Inter-prediction candidate index coding independent of inter-prediction candidate list construction in video coding
AssigneeTexas Instruments Incorporated
InventorMinhua Zhou
Life of this patent
Estimating the term…
Abstract
Methods are provided for inter-prediction candidate index coding independent of the construction of the corresponding inter-prediction candidate list, i.e., a merging candidate list or an advanced motion vector predictor list. A maximum allowed number of inter-prediction candidates for an inter-prediction candidate list is used for encoding the inter-prediction candidate index in an encoded bit stream. The maximum allowed number may be pre-determined or may be selected by the encoder and encoded in the bit stream. A decoder may then decode the index using the maximum allowed number of inter-prediction candidates independent of the construction of the corresponding inter-prediction candidate list.
Claims31
A method for encoding a picture, comprising:
- determining a single maximum allowed number of candidates for a plurality of merge candidate lists for all of a plurality of prediction units (PU) for a slice region in the picture;
- binarizing a merge candidate index for one of the plurality of prediction units in the slice region using a truncated unary coding in which a maximum codeword size is one less than the single maximum allowed number of candidates in the at least one slice region;
- encoding the single maximum allowed number of candidates in a slice header for the slice region into a bit stream;
- and
- encoding the binarized merge candidate index into the bit stream.
A system for encoding a picture, comprising:
- a motion estimation component configured to determine a single maximum allowed number of candidates for all of a plurality of merge candidate lists for at least one slice region in the picture;
- a controlling component coupled to the motion estimation component and configured to include the single maximum number of candidates in a slice header for the at least one slice region;
- and
- an entropy encoder component, coupled to the motion estimation component and the controlling component, and configured to:
- binarize a merge candidate index for a prediction unit (PU) in the at least one slice region using a truncated unary coding in which a maximum codeword size is one less than the single maximum allowed number of candidates in the at least one slice region;
- encode the slice header into a bit stream;
- and
- encode the binarized merge candidate index into the bit stream.
A method for decoding a picture from an encoded video bit stream, comprising:
- decoding the encoded video bit stream;
- extracting from the decoded video bit stream a single maximum allowed number of merge candidates for a merge candidate list from a slice header for a slice region comprising a plurality of prediction units (PUs), wherein the single maximum allowed number of candidates for the merge candidate list is the single maximum allowed number of merge candidates for all of the plurality of prediction units;
- determining a merge candidate index for a prediction unit (PU), from the plurality of prediction units, from the decoded video bit stream using the single maximum allowed number of merge candidates in the slice region wherein one less than the single maximum allowed number of merge candidates is used as a maximum codeword size for truncated unary decoding of the merge candidate index;
- constructing a merge candidate list for the PU;
- and
- reconstructing the PU using the merge candidate list for the PU and the merge candidate index for the PU.
A video system for decoding a picture from an encoded video bit stream, comprising:
- an entropy decoding component configured to:
- decode a single maximum allowed number of merge candidates for a merge candidate list from a slice header for a slice region of the picture comprising a plurality of prediction units (PUs), wherein the single maximum allowed number of candidates for the merge candidate list is the single maximum allowed number of merge candidates for all of the plurality of prediction units;
- decode a merge candidate index for a prediction unit (PU) from the plurality of prediction units using the single maximum allowed number of merge candidates in the slice region wherein one less than the maximum allowed number of merge candidates is used as a maximum codeword size for truncated unary decoding of the merge candidate index;
- construct a merge candidate list for the PU;
- an inverse quantization component coupled to the entropy decoding component and configured to de-quantize quantized transform coefficients of residual prediction units decoded from the encoded bit stream;
- an inverse transform component coupled to the inverse quantization component and configured to inverse transform the transform coefficients of residual prediction units from the inverse quantization component from a frequency domain to residual prediction units;
- a motion compensation component coupled to the entropy decoding component and configured to form reconstructed prediction units;
- and
- an adder component coupled to the motion compensation component and the inverse transform component and configured to combine residual prediction units and reconstructed prediction units.
A video system for decoding a picture from an encoded video bit stream, comprising:
- an entropy decoding component configured to:
- decode a single maximum allowed number of merge candidates for a merge candidate list from a slice header for a slice region of the picture comprising a plurality of prediction units (PUs), wherein the single maximum allowed number of candidates for the merge candidate list is the single maximum allowed number of merge candidates for all of the plurality of prediction units;
- decode a merge candidate index for a prediction unit (PU) from the plurality of prediction units using the single maximum allowed number of merge candidates in the slice region wherein one less than the single maximum allowed number of merge candidates is used as a maximum codeword size for truncated unary decoding of the merge candidate index;
- construct a merge candidate list for the PU;
- an inverse quantization component coupled to the entropy decoding component and configured to de-quantize quantized transform coefficients of residual prediction units decoded from the encoded bit stream;
- an inverse transform component coupled to the inverse quantization component and configured to inverse transform the transform coefficients of residual prediction units from the inverse quantization component from a frequency domain to residual prediction units;
- a motion compensation component coupled to the entropy decoding component and configured to form reconstructed prediction units;
- an adder component coupled to the motion compensation component and the inverse transform component and configured to combine residual prediction units and reconstructed prediction units;
- a memory coupled to the adder component and configured to store picture;
- and
- a display coupled to the memory to display the picture.
A video system for decoding a picture from an encoded video bit stream, comprising:
- an entropy decoding component configured to:
- decode a single maximum allowed number of merge candidates for a merge candidate list from a slice header for a slice region of the picture comprising a plurality of prediction units (PUs), wherein the single maximum allowed number of candidates for the merge candidate list is the single maximum allowed number of merge candidates for all of the plurality of prediction units;
- decode a merge candidate index for a prediction unit (PU) from the plurality of prediction units PU using the single maximum allowed number of merge candidates in the slice region wherein one less than the single maximum allowed number of merge candidates is used as a maximum codeword size for truncated unary decoding of the merge candidate index;
- extract from the decoded video bit stream a skip mode flag for the PU;
- extract from the decoded bit stream a merge flag for the PU when skip mode is not indicated for the PU by the extracted skip mode flag;
- construct a merge candidate list for the PU;
- an inverse quantization component coupled to the entropy decoding component and configured to de-quantize quantized transform coefficients of residual prediction units decoded from the encoded bit stream;
- an inverse transform component coupled to the inverse quantization component and configured to inverse transform the transform coefficients of residual prediction units from the inverse quantization component from a frequency domain to residual prediction units;
- a motion compensation component coupled to the entropy decoding component and configured to form reconstructed prediction units;
- and
- an adder component coupled to the motion compensation component and the inverse transform component and configured to combine residual prediction units and reconstructed prediction units.
A video system for decoding a picture from an encoded video bit stream, comprising:
- an entropy decoding component configured to:
- decode a single maximum allowed number of merge candidates for a merge candidate list from a slice header for a slice region of the picture comprising a plurality of prediction units (PUs), wherein the single maximum allowed number of candidates for the merge candidate list is the single maximum allowed number of merge candidates for all of the plurality of prediction units;
- decode a merge candidate index for a prediction unit (PU) from the plurality of prediction units using the single maximum allowed number of merge candidates in the slice region wherein one less than the single maximum allowed number of merge candidates is used as a maximum codeword size for truncated unary decoding of the merge candidate index;
- extract from the decoded video bit stream a skip mode flag for the PU;
- extract from the decoded bit stream a merge flag for the PU when skip mode is not indicated for the PU by the extracted skip mode flag;
- construct a merge candidate list for the PU;
- an inverse quantization component coupled to the entropy decoding component and configured to de-quantize quantized transform coefficients of residual prediction units decoded from the encoded bit stream;
- an inverse transform component coupled to the inverse quantization component and configured to inverse transform the transform coefficients of residual prediction units from the inverse quantization component from a frequency domain to residual prediction units;
- a motion compensation component coupled to the entropy decoding component and configured to form reconstructed prediction units;
- an adder component coupled to the motion compensation component and the inverse transform component and configured to combine residual prediction units and reconstructed prediction units;
- a memory coupled to the adder component and configured to store picture;
- and
- a display coupled to the memory to display the picture.
Dependencies are inferred from the claim text.