Method and device for coding an image block, corresponding decoding method and decoding device

Patent metadata

Field	Value
Publication number	US-10250877-B2
Application number	US-201314372021-A
Country	US
Kind code	B2
Filing date	Jan 10, 2013
Priority date	Jan 13, 2012
Publication date	Apr 2, 2019
Grant date	Apr 2, 2019

Abstract

A method is described for reconstructing a current block of a current image from at least one block of a reconstructed reference image at a different size from the size of the current image. The method comprises the steps:
motion compensate the reference block with at least one interpolation filter that depends on the size of said reference image and on the size of said current image, decode, for the current block, a residue block, and reconstruct the current block from the residue block and from the motion compensated reference block.

First claim

The invention claimed is:

1. A decoding method of a binary stream to reconstruct a current block of a current image from a reference block of a reference image reconstructed at a different size from the size of said current image, said reconstructed reference image being stored in a decoded picture buffer comprising:
motion compensating said reference block of said reconstructed reference image by applying a single horizontal filter GFH and a single vertical filter GFv successively on the lines and on the columns of pixels of said reference block,
decoding, for the current block, a residue block, and
reconstructing the current block from said residue block and from said motion compensated reference block,
wherein said single vertical filter GFv applied on a pixel s is such that GFv(s)=MCIFv(SCFv(s)), where MCIFv is a vertical motion compensation interpolation filter and SCFv is a vertical resampling filter, MCIFv and SCFv being applied jointly and wherein said single horizontal filter GFH applied on a pixel u is such that GFH(u)=MCIFH(SCFH(u)), where MCIFH is a horizontal motion compensation interpolation filter and SCFH is a horizontal resampling filter, MCIFH and SCFH being applied jointly and wherein no resampled version of said reconstructed reference image is stored in the decoded picture buffer.

2. The decoding method according to claim 1 , wherein the single horizontal filter and the single vertical filter being defined by filtering coefficients, said decoding method further comprises a decoding step of said filtering coefficients.

3. The decoding method according to claim 1 , wherein said binary stream is a scalable stream comprising a base layer coding low resolution images and at least one enhancement layer coding high resolution images, said current image is one of said high resolution images and said reference image is one of the low resolution images.

4. A coding method of a current block of a current image from a reference block of a reference image reconstructed at a different size from the size of said current image, said reconstructed reference image being stored in a decoded picture buffer comprising:
motion compensating said reference block of said reconstructed reference image by applying a single horizontal filter GFH and a single vertical filter GFv successively on the lines and on the columns of pixels of said reference block,
calculating, for the current block, a residue block from said current block and from said motion compensated reference block, and
coding the residue block in a binary stream,
wherein said single vertical filter GFv applied on a pixel s is such that GFv(s)=MCIFv(SCFv(s)), where MCIFv is a vertical motion compensation interpolation filter and SCFv is a vertical resampling filter, MCIFv and SCFv being applied jointly and wherein said single horizontal filter GFH applied on a pixel u is such that GFH(u)=MCIFH(SCFH(u)), where MCIFH is a horizontal motion compensation interpolation filter and SCFH is a horizontal resampling filter, MCIFH and SCFH being applied jointly and wherein no resampled version of said reconstructed reference image is stored in the decoded picture buffer.

5. The coding method according to claim 4 , wherein the single horizontal filter and the single vertical filter being defined by filtering coefficients, said coding method further comprises a coding step of said filtering coefficients.

6. The coding method according to claim 4 , wherein said binary stream is a scalable stream comprising a base layer coding low resolution images and at least one enhancement layer coding high resolution images, said current image is one of said high resolution images and said reference image is one of the low resolution images.

7. A decoding device comprising at least one circuit configured to:
access, from a decoded picture buffer, a reference image reconstructed at a size different from the size of a current image;
motion compensate a reference block of said reconstructed reference image by applying a single horizontal filter GFH and a single vertical filter GFv successively on the lines and on the columns of pixels of said reference block,
decode, for a current block of said current image, a residue block from a binary stream, and
reconstruct the current block from said residue block and from said motion compensated reference block,
wherein said single vertical filter GFv applied on a pixel s is such that GFv(s)=MCIFv(SCFv(s)), where MCIFv is a vertical motion compensation interpolation filter and SCFv is a vertical resampling filter, MCIFv and SCFv being applied jointly and wherein said single horizontal filter GFH applied on a pixel u is such that GFH(u)=MCIFH(SCFH(u)), where MCIFH is a horizontal motion compensation interpolation filter and SCFH is a horizontal resampling filter, MCIFH and SCFH being applied jointly and wherein no resampled version of said reconstructed reference image is stored in the decoded picture buffer.

8. A coding device comprising at least one circuit configured to:
access, from a decoded picture buffer, a reference image reconstructed at a size different from the size of a current image;
motion compensate a reference block of said reconstructed reference image by applying a single horizontal filter GFH and a single vertical filter GFv successively on the lines and on the columns of pixels of said reference block,
calculate, for a current block of said current image, a residue block from said current block and from said motion compensated reference block, and
code the residue block in a binary stream,
wherein said single vertical filter GFv applied on a pixel s is such that GFv(s)=MCIFv(SCFv(s)), where MCIFv is a vertical motion compensation interpolation filter and SCFv is a vertical resampling filter, MCIFv and SCFv being applied jointly and wherein said single horizontal filter GFH applied on a pixel u is such that GFH(u)=MCIFH(SCFH(u)), where MCIFH is a horizontal motion compensation interpolation filter and SCFH is a horizontal resampling filter, MCIFH and SCFH being applied jointly and wherein no resampled version of said reconstructed reference image is stored in the decoded picture buffer.

9. The decoding device according to claim 7 , wherein the single horizontal filter and the single vertical filter being defined by filtering coefficients, said decoding device further comprises a module configured to decode said filtering coefficients.

10. The decoding device according to claim 7 , wherein said binary stream is a scalable stream comprising a base layer coding low resolution images and at least one enhancement layer coding high resolution images, said current image is one of said high resolution images and said reference image is one of the low resolution images.

11. The coding device according to claim 8 , wherein the single horizontal filter and the single vertical filter being defined by filtering coefficients, said coding device further comprises a module configured to code said filtering coefficients.

12. The coding device according to claim 8 , wherein said binary stream is a scalable stream comprising a base layer coding low resolution images and at least one enhancement layer coding high resolution images, said current image is one of said high resolution images and said reference image is one of the low resolution images.

Assignees

• Thomson Licensing
• Interdigital Madison Patent Holdings

Inventors

• Bordes Philippe
• Andrivon Pierre
• Salmon Philippe

Classifications

• H04N19/82

  involving filtering within a prediction loop · CPC title

• H04N19/146

  Data rate or code amount at the encoder output · CPC title

• H04N19/176

  the region being a block, e.g. a macroblock · CPC title

• H04N19/423

  characterised by memory arrangements (H04N19/433 takes precedence) · CPC title

• H04N19/105

  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 title