Hierarchical motion estimation employing nonlinear scaling and adaptive source block size

Patent metadata

Field	Value
Publication number	US-9532048-B2
Application number	US-201213977392-A
Country	US
Kind code	B2
Filing date	Mar 15, 2012
Priority date	Mar 15, 2012
Publication date	Dec 27, 2016
Grant date	Dec 27, 2016

Abstract

Systems, devices and methods are described including applying nonlinear scaling to a current image frame and a reference image frame to generate at least a corresponding current image layer and a corresponding reference image layer. Hierarchical motion estimation may then be performed using the nonlinearly scaled image layers. Further, source block size may be adaptively determined in a downsampled image layer and hierarchical motion estimation may be performed using the adaptively sized source blocks.

First claim

What is claimed:

1. A motion estimation method, comprising:
receiving, via a video coder, a current image frame and a reference image frame;
applying, via the video coder, nonlinear scaling to the current image frame to generate at least a first current image layer;
applying, via the video coder, nonlinear scaling to the reference image frame to generate at least a first reference image layer; and
performing, via the video coder, hierarchical motion estimation in response, at least in part, to a source block of the first current image layer and to the firs reference image layer.

2. The method of claim 1 , wherein applying nonlinear scaling to the current image frame comprises scaling each dimension of the current image frame by a scaling factor having a value greater than two.

3. The method of claim 1 , wherein the scaling factor comprises a value greater than two and less than four.

4. The method of claim 1 , wherein performing hierarchical motion estimation comprise:
generating a motion vector field corresponding to the source block of the first current image layer; and
identifying a best candidate motion vector in the motion vector field.

5. The method of claim 1 , further comprising:
applying nonlinear scaling to the first current image layer to generate a second current image layer.

6. The method of claim 1 , further comprising:
Applying linear scaling to the first current image layer to generate a second current image layer.

7. The method of claim 1 , wherein applying linear scaling to the first current image layer comprises scaling each dimension of the first current image layer by a factor of two.

8. The method of claim 1 , further comprising
adaptively determining a size of the source block in the nonlinear scaled current image layer; and
wherein performing hierarchical motion estimation further comprises: performing hierarchical motion estimation for a macroblock of the current image frame using, at least in part, the source block and the nonlinear scaled referenced image layer,
wherein adaptively determining the size of the source block comprises determining the size of the source block in response to image characteristics associated with a location of the source block in the nonlinear scaled current image layer.

9. At least one non-transitory machine readable medium comprising a plurality of instructions that in response to being executed on a computing device, cause the computing device to perform operation, comprising:
receive a current image frame and a reference image frame;
applying nonlinear scaling to the current image frame to generate a t least a first current image layer;
applying nonlinear scaling to the reference image frame to generate at least a first reference image layer; and
performing hierarchical motion estimation in response, at least in part, to a source block of the first current image layer and to the first reference image layer.

10. The non-transitory machine readable medium of claim 9 , further comprising:
adaptively determining a size of the source block in the nonlinear scaled current image layer; and
wherein performing hierarchical motion estimation further comprises: performing hierarchical motion estimation for a macroblock of the current image frame using, at least in part, the source block and the nonlinear scaled referenced image layer,
wherein adaptively determining the size of the source block comprises determining the size of the source block in response to image characteristics associated with a location of the source block in the nonlinear scaled current image layer.

11. A system for video coding on a computer, comprising:
means for receiving a current image frame and a reference image frame;
means for applying nonlinear scaling to the current image frame to generate at least a first current image layer;
means for applying nonlinear scaling to the reference image frame to generate at least a first reference image layer; and
means for performing hierarchical motion estimation in response, at least in part, to a source block of the first current image layer and to the first reference image layer.

12. The system of claim 11 , wherein applying nonlinear scaling to the current image frame comprises scaling each dimension of the current image frame by a scaling factor having a value greater than two.

13. The system of claim 11 , wherein the scaling factor comprises a value greater than two and less than four.

14. The system of claim 11 , wherein performing hierarchical motion estimation comprises:
means for generating a motion vector field corresponding to the source block of the first current image layer; and
means for identifying a best candidate motion vector in the motion vector field.

15. The system of claim 11 , further comprising:
means for applying nonlinear scaling to the first current image layer to generate a second current image layer.

16. The system of claim 11 , further comprising:
means for applying linear scaling to the first current image layer to generate a second current image layer, wherein applying linear scaling to the first current image layer comprises scaling each dimension of the first current image layer by a factor of two.

17. The system of claim 11 , further comprising:
means for adaptively determining a size of the source block in the nonlinear scaled current image layer; and
wherein performing hierarchical motion estimation further comprises: performing hierarchical motion estimation for a macroblock of the current image frame using, at least in part, the source block and the nonlinear scaled referenced image layer,
wherein adaptively determining the size of the source block comprises determining the size of the source block in response to image characteristics associated with a location of the source block in the nonlinear scaled current image layer.

18. A motion estimation method, comprising:
generating a downsampled image layer from an image frame;
adaptively determining a size of a source block in the downsampled image layer; and
performing hierarchical motion estimation for a macroblock of the image frame using, at least in part, the source block and a downsampled referenced image layer,
wherein adaptively determining the size of the source block comprises determining the size of the source block in response to image characteristics associated with a location of the source block in the downsampled image layer.

19. The method of claim 18 , wherein the image characteristics comprise at least one of temporal or spatial image characteristics.

20. The method of claim 19 , wherein adaptively determining the size of the source block comprises determining the source block size in response to preprocessing of the image frame.

21. The method of claim 18 , wherein preprocessing of the image frame comprises determining image complexity of the image frame, wherein adaptively determining the size of the source block comprises determining the source block size in real time in response to motion vector costing.

22. A system for video coding on a computer, comprising:
means for generating a downsampled image layer from an image frame;
means for adaptively determining a size of a source block in the down sampled image layer; and
means for performing hierarchical motion estimation for a macroblock of the image frame using, at least in part, the source block and a downsampled reference image layer,
wherein adaptively determining the size of the source block compr

Assignees

• Tanner Jason D
• Jiang Hong
• Intel Corp

Inventors

• Tanner Jason D
• Jiang Hong

Classifications

• H04N19/0066 Primary

  Electricity · mapped topic

• H04N19/53 Primary

  Multi-resolution motion estimation; Hierarchical motion estimation · CPC title

• H04N19/56 Primary

  Motion estimation with initialisation of the vector search, e.g. estimating a good candidate to initiate a search · CPC title

• G06T7/2073

  Physics · mapped topic

• G06T7/207

  for motion estimation over a hierarchy of resolutions (multi-resolution motion estimation or hierarchical motion estimation for coding, decoding, compressing or decompressing digital video signals H04N19/53) · CPC title