Method and device for processing video data, and method for storing video bit stream
Through the cross-component linear model prediction method, the conversion between chroma blocks and luminance blocks is optimized by using the down-sampled co-located adjacent top luminance samples, which solves the low efficiency problem in the existing technology and improves the video encoding quality and data rate utilization efficiency.
Patent Information
- Application Number
- CN202180023235.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2020-03-29
- Filing Date
- 2021-03-22
- Publication Date
- 2025-09-16
- Estimated Expiration
- 2041-03-22
AI Technical Summary
Existing video coding and decoding technologies suffer from low efficiency and poor encoding quality when processing the conversion between chrominance blocks and luminance blocks. In particular, it is difficult to effectively utilize the differences in the human visual system's sensitivity to color and brightness in video signals with different sampling rates.
A cross-component linear model prediction method is adopted. The parameters of the cross-component linear model are derived by using the downsampled co-located adjacent top luminance samples, and the prediction conversion of the chrominance block is performed based on these parameters. Specific downsampling filters and rules are combined to fill or utilize the available samples to optimize the conversion process between video blocks and bitstreams.
It improves the efficiency of video encoding and decoding and enhances the encoding quality, especially in video signals with different chroma sampling rates, reducing the data rate and maintaining the visual effect.
Smart Images
Figure CN115606177B_ABST
Abstract
Claims
1. A method for processing video data, comprising: For conversion between a chroma block of a video and a bitstream of the video, determining a prediction mode to be applied to the chroma block, wherein in the prediction mode, prediction samples of the chroma block are derived based on co-located luma samples of chroma samples in the chroma block and a cross-component linear model; as well as performing said converting based on said determining, Wherein, when a condition set is satisfied, three rows of adjacent top luminance samples of the co-located luminance block of the chrominance block are used to derive downsampled adjacent top luminance samples, and the downsampled adjacent top luminance samples are used to derive one or more parameters of the cross-component linear model.
2. The method according to claim 1, wherein The condition set is related to at least one of: a color format of the chroma block, a position of the co-located luma block, and availability of the three rows of adjacent top luma samples.
3. The method according to claim 1, wherein The condition set includes that the color format of the chroma block is 4:2:
0.
4. The method according to claim 1, wherein The condition set includes that the co-located luma block is not located at a boundary of a codec tree unit.
5. The method according to claim 1, wherein The condition set includes a value of sps_chroma_vertical_collocated_flag included in the bitstream being equal to 1.
6. The method according to claim 1, wherein The one or more parameters of the cross-component linear model are further derived based on the downsampled neighboring left luma sample, the selected neighboring top chroma sample, and the selected neighboring left chroma sample, and The down-sampled adjacent top luminance sample and the selected adjacent top chrominance sample are obtained before the down-sampled adjacent left luminance sample and the selected adjacent left chrominance sample.
7. The method according to claim 6, wherein: The one or more parameters of the cross-component linear model are derived based on four downsampled neighboring luma samples and four selected neighboring chroma samples, wherein the four downsampled adjacent luma samples include at least one of the downsampled adjacent top luma sample and the downsampled adjacent left luma sample, and The four selected adjacent chroma samples include at least one of the selected adjacent top chroma sample and the selected adjacent left chroma sample.
8. The method according to claim 1, wherein The converting includes encoding the video into the bitstream.
9. The method according to claim 1, wherein: The converting includes decoding the video from the bitstream.
10. An apparatus for processing video data, comprising a processor and non-transitory memory having instructions thereon, wherein the instructions, when executed by the processor, cause the processor to: For conversion between a chroma block of a video and a bitstream of the video, determining a prediction mode to be applied to the chroma block, wherein in the prediction mode, prediction samples of the chroma block are derived based on co-located luma samples of chroma samples in the chroma block and a cross-component linear model; and performing said converting based on said determining, in, When a condition set is met, three rows of neighboring top luma samples of a co-located luma block of the chroma block are used to derive downsampled neighboring top luma samples, and the downsampled neighboring top luma samples are used to derive one or more parameters of the cross-component linear model.
11. The device according to claim 10, wherein The condition set is related to at least one of: a color format of the chroma block, a position of the co-located luma block, and availability of the three rows of adjacent top luma samples.
12. The device according to claim 10, wherein The condition set includes that the color format of the chroma block is 4:2:
0.
13. The device according to claim 10, wherein The condition set includes that the co-located luma block is not located at a boundary of a codec tree unit.
14. The device according to claim 10, wherein The condition set includes a value of sps_chroma_vertical_collocated_flag included in the bitstream being equal to 1.
15. The device according to claim 10, wherein The one or more parameters of the cross-component linear model are further derived based on the downsampled neighboring left luma sample, the selected neighboring top chroma sample, and the selected neighboring left chroma sample, and The down-sampled adjacent top luminance sample and the selected adjacent top chrominance sample are obtained before the down-sampled adjacent left luminance sample and the selected adjacent left chrominance sample.
16. The device according to claim 15, wherein The one or more parameters of the cross-component linear model are derived based on four downsampled neighboring luma samples and four selected neighboring chroma samples, wherein the four downsampled adjacent luma samples include at least one of the downsampled adjacent top luma sample and the downsampled adjacent left luma sample, and The four selected adjacent chroma samples include at least one of the selected adjacent top chroma sample and the selected adjacent left chroma sample.
17. A non-transitory computer-readable storage medium storing instructions that cause a processor to: For conversion between a chroma block of a video and a bitstream of the video, determining a prediction mode to be applied to the chroma block, wherein in the prediction mode, prediction samples of the chroma block are derived based on co-located luma samples of chroma samples in the chroma block and a cross-component linear model; and performing said converting based on said determining, in, When a condition set is met, three rows of neighboring top luma samples of a co-located luma block of the chroma block are used to derive downsampled neighboring top luma samples, and the downsampled neighboring top luma samples are used to derive one or more parameters of the cross-component linear model.
18. The non-transitory computer-readable storage medium of claim 17, wherein: The condition set is related to at least one of: a color format of the chroma block, a position of the co-located luma block, and availability of the three rows of adjacent top luma samples.
19. A method for storing a video bitstream, comprising: Determining a prediction mode to be applied to a chroma block of the video, wherein in the prediction mode, prediction samples of the chroma block are derived based on co-located luma samples of chroma samples in the chroma block and a cross-component linear model; generating the bitstream based on the determination, and storing the bitstream in a non-transitory computer-readable recording medium, Wherein, when a condition set is satisfied, three rows of adjacent top luminance samples of the co-located luminance block of the chrominance block are used to derive downsampled adjacent top luminance samples, and the downsampled adjacent top luminance samples are used to derive one or more parameters of the cross-component linear model.
20. The method according to claim 19, wherein The condition set is related to at least one of: a color format of the chroma block, a position of the co-located luma block, and availability of the three rows of adjacent top luma samples.
Citation Information
Patent Citations
Chroma block prediction method and apparatus
CN110881126A
Size dependent downsampling in cross-component linear model
CN110896480A
CCLM-based intra-prediction method and device
WO2020009357A1