Method and apparatus for syntax redundancy removal in palette coding
A method and apparatus for palette coding of a block of video data using a candidate prediction mode list with syntax redundancy removed are disclosed. In one embodiment, whether a redundant prediction mode exists in the candidate prediction mode list for the current samples of the current block is determined based on the candidate prediction mode list and the previous prediction mode associated with the previous samples. If the redundant prediction mode exists in the candidate prediction mode list, the redundant prediction mode is removed from the candidate prediction mode list to generate a reduced candidate prediction mode list. In another embodiment, whether a redundant predictor exists in a candidate predictor list for a current sample of the current block is determined based on a condition related to one or more predictors for the current sample of the current block.
1. A method of video decoding according to at least a palette coding mode, the method comprising:
receiving coded da associated with a current block of a picture;
for a current pixel of the current block, determining whether a last palette prediction mode for a last coded pixel of the current block is a copy-above-run mode that corresponds to a first number of pixels having same values as those of corresponding above neighboring pixels of the first number of pixels;
in response to the last palette prediction mode being determined to be the copy-above-run mode, decoding or deriving a current palette prediction mode for the current pixel different from the copy-above-run mode;
in response to the last palette prediction mode being determined to be not the copy-above-run mode, decoding the current palette prediction mode according to a syntax element from the coded data that is indicative of whether the current palette prediction mode is the copy-above-run mode;
determining a palette index map for the current block according to at least the last palette prediction mode for the last coded pixel and the current palette prediction mode for the current pixel; and
reconstructing pixel values of the current block according to the palette index map.
2. The method of claim 1 , further comprising:
determining whether a palette index of an above pixel of the current pixel and a palette index of a left pixel of the current pixel have a same value;
in response to the palette index of the above pixel and the palette index of the left pixel being determined to have the same value, decoding the current palette prediction mode for the current pixel according to a first list of candidates; and
in response to the palette index of the above pixel and the palette index of the left pixel being determined to have different values, decoding the current palette prediction mode for the current pixel according to a second list of candidates that has more candidates than the first list of candidates.
3. The method of claim 2 , wherein
the first list of candidates includes either a copy-top mode or a copy-left mode,
the second list of candidates includes both the copy-top mode and the copy-left mode,
the copy-top mode corresponds to using the palette index of the above pixel as a palette index of the current pixel, and
the copy-left mode corresponds to using the palette index of the left pixel as the palette index of the current pixel.
4. The method of claim 1 , further comprising:
in response to the current pixel being not in a top row of the current block, decoding the current palette prediction mode for the current pixel according to a third list of candidates that includes a copy-top mode; and
in response to the current pixel being in the top row of the current block, decoding the current palette prediction mode for the current pixel according to a fourth list of candidates that excludes the copy-top mode,
wherein the copy-top mode corresponds to using a palette index of an above pixel of the current pixel as a palette index of the current pixel.
5. The method of claim 4 , wherein
the third list of candidates further includes a copy-top-right mode or a copy-top-left mode,
the fourth list of candidates further excludes the copy-top-right mode and the copy-top-left mode,
the copy-top-right mode corresponds to using a palette index of a top-right pixel of the current pixel as the palette index of the current pixel, and
the copy-top-left mode corresponds to using a palette index of a top-left pixel of the current pixel as the palette index of the current pixel.
6. The method of claim 4 , wherein
the third list of candidates further includes a copy-above-N-row mode,
the fourth list of candidates further excludes the copy-above-N-row mode,
the copy-above-N-row mode corresponds to using a palette index of a pixel that is N row above the current pixel as the palette index of the current pixel, and
N is a positive integer.
7. The method of claim 1 , further comprising:
in response to the current pixel being not in a left column of the current block, decoding the current palette prediction mode for the current pixel according to a fifth list of candidates that includes a copy-left mode,
in response to the current pixel being in the left column of the current block, decoding the current palette prediction mode for the current pixel according to a sixth list of candidates that excludes the copy-left mode,
wherein the copy-left mode corresponds to using a palette index of a left pixel of the current pixel as a palette index of the current pixel.
8. The method of claim 7 , wherein
the fifth list of candidates further includes a copy-top-left mode or a copy-down-left mode,
the sixth list of candidates further excludes the copy-top-left mode and the copy-down-left mode,
the copy-top-left mode corresponds to using a palette index of a top-left pixel of the current pixel as the palette index of the current pixel, and
the copy-down-left mode corresponds to using a palette index of a down-left pixel of the current pixel as the palette index of the current pixel.
9. An apparatus for video decoding according to at least a palette coding mode, the apparatus comprising:
circuitry configured to:
receive coded data associated with a current block of a picture;
for a current pixel of the current block, determine whether a last palette prediction mode for a last coded pixel of the current block is a copy-above-run mode that corresponds to a first number of pixels having same values as those of corresponding above neighboring pixels of the first number of pixels;
in response to the last palette prediction mode being determined to be the copy-above-run mode, decode or derive a current palette prediction mode for the current pixel different from the copy-above-run mode;
in response to the last palette prediction mode being determined to be not the copy-above-run mode, decode the current palette prediction mode according to a syntax element from the coded data that is indicative of whether the current palette prediction mode is the copy-above-run mode;
determine a palette index map for the current block according to at least the last palette prediction mode for the last coded pixel and the current palette prediction mode for the current pixel; and
reconstruct pixel values of the current block according to the palette index map.
10. The apparatus of claim 9 , wherein the circuitry is further configured to:
determine whether a palette index of an above pixel of the current pixel and a palette index of a left pixel of the current pixel have a same value;
in response to the palette index of the above pixel and the palette index of the left pixel being determined to have the same value, decode the current palette prediction mode for the current pixel according to a first list of candidates; and
in response to the palette index of the above pixel and the palette index of the left pixel being determined to have different values, decode the current palette prediction mode for the current pixel according to a second list of candidates that has more candidates than the first list of candidates.
11. The apparatus of claim 10 , wherein
the first list of candidates includes either a copy-top mode or a copy-left mode,
the second list of candidates includes both the copy-top mode and the copy-left mode,
the copy-top mode corresponds to using the palette index of the above pixel as a palette index of the current pixel, and
the copy-left mode corresponds to using the palette index of the left pixel as the palette index of the current pixel.
12. The apparatus of claim 9 , wherein the circuitry is further configured to:
in response to the current pixel being not in a top row of the current block, decode the current palette prediction mode for the current pixel according to a third list of candidates that includes a copy-top mode; and
in response to the current pixel being in the top row of the current block, decode the current palette prediction mode for the current pixel according to a fourth list of candidates that excludes the copy-top mode,
wherein the copy-top mode corresponds to using a palette index of an above pixel of the current pixel as a palette index of the current pixel.
13. The apparatus of claim 12 , wherein
the third list of candidates further includes a copy-top-right mode or a copy-top-left mode,
the fourth list of candidates further excludes the copy-top-right mode and the copy-top-left mode,
the copy-top-right mode corresponds to using a palette index of a top-right pixel of the current pixel as the palette index of the current pixel, and
the copy-top-left mode corresponds to using a palette index of a top-left pixel of the current pixel as the palette index of the current pixel.
14. The apparatus of claim 4 , wherein
the third list of candidates further includes a copy-above-N-row mode,
the fourth list of candidates further excludes the copy-above-N-row mode,
the copy-above-N-row mode corresponds to using a palette index of a pixel that is N row above the current pixel as the palette index of the current pixel, and
N is a positive integer.
15. The apparatus of claim 9 , wherein the circuitry is further configured to:
in response to the current pixel being not in a left column of the current block, decode the current palette prediction mode for the current pixel according to a fifth list of candidates that includes a copy-left mode,
in response to the current pixel being in the left column of the current block, decode the current palette prediction mode for the current pixel according to a sixth list of candidates that excludes the copy-left mode,
wherein the copy-left mode corresponds to using a palette index of a left pixel of the current pixel as a palette index of the current pixel.
16. The apparatus of claim 15 , wherein
the fifth list of candidates further includes a copy-top-left mode or a copy-down-left mode,
the sixth list of candidates further excludes the copy-top-left mode and the copy-down-left mode,
the copy-top-left mode corresponds to using a palette index of a top-left pixel of the current pixel as the palette index of the current pixel, and
the copy-down-left mode corresponds to using a palette index of a down-left pixel of the current pixel as the palette index of the current pixel.