Method and apparatus of adaptive loop filtering

Patent metadata

Field	Value
Publication number	US-9094658-B2
Application number	US-201113093068-A
Country	US
Kind code	B2
Filing date	Apr 25, 2011
Priority date	May 10, 2010
Publication date	Jul 28, 2015
Grant date	Jul 28, 2015

Abstract

In a block-based motion compensated system, coding noises may arise due to lossy operations such as quantization. Deblocking is often used to alleviate visibility of coding artifacts by applying filtering across block boundaries adaptively. In the recent HEVC development, adaptive loop filtering (ALF) is being adopted to process deblocked reconstruction frames. Adaptive loop filtering is applied to reconstructed video data across multiple frames to reduce visibility of coding artifacts in the temporal domain. In a HEVC system, conventional ALF can be selectively turned ON or OFF for each block in a frame or a slice. The advanced ALF disclosed here allows a choice of multiple filter sets that can be applied to the reconstructed video data adaptively. Furthermore, the advanced ALF adopts a new unit for the filtering process to apply so that more flexible and/or localized processing can be provided. Furthermore, the advanced ALF allows a decoder to derive respective information for applying the filtering process without side information.

First claim

The invention claimed is:

1. A method for processing video using one adaptive processing selected from adaptive loop filter (ALF), adaptive offset (AO) and adaptive clipping (AC), wherein a picture is partitioned into coding units for applying coding process to each coding unit, the method comprising:
receiving input data to said one adaptive processing selected from the ALF, the AO and the AC, wherein the input data corresponds to reconstructed or reconstructed-and-deblocked coding units of the picture;
dividing the input data into multiple filter units, wherein at least one filter unit includes two or more boundary-aligned reconstructed or reconstructed-and-deblocked coding units;
selecting a candidate filter from a candidate filter set for said one adaptive processing selected from the ALF, the AO and the AC, wherein the candidate filter set consists of at least two candidate filters for the ALF or at least two candidate filters for the AO and the AC;
applying said one adaptive processing selected from the ALF, the AO and the AC using the candidate filter to each filter unit to generate each processed filter unit by using the candidate filter selected to all boundary-aligned reconstructed or reconstructed-and-deblocked coding units in each filter unit; and
providing processed video data consisting of the processed filter units.

2. The method of claim 1 , wherein boundaries of the filter units are boundaries of the coding units and each of said filter units contains at least one coding unit.

3. The method of claim 1 , wherein two or more of the filter units are merged as indicated by a merge index to share the at least one filter.

4. The method of claim 3 , wherein said two or more of the filter units are spatially neighboring filter units, wherein said spatially neighboring filter units are formed by path scanning through the filter units using a scan pattern selected from a group consisting of horizontal scan, vertical scan, z-scan, snake scan, Hilbert scan, a pre-defined scan pattern, and a user-defined scan pattern.

5. The method of claim 4 , wherein information associated with the scan pattern, a choice of whether to allow use of the merge index, or both are incorporated in a sequence level, a picture level, a slice level, a coding unit level, a filter unit level or a filter control unit level.

6. The method of claim 1 , wherein each of the filter units is further partitioned into third blocks corresponding to filter control units, a filter control flag is associated with each of the filter control units to select one filter from said at least one filter of the filter unit selected for applying adaptive loop filtering, adaptive offset or adaptive clipping.

7. The method of claim 6 , wherein each of filter units is partitioned into filter control units using quadtree partitioning, bock partitioning, prediction unit synchronized partitioning, or transform unit synchronized partitioning.

8. The method of claim 6 , wherein the filter control flag is used to indicate filter ON/OFF control when one filter is selected for the filter unit or the filter control flag is used to indicate one of multiple filters when more than one filter are selected for the filter unit.

9. The method of claim 1 , wherein the input data corresponds reconstructed or constructed-and-deblocked luma component and reconstructed or constructed-and-deblocked chroma component, and the reconstructed or constructed-and-deblocked luma component and the reconstructed or constructed-and-deblocked chroma component share filter information associated with said one adaptive processing selected from the ALF, the AO and the AC, and the filter information corresponds to filter control unit partitioning, filter selection, filter control flag, filter shape, filter coefficients, or a combination of the above.

10. The method of claim 9 , wherein sharing the filter information associated with said one adaptive processing selected from the ALF, the AO and the AC by the reconstructed or constructed-and-deblocked luma component and the chroma component is enabled or disable according to a sharing switch.

11. The method of claim 10 , wherein information associated with the sharing switch is incorporated in a sequence level, a picture level, a slice level, a coding unit level, a filter unit level or a filter control unit level.

12. The method of claim 10 , wherein the information associated with adaptive loop filtering, adaptive offset or adaptive clipping for the chroma component is derived from the information associated with respective adaptive loop filtering, adaptive offset or adaptive clipping for the luma component.

13. The method of claim 1 , wherein filter information associated with said one adaptive processing selected from the ALF, the AO and the AC is explicitly incorporated in a bitstream or is derived implicitly based on the input data according to a selection method.

14. The method of claim 13 , wherein the selection method is indicated by a syntax element in a sequence level, a picture level, a slice level, a coding unit level, a filter unit level or a filter control unit level.

15. The method of claim 13 , wherein the input data is classified into multiple categories using classification, and one filter is selected from the candidate filter set for each of the categories , the classification is based on a first characteristic derived from the input data, wherein the first characteristic is selected from a first group consisting of pixel intensity, edge activity, edge orientation, edge intensity, mode information, quantization parameter, residual energy, regional feature, motion information, and a combination of the above.

16. The method of claim 15 , wherein more than one first characteristics are adaptively selected for the classification according to a classifier indicator, wherein the classifier indicator is incorporated in a sequence level, a picture level, a slice level, a coding unit level, a filter unit level or a filter control unit level.

17. The method of claim 15 , wherein the multiple categories are further classified using a second characteristic derived from the input data, wherein the second characteristic is selected from a second group consisting of pixel intensity, edge activity, edge orientation, edge intensity, mode information, quantization parameter, residual energy, regional feature, motion information, and a combination of the above, and wherein the second characteristic is different from the first characteristic.

18. The method of claim 15 , wherein the regional feature is derived according to the characteristics for the filter unit, the coding unit or a pixel location.

19. A method for processing video using one adaptive processing selected from adaptive loop filter (ALF), adaptive offset (AO) and adaptive clipping (AC), wherein a picture is partitioned into coding units for applying coding process to each coding unit, the method comprising:
receiving input data to said one adaptive processing selected from the ALF, the AO and the AC, wherein the input data corresponds to reconstructed or reconstructed-and-deblocked coding units of the picture;
dividing the input data into multiple filter units;
selecting a candidate filter from a candidate filter set for said one adaptive processing selected from the ALF, the AO and the AC;
providing options consisting of signaling selection information associated with the candidate filter set from the candidate filter set for said one adaptive processing selected from the ALF, the AO and the AC explicitly in a bitstrea

Assignees

• Fu Chih-Ming
• Chen Ching-Yeh
• Tsai Chia-Yang
• Huang Yu-Wen
• Lei Shaw-Min
• Mediatek Inc

Inventors

• Fu Chih-Ming
• Chen Ching-Yeh
• Tsai Chia-Yang
• Huang Yu-Wen
• Lei Shaw-Min

Classifications

• H04N19/86

  involving reduction of coding artifacts, e.g. of blockiness · CPC title

• H04N19/10

  using adaptive coding · CPC title

• H04N19/117 Primary

  Filters, e.g. for pre-processing or post-processing (sub-band filter banks H04N19/635) · CPC title

• H04N19/119

  Adaptive subdivision aspects, e.g. subdivision of a picture into rectangular or non-rectangular coding blocks · CPC title

• H04N19/439

  using cascaded computational arrangements for performing a single operation, e.g. filtering · CPC title