US-12627796-B2 - Adaptive cross-component sample offset filtering parameters

Abstract

This disclosure describes a set of advanced video coding technologies and is particular related to cross-component sample offset (CCSO) filtering of reconstructed samples. For example, CCSO filter parameters may be allowed to vary from a filtering unit to another filtering unit. A filtering unit may be a filtering block or a filtering region containing multiple spatially adjacent filtering blocks in a reconstructed frame. Corresponding allowed options of CCSO filter parameter combinations may be specified. Each filtering unit may select from the allowed combination options. The selected combination options may be signaled in the bitstream or derived in some other manners. Allowing such CCSO filter parameter variation within a frame may provide filtering quality gain that outweighs a cost of overhead in signaling.

Inventors

• Xin Zhao
• Samruddhi Yashwant KAHU
• Shan Liu

Assignees

• Tencent America LLC

Dates

Publication Date

20260512

Application Date

20230913

Claims (20)

1. 1 . A method for in-loop filtering of a video bitstream, comprising: reconstructing a frame from the video bitstream to generate reconstructed samples of at least a first color component and a second color component; selecting, for a filtering unit at a level lower than the reconstructed frame, a combination of cross-component sample offset (CCSO) filtering parameters among a plurality of predefined candidate combinations of CCSO filtering parameters for the filtering unit in the reconstructed frame, each combination of the plurality of predefined candidate combinations of CCSO filtering parameters comprises a plurality of filtering parameters with specific parameter values; determining a CCSO filter according to the selected combination of CCSO filtering parameters; and applying the CCSO filter to the reconstructed samples of the first color component in the filtering unit to generate offsets to apply to the reconstructed samples of the second color component in the filtering unit.
2. 2 . The method of claim 1 , wherein each of the plurality of predefined candidate combinations of CCSO filtering parameters comprise: a filter shape parameter indicating a number of CCSO filter taps and tap positions; a quantization step size for discretizing cross component tap-to-sample deltas; and a number of bands indicating a number of sample amplitude bands within an allowed range of sample values.
3. 3 . The method of claim 2 , wherein the plurality of predefined candidate combinations of CCSO filtering parameters consist of four candidate combinations.
4. 4 . The method of claim 3 , wherein the four candidate combinations differ in both the filter shape parameter and the number of bands.
5. 5 . Method of claim 3 , wherein the four candidate combinations share a same filter shape parameter.
6. 6 . The method of claim 3 , wherein the four candidate combinations share a same quantization step size.
7. 7 . The method of claim 2 , wherein the combination of CCSO filtering parameter among the plurality of predefined candidate combinations of CCSO filtering parameters is selected according to a syntax element explicitly signaled in the video bitstream.
8. 8 . The method of claim 2 , wherein the CCSO filter comprises a CCSO lookup table for determining cross component sample offsets for the filtering unit.
9. 9 . The method of claim 8 , wherein each of the plurality of predefined candidate combinations of CCSO filtering parameters is associated with at least one CCSO lookup table.
10. 10 . The method of claim 9 , wherein CCSO lookup table for each of the plurality of predefined candidate combinations of CCSO filtering parameters is signaled in the video bitstream.
11. 11 . The method of claim 2 , wherein the plurality of predefined candidate combinations of CCSO filtering parameters are the same among all filtering units in the reconstructed frame.
12. 12 . The method of claim 2 , wherein the plurality of predefined candidate combinations of CCSO filtering parameters differ between at least two different filtering units in the reconstructed frame.
13. 13 . (v) The method of claim 12 , wherein the plurality of predefined candidate combinations of CCSO filtering parameters are associated with combination set identifiers.
14. 14 . The method of claim 13 , wherein a combination set identifier for the filtering unit to select the combination of CCSO filtering parameters is signaled in the video bitstream.
15. 15 . The method of claim 13 , wherein the filtering unit comprises a filtering block or a filtering region comprising a plurality of spatially adjacent filtering blocks.
16. 16 . The method of claim 2 , wherein values of CCSO filtering parameters for the plurality of predefined candidate combinations of CCSO filtering parameters are signaled in high level syntax in a sequence, frame, slice, or tile header.
17. 17 . An electronic device comprising a memory for storing instructions and a processor for executing the instructions to: reconstruct a frame from a video bitstream to generate reconstructed samples of at least a first color component and a second color component; select, for a filtering unit at a level lower than the reconstructed frame, a combination of cross-component sample offset (CCSO) filtering parameters among a plurality of predefined candidate combinations of CCSO filtering parameters for the filtering unit in the reconstructed frame, each combination of the plurality of predefined candidate combinations of CCSO filtering parameters comprises a plurality of filtering parameters with specific parameter values; determine a CCSO filter according to the selected combination of CCSO filtering parameters; and apply the CCSO filter to the reconstructed samples of the first color component in the filtering unit to generate offsets to apply to the reconstructed samples of the second color component in the filtering unit.
18. 18 . The electronic device of claim 17 , wherein each of the plurality of predefined candidate combinations of CCSO filtering parameters comprise: a filter shape parameter indicating a number of CCSO filter taps and tap positions; a quantization step size for discretizing cross component tap-to-sample deltas; and a number of bands indicating a number of sample amplitude bands within an allowed range of sample values.
19. 19 . The electronic device of claim 17 , wherein the combination of CCSO filtering parameter among the plurality of predefined candidate combinations of CCSO filtering parameters is selected according to a syntax element explicitly signaled in the video bitstream.
20. 20 . A non-transitory computer-readable storage medium for storing computer instructions, wherein the computer instructions, when executed by at least one processor of an electronic device, are configured to cause the electronic device to generate a video bitstream of a video by: reconstructing a frame post encoding to generate reconstructed samples of at least a first color component and a second color component of the frame; selecting, for a filtering unit at a level lower than the reconstructed frame, a combination of cross-component sample offset (CCSO) filtering parameters among a plurality of predefined candidate combinations of CCSO filtering parameters for the filtering unit in the reconstructed frame, each combination of the plurality of predefined candidate combinations of CCSO filtering parameters comprises a plurality of filtering parameters with specific parameter values; determining a CCSO filter according to the selected combination of CCSO filtering parameters; applying the CCSO filter to the reconstructed samples of the first color component in the filtering unit to generate offsets to apply to the reconstructed samples of the second color component in the filtering unit to generate filtered second color component of the filtering unit; using the filtered second components of the filtering unit to encode another portion of the video into the video bitstream; and signaling the selected combination of cross component CCSO filtering parameters for the filtering unit in the video bitstream.

Description

INCORPORATION BY REFERENCE This disclosure is based on and claims the benefit of priority to U.S. Provisional Application No. 63/447,308, entitled “ADAPTIVE CROSS-COMPONENT SAMPLE OFFSET FILTER”, filed on Feb. 21, 2023, which is herein incorporated by reference in its entirety. TECHNICAL FIELD This disclosure generally describes a set of advanced video coding technologies, and is specifically related to cross-component sample offset filtering. BACKGROUND Uncompressed digital video can include a series of pictures, and may specific bitrate requirements for storage, data processing, and for transmission bandwidth in streaming applications. One purpose of video coding and decoding can be the reduction of redundancy in the uncompressed input video signal, through various compression techniques. SUMMARY This disclosure describes a set of advanced video coding technologies and is particular related to cross-component sample offset (CCSO) filtering of reconstructed samples. For example, CCSO filter parameters may be allowed to vary from a filtering unit to another filtering unit. A filtering unit may be a filtering block or a filtering region containing multiple spatially adjacent filtering block in a reconstructed frame. Corresponding allowed options of CCSO filter parameter combinations may be specified. Each filtering unit may select from the allowed options. The selected options may be signaled in the bitstream or derived in some other manners. Allowing such CCSO filter parameter variation within a frame may provide filtering quality gain that outweigh a cost of overhead in signaling. In some example implementations, a method for in-loop filtering of a video bitstream is disclosed. The method may include reconstructing a frame from the video bitstream to generate reconstructed samples of at least a first color component and a second color component; selecting, for a filtering unit at a level lower than the reconstructed frame, a combination of cross-component sample offset (CCSO) filtering parameters among a plurality of candidate combinations of CCSO filtering parameters for the filtering unit in the reconstructed frame; determining a CCSO filter according to the selected combination of CCSO filtering parameters; and applying the CCSO filter to the reconstructed samples of the first color component in the filtering unit to generate offsets to apply to the reconstructed samples of the second color component in the filtering unit. In the example implementation above, each of the plurality of candidate combinations of CCSO filtering parameters may include a filter shape parameter indicating a number of CCSO filter taps and tap positions; a quantization step size for discretizing cross component tap-to-sample deltas; and a number of bands indicating a number of sample amplitude bands within an allowed range of sample values. In any one of the example implementations above, the plurality of candidate combinations of CCSO filtering parameters consist of four candidate combinations. In any one of the example implementations above, the four candidate combinations differ in both the filter shape parameter and the number of bands. In any one of the example implementations above, four candidate combinations share a same filter shape parameter. In any one of the example implementations above, the four candidate combinations share a same quantization step size. In any one of the example implementations above, the combination of CCSO filtering parameter among the plurality of candidate combinations of CCSO filtering parameters is selected according to a syntax element explicitly signaled in the video bitstream. In any one of the example implementations above, the CCSO filter comprises a CCSO lookup table for determining cross component sample offsets for the filtering unit. In any one of the example implementations above, wherein each of the plurality of candidate combinations of CCSO filtering parameters is associated with at least one CCSO lookup table. In any one of the example implementations above, wherein CCSO lookup table for each of the plurality of candidate combinations of CCSO filtering parameters is signaled in the video bitstream. In any one of the example implementations above, wherein the plurality of candidate combinations of CCSO filtering parameters are the same among all filtering units in the reconstructed frame. In any one of the example implementations above, the plurality of candidate combinations of CCSO filtering parameters differ between at least two different filtering units in the reconstructed frame. In any one of the example implementations above, the plurality of candidate combinations of CCSO filtering parameters are associated with combination set identifiers. In any one of the example implementations above, wherein a combination set identifier for the filtering unit to select the combination of CCSO filtering parameters is signaled in the video bitstream. In any one of the example implemen