Residual coding with reduced usage of local neighborhood
Different implementations are described, particularly implementations for video encoding and decoding using entropy coding of quantized transform coefficients. The method comprises: determining context model indices for one or more syntax elements of quantized transform coefficients of a transform block, wherein different context model indices are obtained for quantized transform coefficients in different regions of the transform block and a single context index is used for at least one region of the transform block; and selecting for the one or more syntax elements associated with a particular quantized transform coefficient one of several context models based on the determined content model index.
1. A method for encoding a picture, comprising:
splitting a chroma transform block into two regions, wherein a first region comprises one or more lower frequency quantized transform coefficients and a second region comprises a plurality of higher frequency quantized transform coefficients;
determining context model indices for one or more syntax elements of quantized transform coefficients of the chroma transform block, wherein different context model indices are obtained for quantized transform coefficients in different regions of the chroma transform block, and wherein a single context index is used for said first region of the chroma transform block and another single context index is used for said second region of the chroma transform block;
splitting a luma transform block corresponding to the chroma transform block into a plurality of regions, wherein each region of said plurality of regions comprises one or more quantized transform coefficients, and wherein multiple context indices are used for each of said plurality of regions for the luma transform block;
selecting for the one or more syntax elements associated with a particular quantized transform coefficient one of several context models based on the determined content model index, and
binary arithmetic encoding the one or more syntax elements according to the selected context model.
2. The method of claim 1 , wherein in the first region of the chroma transform block a first predefined context index is used and in a second region of the chroma transform block a second predefined context index is used.
3. The method of claim 1 , further comprising:
determining a channel type of the transform block and performing the determination of context model indices and selection of context models depending on the determined channel type.
4. The method of claim 1 , wherein a context model index for a quantized transform coefficient for the luma transform block depends on neighboring higher frequency quantized transform coefficients.
5. A non-transitory machine readable medium having stored thereon machine executable instructions that, when executed, implement an encoding method according to claim 1 .
6. An apparatus for encoding a picture, comprising one or more processors, wherein said one or more processors are configured to:
split a chroma transform block into two regions, wherein a first region comprises one or more lower frequency quantized transform coefficients and a second region comprises a plurality of higher frequency quantized transform coefficients;
determine context model indices for one or more syntax elements of quantized transform coefficients of the chroma transform block, wherein different context model indices are obtained for quantized transform coefficients in different regions of the chroma transform block, and wherein a single context index is used for said first region of the chroma transform block and another single context index is used for said second region of the chroma transform block;
split a luma transform block corresponding to the chroma transform block into a plurality of regions, wherein each region of said plurality of regions comprises one or more quantized transform coefficients, and wherein multiple context indices are used for each of said plurality of regions for the luma transform block;
select for the one or more syntax elements associated with a particular quantized transform coefficient one of several context models based on the determined content model index; and
binary arithmetic encode the one or more syntax elements according to the selected context model.
7. The apparatus of claim 6 , wherein in said first region of the chroma transform block a first predefined context index is used and in a second region of the chroma transform block a second predefined context index is used.
8. The apparatus of claim 6 , wherein said one or more processors are further configured to determine a channel type of the transform block and to perform the determination of context model indices and selection of context models depending on the determined channel type.
9. The apparatus of claim 6 , wherein a context model index for a quantized transform coefficient for the luma transform block depends on neighboring higher frequency quantized transform coefficients.
10. A method for decoding a picture, comprising:
splitting a chroma transform block into two regions, wherein a first region comprises one or more lower frequency quantized transform coefficients and a second region comprises a plurality of higher frequency quantized transform coefficients;
determining context model indices for one or more syntax elements of quantized transform coefficients of the chroma transform block, wherein different context model indices are obtained for quantized transform coefficients in different regions of the chroma transform block, and wherein a single context index is used for said first region of the chroma transform block and another single context index is used for said second region of the chroma transform block;
splitting a luma transform block corresponding to the chroma transform block into a plurality of regions, wherein each region of said plurality of regions comprises one or more quantized transform coefficients, and wherein multiple context indices are used for each of said plurality of regions for the luma transform block;
selecting for the one or more syntax elements associated with a particular quantized transform coefficient one of several context models based on the determined content model index; and
binary arithmetic decoding the one or more syntax elements according to the selected context model.
11. The method of claim 10 , wherein in the first region of the chroma transform block a first predefined context index is used and in a second region of the chroma transform block a second predefined context index is used.
12. The method of claim 10 , further comprising
determining a channel type of the transform block and performing the determination of context model indices and selection of context models depending on the determined channel type.
13. The method of claim 10 , wherein a context model index for a quantized transform coefficient for the luma transform block depends on neighboring higher frequency quantized transform coefficients.
14. A non-transitory machine readable medium having stored thereon machine executable instructions that, when executed, implement a decoding method according to claim 10 .
15. An apparatus for decoding a picture, comprising one or more processors, wherein said one or more processors are configured to:
split a chroma transform block into two regions, wherein a first region comprises one or more lower frequency quantized transform coefficients and a second region comprises a plurality of higher frequency quantized transform coefficients;
determine context model indices for one or more syntax elements of quantized transform coefficients of the chroma transform block, wherein different context model indices are obtained for quantized transform coefficients in different regions of the chroma transform block, and wherein a single context index is used for said first region of the chroma transform block and another single context index is used for said second region of the chroma transform block;
split a luma transform block corresponding to the chroma transform block into a plurality of regions, wherein each region of said plurality of regions comprises one or more quantized transform coefficients, and wherein multiple context indices are used for each of said plurality of regions for the luma transform block;
select for the one or more syntax elements associated with a particular quantized transform coefficient one of several context models based on the determined content model index; and
binary arithmetic decode the one or more syntax elements according to the selected context model.
16. The apparatus of claim 15 , wherein in said first region of the chroma transform block a first predefined context index is used and in a second region of the chroma transform block a second predefined context index is used.
17. The apparatus of claim 15 , wherein said one or more processors are further configured to determine a channel type of the transform block and to perform the determination of context model indices and selection of context models depending on the determined channel type.
18. The apparatus of claim 15 , wherein a context model index for a quantized transform coefficient for the luma transform block depends on neighboring higher frequency quantized transform coefficients.