Method and apparatus of simplified luma-based chroma intra prediction

Patent metadata

Field	Value
Publication number	US-9794557-B2
Application number	US-201314380709-A
Country	US
Kind code	B2
Filing date	Jan 24, 2013
Priority date	Apr 16, 2012
Publication date	Oct 17, 2017
Grant date	Oct 17, 2017

Abstract

A method and apparatus for chroma intra prediction is based on reconstructed luma pixels and chroma pixels, where the chroma intra prediction is based on a linear model of derived co-located current luma pixels of the current luma block scaled by a scaling factor. The scaling factor comprises a product term of a division factor and a scaled covariance-like value associated with neighboring reconstructed luma and chroma pixels of a current block. The division factor is related to a first data range divided with rounding by a scaled variance-like value associated with the neighboring reconstructed luma pixels of the current block. The scaled covariance-like value, the first data range, or both of the scaled covariance-like value and the first data range are dependent on the internal bit depth, with which the chroma signal is processed during video coding process, according to an embodiment of the present invention.

First claim

The invention claimed is:

1. A method of chroma intra prediction based on reconstructed luma pixels and chroma pixels, the method comprising:
receiving neighboring reconstructed luma pixels and current reconstructed luma pixels of a current luma block from a media or a processor;
receiving neighboring reconstructed chroma pixels of a current chroma block from the media or the processor, wherein the current chroma block is co-located with the current luma block;
deriving chroma intra prediction for chroma pixels of the current chroma block based on derived co-located current luma pixels of the current luma block scaled by a scaling factor, wherein the scaling factor comprises a product term of a division factor and a scaled covariance value associated with the neighboring reconstructed luma pixels of the current luma block and the neighboring reconstructed chroma pixels of the current chroma block, wherein the division factor is related to a first data range divided with rounding by a scaled variance value associated with the neighboring reconstructed luma pixels of the current luma block; and
providing the chroma intra prediction for encoding or decoding of the chroma pixels of the current chroma block;
wherein said scaled covariance value is derived based on a covariance value associated with the neighboring reconstructed luma pixels of the current luma block and the neighboring reconstructed chroma pixels of the current chroma block, wherein said scaled covariance value corresponds to right-shifting said covariance value by a first number of bits related to said covariance value and the internal bit depth with which chroma signal is processed during video coding process, a1s=a1>>Max(0, Log2(abs(a1))−(BitDepth C −2)), wherein a1s represents said scaled covariance value, a1 represents said covariance value, and BitDepth C represents an internal bit depth with which chroma signal is processed during video coding process.

2. The method of claim 1 , wherein the neighboring reconstructed luma pixels comprise a first group of the reconstructed luma pixels above a top boundary of the current luma block and a second group of the reconstructed luma pixels to a left boundary of the current luma block.

3. The method of claim 1 , wherein said derived co-located current luma pixels of the current luma block are derived from the current reconstructed luma pixels of the current luma block at pixel locations co-located with the chroma pixels of the current chroma block.

4. The method of claim 1 , wherein the first data range corresponds to 2^(BitDepth C −2) or 2 (BitDepth C −1).

5. The method of claim 1 , wherein said scaled variance value is derived based on a variance value associated with the neighboring reconstructed luma pixels of the current luma block, wherein said scaled variance value corresponds to right-shifting said variance value by a second number of bits related to said variance value.

6. The method of claim 5 , wherein a2s=abs(a2>>Max(0, Log 2(abs(a2))−5)), wherein a2s represents said scaled variance value and a2 represents said variance value.

7. The method of claim 6 , wherein the division factor corresponds to (2^(BitDepth C −2)+a2s/2)/a2s, (2^(BitDepth C −1)+a2s/2)/a2s, (2^(BitDepth C +1)+a2s/2)/a2s, (2^(BitDepth C +2)+a2s/2)/a2s, (2^(BitDepth C +3)+a2s/2)/a2s, or (2^(BitDepth C +4)+a2s/2)/a2s.

8. The method of claim 7 , wherein said derived co-located current luma pixels of the current luma block scaled by the scaling factor are right shifted by a shift factor related to the product term during said deriving chroma intra prediction for chroma pixels of the current chroma block, wherein the shift factor corresponds to (BitDepth C −2−Max(0, Log 2(abs(a3))−6)), (BitDepth C −1−Max(0, Log 2(abs(a3))−6)), (BitDepth C +1−Max(0, Log 2(abs(a3))−6)), (BitDepth C +2−Max(0, Log 2(abs(a3))−6)), (BitDepth C +3−Max(0, Log 2(abs(a3))−6)), or (BitDepth C +4−Max(0, Log 2(abs(a3))−6)) for a respective division factor, wherein a3 represents a processed result corresponding to the product term being right shifted with rounding and clipped to (−2 15 , 2 15 −1) if said scaled variance value is equal to or greater than one, and wherein a3 is 0 otherwise.

9. The method of claim 7 , wherein the division factor is implemented by a look-up table for a selected internal bit depth of chroma signal with the scaled variance value as a table input.

10. The method of claim 9 , wherein the look-up table omits table entries for the scaled variance value from 1 to 6 in case of the division factor corresponding to either (2^(BitDepth C +1)+a2s/2)/a2s, (2^(BitDepth C +2)+a2s/2)/a2s and for the scaled variance value from 1 to 31 in case of the division factor corresponding to either (2^(BitDepth C +3)+a2s/2)/a2s, or (2^(BitDepth C +4)+a2s/2)/a2s.

11. The method of claim 10 , wherein a3 is set to 0 if a2s is less than 7 in case of the division factor corresponding to either (2^(BitDepth C +1)+a2s/2)/a2s, (2^(BitDepth C +2)+a2s/2)/a2s and wherein a3 is set to 0 if a2s is less than 32 in case of the division factor corresponding to either (2^(BitDepth C +3)+a2s/2)/a2s, or (2^(BitDepth C +4)+a2s/2)/a2s.

12. The method of claim 10 , wherein the division factor is set to a first value of the division factor corresponding to a2s equal to 7 if a2s is less than 7 or a2s is less than 7 and larger than 0, in case of the division factor corresponding to either (2^(BitDepth C +1)+a2s/2)/a2s, (2^(BitDepth C +2)+a2s/2)/a2s and wherein a3 is set to a second value of the division factor corresponding to a2s equal to 32 if a2s is less than 32 or a2s is less than 32 and larger than 0, in case of the division factor corresponding to either (2^(BitDepth C +3)+a2s/2)/a2s, or (2^(BitDepth C +4)+a2s/2)/a2s.

13. The method of claim 1 , wherein said scaled covariance value is derived based on a covariance value associated with the neighboring reconstructed luma pixels of the current luma block and the neighboring reconstructed chroma pixels of the current chroma block, wherein said scaled covariance value corresponds to right-shifting said covariance value by a first number of bits related to said covariance value and an internal bit depth with which chroma signal is processed during video coding process.

14. The method of claim 13 , wherein a1s=a1>>Max(0, Log 2(abs(a1))−(BitDepth C −2)), wherein a1s represents said scaled covariance value, a1 represents said covariance value, and BitDepth C represents an internal bit depth with which chroma signal is processed during video coding process.

15. The method of claim 14 , wherein the first data range corresponds to 2^(BitDepth C −2) or 2 (BitDepth C −1).

16. An apparatus of chroma intra prediction based on reconstructed luma pixels and chroma pixels, the apparatus comprising:
at least one circuit, the at least one circuit configured for:
receiving neighboring reconstructed luma pixels and current reconstructed luma pixels of a current luma block from a media or a processor;
receiving neighboring reconstructed chroma pixels of a current chroma block from the media or the processor, wherein the current chroma block is co-located with the current luma block;
deriving chroma intra prediction for chroma pixels of the current chroma block based on derived co-located current luma pixels of the current luma block scaled by a scaling factor, wherein the scaling factor comprises a product term of a division factor and a scaled covariance value associated with the neighboring reconstructed luma pixels of the current luma block and the neighboring reconstructed chroma pixels of the current chroma block, wherein the division factor

Assignees

• Mediatek Singapore Pte Ltd
• Mediatek Singapore Pte Ltd

Inventors

• Hsu Juliana
• Guo Mei
• Guo Xun

Classifications

• H04N19/42

  characterised by implementation details or hardware specially adapted for video compression or decompression, e.g. dedicated software implementation (H04N19/635 takes precedence) · CPC title

• H04N19/186

  the unit being a colour or a chrominance component · CPC title

• H04N19/00763 Primary

  Electricity · mapped topic

• H04N19/593 Primary

  involving spatial prediction techniques · CPC title

• H04N19/169 Primary

  characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding · CPC title