Mode-dependent coefficient scanning and directional transforms for different colour sampling formats
A method of coding 4:2:2 or 4:4:4 video data comprises predicting luminance and/or chrominance samples of an image from other respective reference samples derived from the same image according to a prediction mode associated with a sample to be predicted, the prediction mode being selected for each of a plurality of blocks of samples, from a set of two or more candidate prediction modes; detecting differences between the samples and the respective predicted samples; selecting a frequency-separation transform from two or more candidate frequency separation transforms according to the prediction mode associated with a current block of samples using a mapping between transform and prediction mode, the mapping between different, as between chrominance and luminance samples, for at least the 4:4:4: format; and encoding the detected differences by frequency-separating the differences, using the selected frequency-separation transform.
1. Apparatus for decoding video data in which differences between predicted and original samples are frequency-separated and encoded, comprising:
a predictor configured to predict luminance and chrominance samples of an image from other respective reference samples derived from the same image according to a prediction mode associated with a sample to be predicted, the prediction mode being selected for each of a plurality of blocks of samples, from a set of two or more candidate prediction modes;
a selector configured to select a scan pattern from a set of two or more candidate scan patterns, each scan pattern defining an order of encoding the set of frequency-separated coefficients, in dependence upon the prediction mode for that block of samples using a mapping between scan pattern and prediction mode, the mapping being different, as between chrominance and luminance samples; and
a decoder configured to decode frequency-separated difference data representing a frequency-separated version of data indicative of differences between the samples to be decoded and respective predicted samples, in an order of frequency-separated coefficients according to the selected scan pattern,
wherein the video data comprises 4:2:2 and 4:4:4 video data and the mapping is different for 4:2:2 luminance and 4:2:2 chrominance data,
wherein the mapping is different for 4:2:2 and 4:4:4 video data, and
wherein a filter is used on the other respective reference samples when predicting the luminance and chrominance samples.
2. Apparatus for coding video data in which differences between predicted and original samples are frequency-separated and encoded, comprising:
a predictor configured to predict luminance and chrominance samples of an image from other respective reference samples derived from the same image according to a prediction mode associated with a sample to be predicted, the prediction mode being selected for each of a plurality of blocks of samples, from a set of two or more candidate prediction modes;
a detector configured to detect differences between the samples and the respective predicted samples;
a transform unit configured to frequency-separate the detected differences for a block of samples, using a frequency-separation transform, to generate a corresponding set of frequency-separated coefficients;
a selector configured to select a scan pattern from a set of two or more candidate scan patterns, each scan pattern defining an order of encoding the set of frequency-separated coefficients, in dependence upon the prediction mode for that block of samples; and
an encoder configured to encode the frequency-separated difference data in an order of frequency-separated coefficients according to the selected scan pattern,
wherein the video data comprises 4:2:2 and 4:4:4 video data and a mapping between scan pattern and prediction mode, the mapping being different, as between chrominance and luminance samples,
wherein the video data comprises 4:2:2 and 4:4:4 video data and the mapping is different for 4:2:2 luminance and 4:2:2 chrominance data,
wherein the mapping is different for 4:2:2 and 4:4:4 video data, and
wherein a filter is used on the other respective reference samples when predicting the luminance and chrominance samples.
3. A method of coding video data in which differences between predicted and original samples are frequency-separated and encoded, comprising:
predicting luminance and chrominance samples of an image from other respective reference samples derived from the same image according to a prediction mode associated with a sample to be predicted, the prediction mode being selected for each of a plurality of blocks of samples, from a set of two or more candidate prediction modes;
detecting differences between the samples and the respective predicted samples;
frequency-separating the detected differences for a block of samples, using a frequency-separation transform, to generate a corresponding set of frequency-separated coefficients;
selecting a scan pattern from a set of two or more candidate scan patterns, each scan pattern defining an order of encoding the set of frequency-separated coefficients, in dependence upon the prediction mode for that block of samples using a mapping between scan pattern and prediction mode, the mapping being different, as between chrominance and luminance samples;
providing a mapping between prediction mode and scan pattern; and
encoding the frequency-separated difference data in an order of frequency-separated coefficients according to the selected scan pattern,
wherein the video data comprises 4:2:2 and 4:4:4 video data and the mapping is different for 4:2:2 luminance and 4:2:2 chrominance data,
wherein the mapping is different for 4:2:2 and 4:4:4 video data, and
wherein filtering is used on the other respective reference samples when predicting the luminance and chrominance samples.
4. A method according to claim 3 , in which the prediction mode associated with a block of samples to be predicted indicates a prediction direction defining one or more other respective reference samples from which each sample of that block is to be predicted.
5. A method according to claim 3 , in which the size of a current block of luminance samples is 4×4 or 8×8 samples.
6. A method according to claim 3 , comprising:
selecting the size of the current block of samples from a set of candidate sizes;
applying the step of selecting a scan pattern if the selected block size is one of a predetermined subset of the set of candidate sizes.
7. A method according to claim 3 , comprising:
applying the step of selecting a scan pattern only in respect of luminance samples.
8. A method according to claim 3 , in which the set of candidate scan patterns is different for use in respect of luminance and chrominance samples.
9. A method according to claim 3 , in which the selecting step is configured to select a horizontal scan pattern in respect of a set of predominantly horizontal prediction modes, to select a vertical scan pattern in respect of a set of predominantly vertical prediction modes, and to select a diagonal scan pattern in respect of other prediction modes.
10. A method of decoding video data in which differences between predicted and original samples are frequency-separated and encoded, comprising:
predicting luminance and chrominance samples of an image from other respective reference samples derived from the same image according to a prediction mode associated with a sample to be predicted, the prediction mode being selected for each of a plurality of blocks of samples, from a set of two or more candidate prediction modes;
selecting a scan pattern from a set of two or more candidate scan patterns, each scan pattern defining an order of encoding the set of frequency-separated coefficients, in dependence upon the prediction mode for that block of samples using a mapping between scan pattern and prediction mode, the mapping being different, as between chrominance and luminance samples; and
decoding frequency-separated difference data representing a frequency-separated version of data indicative of differences between the samples to be decoded and respective predicted samples, in an order of frequency-separated coefficients according to the selected scan pattern,
wherein the video data comprises 4:2:2 and 4:4:4 video data and the mapping is different for 4:2:2 luminance and 4:2:2 chrominance data,
wherein the mapping is different for 4:2:2 and 4:4:4 video data, and
wherein filtering is used on the other respective reference samples when predicting the luminance and chrominance samples.
11. A method according to claim 3 , in which the video data is 4:2:2 or 4:4:4 video data.
12. A method according to claim 3 , in which the mapping is different, as between chrominance and luminance samples, for at least the 4:4:4 format.
13. Video storage, transmission, capture or display apparatus comprising apparatus according to claim 1 .
14. Video storage, transmission, capture or display apparatus comprising apparatus according to claim 2 .
15. A non-transitory computer readable medium including computer program instructions, which when executed by a computer causes the computer to perform the method of claim 3 .
16. A non-transitory computer readable medium including computer program instructions, which when executed by a computer causes the computer to perform the method of claim 10 .
17. A video coding method in which luminance and chrominance samples of an image associated with video data are predicted from other respective reference samples derived from the same image according to a prediction mode associated with a sample to be predicted, the chrominance samples representing first and second chrominance components;
the method comprising:
providing a mapping between prediction mode and scan pattern;
selecting, for at least some samples, the same prediction mode for each of the luminance and chrominance components corresponding to an image region; and
selecting either the same prediction mode or different respective prediction modes for each of the luminance and chrominance components corresponding to an image region, the selection being made according to an image sequence, an image, or a region of an image,
wherein the video data comprises 4:2:2 and 4:4:4 video data and the mapping is different for 4:2:2 luminance and 4:2:2 chrominance data,
wherein the mapping is different for 4:2:2 and 4:4:4 video data, and
wherein filtering is used on the other respective reference samples when predicting the luminance and chrominance samples.
18. The apparatus according to claim 1 , wherein the filter is used on 4:4:4 video data.
19. The apparatus according to claim 18 , wherein the filter is a smoothing filter.
20. The apparatus according to claim 2 , wherein the filter is used on 4:4:4 video data.
21. The method according to claim 3 , wherein the filtering is used on 4:4:4 video data.
22. The method according to claim 10 , wherein the filtering is used on 4:4:4 video data.