IP Library Granted Patent US 8,885,730
Granted Patent B2
US 8,885,730 · App. 13/377,160 · Granted Nov 11, 2014

Image coding method, image decoding method, and apparatuses therefor

Inventors: Matthias Narroschke (Rodgau-Dudenhofen, DE); Steffen Wittmann (Moerfelden-Walldorf, DE)
Assignee: Panasonic Corporation
H04N19/00963H04N19/00909H04N19/00884H04N19/00781H04N19/00545H04N19/00272H04N19/00278H04N19/00569H04N19/00066
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 8,885,730
App. No.
13/377,160
Granted
Nov 11, 2014
Kind
B2
Abstract

The image coding method for improving the filtering quality and suppressing increase in the transmission bandwidth includes: selecting filter data elements of the first and second component filters from a filter memory ( 1330 ), and obtaining indexes stored in association with the selected filter data elements in the filter memory ( 1330 ) (S 10 , S 11 ); coding a macroblock using the two filter data elements (S 12 ); adding the two indexes to a coded video signal (S 13 ); calculating the filter data elements of the first and second component filters based on a slice, and assigning indexes to the filter data elements (S 20 , S 21 ); storing, in the filter memory ( 1330 ), a corresponding one of the calculated filter data elements and the index in association with each other for each of the component filters (S 22 ); and adding the two calculated filter data elements to the coded video signal (S 23 ).

Claims (80)

1. An image coding method of coding a video signal to generate a coded video signal that is a bitstream, said method comprising:

selecting, from at least one filter data element stored in a memory, respective filter data elements of one-dimensional component filters of two types that compose a two-dimensional separable filter as two application filter data elements, the one-dimensional component filters being to be applied to a filter-application image area indicated by the video signal;

obtaining an index stored in the memory in association with a corresponding one of the selected two application filter data elements, as an application index, for each of the one-dimensional component filters;

coding the filter-application image area by applying, to the filter-application image area, the one-dimensional component filters of two types using the two application filter data elements selected in said selecting;

adding the two application indexes obtained in said obtaining to the coded filter-application image area in the coded video signal;

identifying filter data elements of the one-dimensional component filters as two designed filter data elements, based on a filter-design image area including the filter-application image area and indicated by the video signal;

assigning an index for identifying a corresponding one of the designed filter data elements to the corresponding one of the designed filter data elements as a designed index for each of the one-dimensional component filters;

storing, in the memory, the designed filter data element and the designed index in association with each other, for each of the one-dimensional component filters, the designed index being assigned to the designed filter data element; and

adding the two designed filter data elements identified in said identifying of filter data elements, to the coded filter-design image area in the coded video signal.

2. The image coding method according to claim 1 ,

wherein in said selecting, after the two designed filter data elements are stored in the memory in said storing, the two application filter data elements are selected from among the filter data elements including the two designed filter data elements.

3. The image coding method according to claim 1 ,

wherein in said selecting, before the two designed filter data elements are stored in the memory in said storing, the two application filter data elements are selected from the at least one filter data element excluding the two designed filter data elements.

4. The image coding method according to claim 1 , further comprising

identifying other two filter data elements of the one-dimensional component filters of two types, based on an other filter-design image area that precedes the filter-design image area in coding order, the other filter-design image area being indicated by the video signal,

wherein in said selecting, the two application filter data elements are selected from among the filter data elements including the two filter data elements identified in said identifying of other two filter data elements.

5. The image coding method according to claim 1 , further comprising

predicting one of the two application indexes of the one-dimensional component filters of two types,

wherein when a prediction index obtained in said predicting is identical to the one of the two application indexes, in said adding of the two application indexes, prediction information indicating that the one of the two application indexes is identical to an index indicating a result of the prediction is added to the coded video signal, instead of the one of the two application indexes.

6. The image coding method according to claim 5 ,

wherein in said predicting, the one of the two application indexes is predicted by selecting any one of indexes of the one-dimensional component filters of two types as the prediction index, the indexes including two indexes obtained from each of surrounding filter-application image areas that surround the filter-application image area and that are included in the filter-design image area.

7. The image coding method according to claim 6 , further comprising

estimating a motion vector for each of the filter-application image area and the surrounding filter-application image areas,

wherein in said predicting, a surrounding filter-application image area from which a motion vector most similar to the motion vector estimated for the filter-application image area is estimated is identified from among the surrounding filter-application image areas, and

one of the two indexes of the one-dimensional component filters of two types is selected as the prediction index, the one of the two application indexes being obtained for the identified surrounding filter-application image area.

8. The image coding method according to claim 7 ,

wherein in said predicting, an index of a one-dimensional component filter of a type identical to a type of the one of the two application indexes for the one-dimensional component filters is selected as the prediction index from among the two indexes of the one-dimensional component filters of two types.

9. The image coding method according to claim 5 ,

wherein in said predicting, the one of the two application indexes is predicted using the other of the two application indexes of the one-dimensional component filters of two types, the other of the two application indexes being obtained for the filter-application image area.

10. The image coding method according to claim 1 ,

wherein when the video signal indicates filter-design image areas including the filter-design image area, said identifying of filter data elements, said assigning, and said storing are repeated for each of the filter-design image areas, and

in said assigning, the two designed indexes are assigned to each of the filter-design image areas so that values indicated by the two designed indexes for each of the filter-design image areas monotonically increase or decrease, according to an order of storing the two designed indexes for each of the filter-design image areas in the memory in said storing or an order of displaying the filter-design image areas.

11. The image coding method according to claim 1 ,

wherein the filter-design image area is a slice, and the filter-application image area is a block, and

in said adding of the two designed filter data elements, the two designed filter data elements are added to a header of the coded slice.

12. The image coding method according to claim 1 ,

wherein the one-dimensional component filters of two types are vertical and horizontal component filters, and

in said identifying of filter data elements, the two designed filter data elements are identified so that an information amount of the designed filter data element of the horizontal component filter is larger than an information amount of the designed filter data element of the vertical component filter.

13. An image decoding method of decoding a coded video signal that is a bitstream, said method comprising:

extracting two indexes of one-dimensional component filters of two types that compose a two-dimensional separable filter from a filter-application image area indicated by the coded video signal, as two application indexes;

obtaining respective filter data elements of the one-dimensional component filters of two types from at least one filter data element stored in a memory as two application filter data elements, the respective filter data elements being stored in association with the two application indexes extracted in said extracting of two indexes;

decoding the filter-application image area by applying, to the filter-application image area, the one-dimensional component filters of two types using the two application filter data elements obtained in said obtaining;

extracting filter data elements of the one-dimensional component filters as designed filter data elements, from a filter-design image area including the filter-application image area and indicated by the video signal;

assigning an index for identifying a corresponding one of the designed filter data elements to the corresponding one of the designed filter data elements as a designed index, for each of the one-dimensional component filters; and

storing the designed filter data element and the designed index in association with each other, in the memory for each of the one-dimensional component filters, the designed index being assigned to the designed filter data element.

14. The image decoding method according to claim 13 ,

wherein when prediction information is extracted in said extracting of two indexes, instead of one of the two application indexes of the one-dimensional component filters of two types:

said method further comprises predicting the one of the two application indexes; and

in said obtaining, a prediction index obtained in said predicting is used as the one of the two application indexes.

15. The image decoding method according to claim 14 ,

wherein in said predicting, the one of the two application indexes is predicted by selecting any one of indexes of the one-dimensional component filters of two types as the prediction index, the indexes including two indexes extracted from each of surrounding filter-application image areas that surround the filter-application image area and that are included in the filter-design image area.

16. The image decoding method according to claim 15 , further comprising

extracting, from the coded video signal, a motion vector for each of the filter-application image area and the surrounding filter-application image areas,

wherein in said predicting, a surrounding filter-application image area from which a motion vector most similar to the motion vector extracted for the filter-application image area is extracted is identified from among the surrounding filter-application image areas, and

the one of the two indexes of the one-dimensional component filters of two types is selected as the prediction index, the one of the two application indexes being extracted from the identified surrounding filter-application image area.

17. An image coding method of coding a video signal to generate a coded video signal that is a bitstream, said method comprising:

selecting, from at least one filter data element stored in a memory, respective filter data elements of one-dimensional component filters of two types that compose a two-dimensional separable filter as two application filter data elements, the one-dimensional component filters being to be applied to a filter-application image area indicated by the video signal;

obtaining an index stored in the memory in association with a corresponding one of the selected two application filter data elements, as an application index, for each of the one-dimensional component filters;

coding the filter-application image area by applying, to the filter-application image area, the one-dimensional component filters of two types using the two application filter data elements selected in said selecting; and

adding the two application indexes obtained in said obtaining to the coded filter-application image area in the coded video signal.

18. An image decoding method of decoding a coded video signal that is a bitstream, said method comprising:

extracting two indexes of one-dimensional component filters of two types that compose a two-dimensional separable filter from a filter-application image area indicated by the coded video signal, as two application indexes;

obtaining respective filter data elements of the one-dimensional component filters of two types from at least one filter data element stored in a memory as two application filter data elements, the respective filter data elements being stored in association with the two application indexes extracted in said extracting of two indexes; and

decoding the filter-application image area by applying, to the filter-application image area, the one-dimensional component filters of two types using the two application filter data elements obtained in said obtaining.

19. An image coding apparatus that codes a video signal to generate a coded video signal that is a bitstream, said apparatus comprising:

a filter data selecting unit configured to select, from at least one filter data element stored in a memory, respective filter data elements of one-dimensional component filters of two types that compose a two-dimensional separable filter as two application filter data elements, the one-dimensional component filters being to be applied to a filter-application image area indicated by the video signal;

an index obtaining unit configured to obtain an index stored in the memory in association with a corresponding one of the selected two application filter data elements, as an application index, for each of the one-dimensional component filters;

a coding unit configured to code the filter-application image area by applying, to the filter-application image area, the one-dimensional component filters of two types using the two application filter data elements selected by said filter data selecting unit;

an index adding unit configured to add the two application indexes obtained by said index obtaining unit to the coded filter-application image area in the coded video signal;

a filter data identifying unit configured to identify filter data elements of the one-dimensional component filters as two designed filter data elements, based on a filter-design image area including the filter-application image area and indicated by the video signal;

an index assigning unit configured to assign an index for identifying a corresponding one of the designed filter data elements to the corresponding one of the designed filter data elements as a designed index for each of the one-dimensional component filters;

a storage processing unit configured to store, in the memory, the designed filter data element and the designed index in association with each other, for each of the one-dimensional component filters, the designed index being assigned to the designed filter data element; and

a filter data adding unit configured to add the two designed filter data elements identified by said filter data identifying unit, to the coded filter-design image area in the coded video signal.

20. An image decoding apparatus that decodes a coded video signal that is a bitstream, said apparatus comprising:

an index extracting unit configured to extract two indexes of one-dimensional component filters of two types that compose a two-dimensional separable filter from a filter-application image area indicated by the coded video signal, as two application indexes;

a filter data obtaining unit configured to obtain respective filter data elements of the one-dimensional component filters of two types from at least one filter data element stored in a memory as two application filter data elements, the respective filter data elements being stored in association with the two application indexes extracted by said index extracting unit;

a decoding unit configured to decode the filter-application image area by applying, to the filter-application image area, the one-dimensional component filters of two types using the two application filter data elements obtained by said filter data obtaining unit;

a filter data extracting unit configured to extract filter data elements of the one-dimensional component filters as designed filter data elements, from a filter-design image area including the filter-application image area and indicated by the video signal;

an index assigning unit configured to assign an index for identifying a corresponding one of the designed filter data elements to the corresponding one of the designed filter data elements as a designed index, for each of the one-dimensional component filters; and

a storage processing unit configured to store the designed filter data element and the designed index in association with each other, in the memory for each of the one-dimensional component filters, the designed index being assigned to the designed filter data element.

Assignments (2)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Oct 31, 2018
From: PANASONIC CORPORATION
To: SOVEREIGN PEAK VENTURES, LLC
Reel/Frame 047914/0675 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Feb 23, 2012
From: NARROSCHKE, MATTHIAS; WITTMANN, STEFFEN
To: PANASONIC CORPORATION
Reel/Frame 027746/0383 →
Priority Claims (1)
EP 09162456 · Jun 10, 2009 · regional
Continuity (1)
Related Publication 20120082242A1 · Apr 5, 2012