Method and apparatus for adaptive transform in video encoding and decoding
For a picture with two or more color components, the prediction residuals for the first color component of a block to be encoded may be transformed with a first transform. The transform coefficients for the first color component may go through quantization, de-quantization and inverse transform to obtain reconstructed prediction residuals. Based on the reconstructed prediction residuals for the first color component, the phases of the transform basis function of the first transform can be adjusted to improve the sparsity of the transformed signal. The prediction residuals for the remaining color components may then be transformed with the adjusted transform. In order to determine the phase shift factor, the reconstructed prediction residuals for the first color component may be transformed with the first transform, adjusted by different candidate phase shift factors, and the candidate phase shift factor that provides a smallest sparsity measure can be selected for the block.
1. A method for video decoding, comprising:
accessing quantized transform coefficients of a first color component and a second color component for a block of a picture;
de-quantizing said quantized transform coefficients of said first and second color components;
obtaining prediction residuals for said first color component of said block, by inverse transforming said de-quantized transform coefficients for said first color component of said block using a first transform, said first transform corresponding to at least a first set of transform basis functions;
determining a phase shift factor based on said obtained prediction residuals for said first color component of said block, by performing:
transforming said prediction residuals for said first color component with a plurality of candidate transforms, wherein each candidate transform is based on said first transform and a candidate phase shift factor of a plurality of candidate phase shift factors,
determining a sparsity based on transform coefficients obtained from a respective one of said plurality of candidate transforms, and
selecting a candidate phase shift factor corresponding to a candidate transform with a smallest sparsity as said phase shift factor;
obtaining prediction residuals for said second color component of said block, by inverse transforming said de-quantized transform coefficients for said second color component of said block using a second transform, said second transform corresponding to at least a second set of transform basis functions, wherein a transform basis function of said first set of transform basis functions is adjusted based on said determined phase shift factor to form a respective transform basis function of said second set of transform basis functions; and
reconstructing said block based on said obtained prediction residuals and a predicted block for said block.
2. The method of claim 1 , wherein said determined phase shift factor is scaled by a subsampling ratio between said first and second color components.
3. A method for video encoding, comprising:
accessing prediction residuals of a first color component and a second color component for a block of a picture;
transforming said prediction residuals for said first color component of said block to form transform coefficients for said first color component of said block, using a first transform, said first transform corresponding to at least a first set of transform basis functions;
determining reconstructed prediction residuals for said first color component of said block based on said transform coefficients for said first color component of said block;
determining a phase shift factor based on said reconstructed prediction residuals for said first color component of said block, by performing:
transforming said reconstructed prediction residuals for said first color component with a plurality of candidate transforms, wherein each candidate transform is based on said first transform and a candidate phase shift factor of a plurality of candidate phase shift factors,
determining a sparsity based on transform coefficients obtained from a respective one of said plurality of candidate transforms, and
selecting a candidate phase shift factor corresponding to a candidate transform with a smallest sparsity as said phase shift factor;
transforming said prediction residuals for said second color component of said block to form transform coefficients for said second color component of said block, using a second transform, said second transform corresponding to at least a second set of transform basis functions, wherein a transform basis function of said first set of transform basis functions is adjusted based on said determined phase shift factor to form a respective transform basis function of said second set of transform basis functions; and
encoding said transform coefficients for said first and second color components of said block.
4. The method of claim 3 , wherein said determined phase shift factor is scaled by a subsampling ratio between said first and second color components.
5. An apparatus for video decoding, comprising:
one or more processors, wherein said one or more processors are configured to:
access quantized transform coefficients of a first color component and a second color component for a block of a picture;
de-quantize said quantized transform coefficients of said first and second color components;
obtain prediction residuals for said first color component of said block, by inverse transforming said de-quantized transform coefficients for said first color component of said block using a first transform, said first transform corresponding to at least a first set of transform basis functions;
determine a phase shift factor based on said obtained prediction residuals for said first color component of said block, by performing:
transforming said prediction residuals for said first color component with a plurality of candidate transforms, wherein each candidate transform is based on said first transform and a candidate phase shift factor of a plurality of candidate phase shift factors,
determining a sparsity based on transform coefficients obtained from a respective one of said plurality of candidate transforms, and
selecting a candidate phase shift factor corresponding to a candidate transform with a smallest sparsity as said phase shift factor;
obtain prediction residuals for said second color component of said block, by inverse transforming said de-quantized transform coefficients for said second color component of said
transform said prediction residuals for said second color component of said block to form transform coefficients for said second color component of said block, using a second transform, said second transform corresponding to at least a second set of transform basis functions, wherein a transform basis function of said first set of transform basis functions is adjusted based on said determined phase shift factor to form a respective transform basis function of said second set of transform basis functions; and
encode said transform coefficients for said first and second color components of said block.
6. The apparatus of claim 5 , wherein said determined phase shift factor is scaled by a subsampling ratio between said first and second color components.
7. An apparatus for video encoding, comprising:
one or more processors, wherein said one or more processors are configured to:
access prediction residuals of a first color component and a second color component for a block of a picture;
transform said prediction residuals for said first color component of said block to form transform coefficients for said first color component of said block, using a first transform, said first transform corresponding to at least a first set of transform basis functions;
determine reconstructed prediction residuals for said first color component of said block based on said transform coefficients for said first color component of said block, by performing:
transforming said reconstructed prediction residuals for said first color component with a plurality of candidate transforms, wherein each candidate transform is based on said first transform and a candidate phase shift factor of a plurality of candidate phase shift factors,
determining a sparsity based on transform coefficients obtained from a respective one of said plurality of candidate transforms, and
selecting a candidate phase shift factor corresponding to a candidate transform with a smallest sparsity as said phase shift factor;
determine a phase shift factor based on said reconstructed prediction residuals for said first color component of said block; block using a second transform, said second transform corresponding to at least a second set of transform basis functions, wherein a transform basis function of said first set of transform basis functions is adjusted based on said determined phase shift factor to form a respective transform basis function of said second set of transform basis functions; and
reconstruct said block based on said obtained prediction residuals and a predicted block for said block.
8. The apparatus of claim 7 , wherein said determined phase shift factor is scaled by a subsampling ratio between said first and second color components.
9. A non-transitory computer readable medium comprising instructions which, when the instructions are executed by a computer, cause the computer to perform the method of claim 1 .
10. A non-transitory computer readable medium comprising instructions which, when the instructions are executed by a computer, cause the computer to perform the method of claim 3 .
11. The method of claim 1 , wherein elements for a transform matrix corresponding to said second transform are represented in integers.
12. The method of claim 3 , wherein elements for a transform matrix corresponding to said second transform are represented in integers.
13. The apparatus of claim 5 , wherein elements for a transform matrix corresponding to said second transform are represented in integers.
14. The apparatus of claim 7 , wherein elements for a transform matrix corresponding to said second transform are represented in integers.