Coding and decoding of transform skipped blocks
AssigneeTelefonaktiebolaget LM Ericsson (publ)
InventorsJacob Ström, Kenneth Andersson
Life of this patent
Estimating the term…
Abstract
Video encoder and decoder and methods therein for transform skipped encoding and decoding of blocks of pixels. The methods comprise a modified use of an existing context model associated with transform coefficients, for deriving contexts for a bitmask indicating which residual values (i.e., not transform coefficients), in a transform skipped residual block, that are greater than one. The methods are applicable to video coding schemes such as HEVC.
Claims12
A method performed by a video decoder, operable according to a video decoding scheme applying residual decoding comprising use of a context model for deriving contexts to be used for decoding flags, in a bitmask, which flags indicate, respectively, whether an absolute value of a non-zero transform coefficients in a transformed residual block is greater than one, where the context model comprises a plurality of states, each state corresponding to a context and a probability, where a procedure for deriving contexts for a number of flags related to transform coefficients is started in a first state, A, and where, after decoding a flag which corresponds to an absolute value of a transform coefficient that is greater than one, the context model goes to a second state, D, which is used for decoding remaining flags in said bitmask,
- the method comprising:
- for a transform-skipped residual block comprising residual values and not transform coefficients, deriving contexts to be used for decoding flags, in a bitmask, that indicate, respectively, whether an absolute value of a non-zero residual value in the transform-skipped block is greater than one, by use of the context model, starting in the second state D;
- and
- decoding the flags in the bitmask, for the transform-skipped residual block, using the derived contexts.
A method performed by a video encoder, operable according to a video encoding scheme applying residual encoding comprising use of a context model for associating contexts to flags in a bitmask, which flags indicate, respectively, whether an absolute value of a non-zero transform coefficients in a transformed residual block is greater than one, where the context model comprises a plurality of states, each state corresponding to a context and a probability, where a procedure for deriving contexts to be associated with a number of flags related to transform coefficients is started in a first state, A, and where, after encoding a flag which corresponds to a transform coefficient greater than one, the context model goes to a second state, D, which is used for encoding remaining flags in said bitmask, the method comprising:
- for a transform-skipped residual block comprising residual values and not transform coefficients, deriving contexts to be associated with flags in a bit map, which flags indicate, respectively, whether an absolute value of a non-zero residual value in the transform-skipped residual block is greater than one, by use of the context model, starting in the second state D;
- and
- encoding the flags in the bit map, for the transform-skipped residual block, using the derived contexts.
A video decoder operable according to a video decoding scheme applying residual decoding comprising use of a context model for deriving contexts to be used for decoding flags in a bitmask, which flags indicate, respectively, whether an absolute value of a non-zero transform coefficients in a transformed residual block is greater than one, where the context model comprises a plurality of states, each state corresponding to a context and a probability, where a procedure for deriving contexts for a number of flags related to transform coefficients is started in a first state, A, and where, after decoding a flag which corresponds to an absolute value of a transform coefficient that is greater than one, the context model goes to a second state, D, which is used for decoding remaining flags in said bitmask;
- the video decoder comprising processing means and a memory comprising instructions, which when executed by the processing means cause the video decoder to:
- for a transform-skipped residual block comprising residual values and not transform coefficients, derive contexts to be used for decoding flags, in a bitmask, which flags indicate, respectively, whether an absolute value of a non-zero residual value in the transform-skipped block is greater than one, by use of the context model, starting in the second state D;
- and
- decode the flags in the bitmask, for the transform-skipped residual block, using the derived contexts.
A video encoder operable according to a video encoding scheme applying residual encoding comprising use of a context model for associating contexts to flags in a bitmask, which flags indicate, respectively, whether an absolute value of a non-zero transform coefficients in a transformed residual block is greater than one, where the context model comprises a plurality of states, each state corresponding to a context and a probability, where a procedure for deriving contexts to be associated with a number of flags related to transform coefficients is started in a first state, A, and where, after encoding a flag which corresponds to a transform coefficient greater than one, the context model goes to a second state, D, which is used for encoding remaining flags in said bitmask,
- the video encoder comprising processing means and a memory comprising instructions, which when executed by the processing means cause the video decoder to:
- for a transform-skipped residual block comprising residual values and not transform coefficients, derive contexts to be associated with flags in a bitmask, which flags indicate, respectively, whether an absolute value of a non-zero residual value in the transform-skipped residual block is greater than one, by use of the context model, starting in the second state D;
- and
- encode the flags in the bit map, for the transform-skipped residual block, using the derived contexts.
Dependencies are inferred from the claim text.