Reference Picture Selection for Sub-Pixel Motion Estimation

• US 20100158120A1
• Filed: 12/23/2008
• Published: 06/24/2010
• Est. Priority Date: 12/23/2008
• Status: Active Grant

• Current Assignee: ARRIS Enterprises LLC (CommScope Holding Company, Inc.)
• Original Assignee: General Instrument Corporation (CommScope Holding Company, Inc.)
• Inventors: Wang, Limin, Luthra, Ajay, Fang, Xue
• Granted Patent: US 8,462,849 B2
• Time in Patent Office: Days
• US Class Current: 375/240.160
• CPC Class Codes: H04N 19/105 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 Class Codes: H04N 19/147 according to rate distortion criteria rate-distortion as a criterion for motion estimation H04N19/567
• CPC Class Codes: H04N 19/172 the region being a picture, frame or field
• CPC Class Codes: H04N 19/196 being specially adapted for the computation of encoding parameters, e.g. by averaging previously computed encoding parameters processing of motion vectors H04N19/513
• CPC Class Codes: H04N 19/46 Embedding additional information in the video signal during the compression process H04N19/517, H04N19/68, H04N19/70 take precedence
• CPC Class Codes: H04N 19/463 by compressing encoding parameters before transmission
• CPC Class Codes: H04N 19/52 by predictive encoding
• CPC Class Codes: H04N 19/523 with sub-pixel accuracy
• CPC Class Codes: H04N 19/573 Motion compensation with multiple frame prediction using two or more reference frames in a given prediction direction

Abstract

A video coding system includes a prediction module configured to perform multi-reference picture sub-pixel motion estimation. Reference pictures are selected for the sub-pixel motion estimation based on full-pixel motion estimation. A cost function is used to select the reference pictures from one or more groups of reference pictures for a block in a macroblock. The cost is based on full-pixel motion estimation errors and a predicted motion vector for the block. The selected reference pictures are used for sub-pixel motion estimation to get the final motion vector for each block.

Claims

1. A video coding system comprising:

a prediction module configured to perform multi-reference picture motion estimation, the prediction module including

a first motion estimation module configured to perform full-pixel motion estimation for a current block in a macroblock using at least one group of reference pictures;

a reference picture selection module configured to select, for the current block, reference pictures from the at least one group of reference pictures based on a cost computed for each reference picture in the at least one group, wherein the cost is based on full-pixel motion estimation;

a second motion estimation module configured to perform sub-pixel motion estimation for the current block using the reference pictures selected by the reference picture selection module; and

a coding module configured to generate a coded current macroblock for the current picture using the sub-pixel motion estimation.

2. The video coding system of claim 1, wherein the cost for the current block is based on a full-pixel motion estimation errors, a full-pixel motion vector for the current block, a predicted motion vector for the current block, and a reference picture index

3. The video coding system of claim 2, wherein

$c=\frac{k}{\mathrm{partition}}$

and the partition is 1 for a block size of 16×

16, the partition is 2 for a block size of 16×

8 or 8×

16, and the partition is 4 for a block size of 8×

8 and below, and wherein k is a constant set to an arbitrary number in the range of 1-32.

4. The video coding system of claim 1, wherein the motion estimation module performs sub-pixel motion estimation for each block using the reference pictures selected for the block.

5. The video coding system of claim 1, wherein the coding module is configured to determine the transform coefficients for the current picture and perform quantization to generate the coded current picture.

6. The video coding system of claim 1, wherein the cost is equal to SAD+c*λ

_motion*R(mv−

pmv)+λ

_motion*R(refidx), where SAD is sum of full-pixel motion estimation errors, mv is a full-pixel motion vector for the current block, pmv is a predicted motion vector (pmv) for the current block, λ

_motion is a constant associated with quantization parameters and picture type, R(mv−

pmv) is a function for coding bits associated with a difference of a particular choice of the full-pixel motion vector (mv) and the pmv, R(refldx) is a function for coding bits associated with reference picture, and c is a constant allowing different weighting to R(mv−

pmv) and R(refldx).

7. The video coding system of claim 1, wherein the at least one group of reference pictures comprises at least one of a group of reference pictures in list L(0) and a group of reference pictures in list L(1).

8. A video coding method comprising:

identifying at least one group of reference pictures to code a current P or B picture;

for each reference picture in the at least one group of reference pictures,

determining a predicted motion vector for a current block of a current macoblock in the current P or B picture; and

determining a cost of a full pixel motion vector using the predicted motion vector; and

selecting at least one reference picture for the current block if in the current P picture and at least two reference pictures for the current block if in the current B picture from the at least one group based on the determined cost for each reference picture in the at least one group of reference pictures.

9. The method of claim 8, further comprising:

using the at least one or the at least two selected reference pictures for sub-pixel motion estimation for the current block in the current picture; and

generating an encoded current picture using the sub-pixel motion estimation.

10. The method of claim 8, wherein selecting at least one reference picture for the current block if in the current P picture comprises:

selecting the at least one reference picture having the lowest cost.

11. The method of claim 8, wherein selecting at least two reference pictures for the current block if in the current B picture comprises:

selecting the at least two reference pictures having the two lowest costs.

12. The method of claim 11, wherein the block in terms of pixels is one of 16×

16, 16×

8, 8×

16, 8×

8, 8×

4, 4×

8 and 4×

4.

13. The method of claim 8, wherein the cost is based on a full-pixel motion estimation errors, a full-pixel motion vector for the current block, a predicted motion vector for the current block, and a reference picture index.

14. The method of claim 8, wherein the at least one group of reference pictures comprises at least one of a group of reference pictures in list L(0) and a group of reference pictures in list L(1).

15. A computer readable medium upon which is embedded programmed instructions which when executed by a processor or other processing circuitry perform a video coding method comprising:

identifying at least one group of reference pictures to code a current P or B picture;

for each reference picture in the at least one group of reference pictures,

determining a predicted motion vector for a current block of a current macoblock in the current P or B picture; and

determining a cost of a full pixel motion vector using the predicted motion vector; and

selecting at least one reference picture for the current block if in the current P picture and at least two reference pictures for the current block if in the current B picture from the at least one group based on the determined cost for each reference picture in the at least one group of reference pictures.

16. The computer readable medium of claim 15, wherein the method further comprises:

using the at least one or the at least two selected reference pictures for sub-pixel motion estimation for the current block in the current picture; and

generating an encoded current picture using the sub-pixel motion estimation.

17. The computer readable medium of claim 15, wherein selecting at least one reference picture for the current block if in the current P picture comprises:

selecting the at least one reference picture having the lowest cost.

18. The computer readable medium of claim 15, wherein selecting at least two reference pictures for the current block if in the current B picture comprises:

selecting the at least two reference pictures having the two lowest costs.

19. The computer readable medium of claim 18, wherein the block partition in terms of pixels is one of 16×

16, 16×

8, 8×

16, 8×

8, 8×

4, 4×

8 and 4'"'"'4.

20. The computer readable medium of claim 15, wherein the cost is based on a full-pixel motion estimation errors, a full-pixel motion vector for the current block, a predicted motion vector for the current block, and a reference picture index.