParam when the absolute value of the second prediction residual is greater than 2, and wherein the updated cRiceParam is adaptively based on |R| and the cRiceParam,

wherein the prediction residual and the second prediction residual are binarized without any prior or subsequent transform and quantization steps being performed, and

wherein the prediction residual and the second prediction residual are used for coding of a video frame.

10. The method of claim 9, wherein binarizing |R| comprises using a truncated Golomb-Rice (TR) code with a Golomb-Rice parameter when the absolute value of the prediction residual, denoted as |R|, is greater than 2, and wherein the Golomb-Rice parameter is denoted as cRiceParam, is an integer, and has a valid range from 0 to n, where n is greater than 3.

11. The method of claim 10, wherein cRiceParam determines another parameter denoted as cTRMax, wherein binarizing cRiceParam comprises using a k-th order exponential Golomb (EG) code when |R1|-3>

cTRMax, and wherein k is a positive integer.

12. The method of claim 9, further comprising determining a minimum value between |R|-3 and a threshold, wherein the threshold is a pre-configured integer greater than 14, and wherein the updated cRiceParam is no less than the cRiceParam when the minimum value equals |R|-3.

13. A method for video coding in a video coder, comprising:

computing, by the video coder, a difference between an original pixel and a prediction pixel to generate a prediction residual;

binarizing, by the video coder, an absolute value of the prediction residual, wherein binarizing |R| comprises using a truncated Golomb-Rice (TR) code with a Golomb-Rice parameter when the absolute value of the prediction residual, denoted as |R|, is greater than 2, and wherein the Golomb-Rice parameter is denoted as cRiceParam, is an integer, and has a valid range from 0 to n, where n is greater than 3;

generating, by the video coder, a plurality of prediction residuals including the prediction residual and a second prediction residual, wherein the plurality of prediction residuals are located in one block and each has a non-zero absolute value;

updating, by the video coder, the cRiceParam to generate an updated cRiceParam after binarizing |R|, wherein updating the cRiceParam is based on a counter and an accumulator, wherein the counter specifies a number of prediction residuals that have been binarized using the TR code, wherein the accumulator specifies a summation of each absolute value of the prediction residuals that have been binarized using the TR code minus 3; and

binarizing, by the video coder, the absolute value of the second prediction residual using the TR code with the updated cRiceParam when an absolute value of the second prediction residual is greater than 2,

wherein the prediction residual the second prediction residual are binarized without any prior or subsequent transform and quantization steps being performed, and

wherein the prediction residual and the second prediction residual are used for coding of a video frame.

14. A video encoder comprising:

a processor configured to set a number of syntax elements to indicate lossless encoding of some or all coding units (CU) in a video frame; and

a transmitter operably coupled to the processor and configured to transmit a bit stream comprising the syntax elements to a video device in communication with the video encoder,

wherein a first syntax element is set to disable quantization parameter (QP) configuration on a CU level,

wherein the first syntax element is included in a sequence parameter set (SPS) of the bit stream, and

wherein the first syntax element is binarized without any prior or subsequent transform and quantization steps being performed.

15. The video encoder of claim 14, wherein a second syntax element is set to specify an initial value of QP for the video frame minus 26, wherein the second syntax element is included in a picture parameter set (PPS) of the bit stream, wherein a third syntax element is set to specify a difference between a slice QP value and the initial value of QP for the video frame, wherein the third syntax element is included in slice headers associated with the video frame.

16. The video encoder of claim 15, wherein the first syntax element is set as cu_qp_delta_enabled_flag=0, wherein the second syntax element is set as pic_init_qp_minus26=26, and wherein the third syntax element is set as slice_qp_delta=0.

17. A video encoder comprising:

a processor configured to set a number of syntax elements to indicate lossless encoding of some or all coding units (CU) in a video frame;

a transmitter operably coupled to the processor and configured to transmit a bit stream comprising the syntax elements to a video decoder for use in generating a visual image on a display device,

wherein a first syntax element is set to enable quantization parameter (QP) configuration on a CU level,

wherein the first syntax element is included in a sequence parameter set (SPS) of the bit stream, and

wherein the first syntax element is binarized without any prior or subsequent transform and quantization steps being performed.

18. The video encoder of claim 17, wherein a second syntax element is set to identify an initial value of QP for the video frame minus 26, wherein the second syntax element is included in a picture parameter set (PPS) of the bit stream, wherein a third syntax element is set to specify a difference between a slice QP value and the initial value of QP for the video frame, wherein a fourth syntax is set to specify a difference between a QP value for a CU and the slice QP value.

19. The video encoder of claim 18, wherein lossless encoding of the CU is indicated by setting the first syntax element as cu_qp_delta_enabled_flag=1, the second syntax element as pic_init_qp_minus26=26, and the third syntax element as slice_qp_delta=0.

20. A method for video encoding in a video coder, comprising:

setting, by the video coder, a number of syntax elements to indicate lossless encoding of some or all coding units (CU) in a video frame; and

transmitting, by the video coder, a bit stream comprising the syntax elements to a video decoder for use in generating a visual image on a display device,

wherein a first syntax element is set to disable quantization parameter (QP) configuration on a CU level,

wherein the first syntax element is included in a sequence parameter set (SPS) of the bit stream,

wherein a second syntax element is set to specify an initial value of QP for the video frame minus 26,

wherein the second syntax element is included in a picture parameter set (PPS) of the bit stream,

wherein a third syntax element is set to specify a difference between a slice QP value and the initial value of QP for the video frame,

wherein the third syntax element is included in slice headers associated with the video frame, and

wherein the syntax elements are binarized without any prior or subsequent transform and quantization steps being performed.

21. A method for video encoding in a video coder, comprising:

setting, by the video coder, a number of syntax elements to indicate lossless encoding of some or all coding units (CU) in a video frame; and

transmitting, by the video coder, a bit stream comprising the syntax elements to a video decoder for use in generating a visual image on a display device,

wherein a first syntax element is set to enable quantization parameter (QP) configuration on a CU level,

wherein the first syntax element is included in a sequence parameter set (SPS) of the bit stream,

wherein a second syntax element is set to identify an initial value of QP for the video frame minus 26,

wherein the second syntax element is included in a picture parameter set (PPS) of the bit stream,

wherein a third syntax element is set to specify a difference between a slice QP value and the initial value of QP for the video frame,

wherein a fourth syntax is set to specify a difference between a QP value for a CU and the slice QP value, and

wherein the syntax elements are binarized without any prior or subsequent transform and quantization steps being performed.