Video encoding and decoding using multiple transform selection

1. A method, comprising:

obtaining a signal including image information including non-camera-captured image content;

determining a block of spatial-domain values for a prediction residual;

replacing, based on information included in the signal, a first transform in a first set of multiple transforms with a second transform in a second set of multiple transforms for the non-camera-captured image content;

transforming the block of spatial-domain values using the second transform; and

encoding at least a portion of the image information based on the transforming of the block of spatial-domain values,

wherein the second transform comprises one of a Haar transform, a Walsh-Hadamard transform, a Slant transform, or an identity transform.

2. The method of claim 1 , wherein the transforms included in the first set of multiple transforms and the second set of multiple transforms are defined based on one or more index values corresponding to entries in a table of transforms.

3. The method of claim 1 , wherein for different block sizes different transforms having differing suitability for differently correlated residuals are applied and below a predetermined transform size a transform is applied which is more suitable for a highly correlated residual.

4. The method of claim 1 , further comprising determining the second transform for replacing the first transform based on at least one of a shape and/or size of the block and/or a prediction mode.

5. The method of claim 4 , wherein below a predetermined transform size the Slant transform is applied and above the predetermined transform size the Haar transform or the Walsh-Hadamard transform is applied.

6. The method of claim 4 , wherein for a non-square block different transforms are applied for a horizontal dimension and a vertical dimension, the shortest of the horizontal dimension or the vertical dimension having a transform applied which is more suitable for a less correlated residual.

7. The method of claim 6 , wherein the Haar transform or the Walsh-Hadamard transform is applied to the shortest of the horizontal dimension or the vertical dimension.

8. A method, comprising:

obtaining a block of transform coefficients from a signal including encoded image information including non-camera-captured image content;

replacing, based on information included in the signal, a first transform in a first set of multiple transforms with a second transform in a second set of multiple transforms for the non-camera-captured image content;

inverse transforming the block of transform coefficients using said second transform; and

decoding at least a portion of the encoded image information based on an inverse transforming of the block of transform coefficients,

wherein the second transform comprises one of a Haar transform, a Walsh-Hadamard transform, a Slant transform, or an identity transform.

9. The method of claim 8 , wherein the transforms included in the first set of multiple transforms and the second set of multiple transforms are defined based on one or more index values corresponding to entries in a table of transforms.

10. The method of claim 9 , comprising:

obtaining at least one of the one or more index values from a received encoded signal; and

applying one or more transforms included in the second set of multiple transforms corresponding to the one or more index values instead of the first transform.

11. The method of claim 8 , further comprising determining the second transform for replacing the first transform based on at least one of a shape and/or size of the block and/or a prediction mode.

12. The method of claim 11 , wherein for a non-square block different transforms are applied for a horizontal dimension and a vertical dimension, the shortest of the horizontal dimension or the vertical dimension having a transform applied which is more suitable for a less correlated residual.

13. The method of claim 12 , wherein the Haar transform or the Walsh-Hadamard transform is applied to the shortest of the horizontal dimension or the vertical dimension.

14. The method of claim 8 , wherein for different block sizes different transforms having differing suitability for differently correlated residuals are applied and wherein below a predetermined transform size a transform is applied which is more suitable for a highly correlated residual.

15. The method of claim 14 , wherein below the predetermined transform size the Slant transform is applied and above the predetermined transform size the Haar transform or the Walsh-Hadamard transform is applied.

16. The method of claim 8 , wherein the second transform has a basis function including a discontinuity and/or constant increase or decrease.

17. An apparatus, comprising:

one or more processors configured to:

obtain a signal including image information including non-camera-captured image content;

determine a block of spatial-domain values for a prediction residual;

replace, based on information included in the signal, a first transform in a first set of multiple transforms with a second transform in a second set of multiple transforms for the non-camera-captured image content;

transform the block of spatial-domain values using the second transform; and

encode at least a portion of the image information based on the transform of the block of spatial-domain values,

wherein the second transform comprises one of a Haar transform, a Walsh-Hadamard transform, a Slant transform, or an identity transform.

18. The apparatus of claim 17 , wherein the transforms included in the first set of multiple transforms and the second set of multiple transforms are defined based on one or more index values corresponding to entries in a table of transforms.

19. The apparatus of claim 17 , wherein for different block sizes different transforms having differing suitability for differently correlated residuals are applied and below a predetermined transform size a transform is applied which is more suitable for a highly correlated residual.

20. The apparatus of claim 17 , wherein the one or more processors are further configured to determine the second transform for replacing the first transform based on at least one of a shape and/or size of the block and/or a prediction mode.

21. The apparatus of claim 20 , wherein below a predetermined transform size the Slant transform is applied and above the predetermined transform size the Haar transform or the Walsh-Hadamard transform is applied.

22. The apparatus of claim 20 , wherein for a non-square block different transforms are applied for a horizontal dimension and a vertical dimension, the shortest of the horizontal dimension or the vertical dimension having a transform applied which is more suitable for a less correlated residual.

23. The apparatus of claim 22 , wherein the Haar transform or the Walsh-Hadamard transform is applied to the shortest of the horizontal dimension or the vertical dimension.

24. An apparatus, comprising:

one or more processors configured to:

obtain a block of transform coefficients from a signal including encoded image information including non-camera-captured image content;

replace, based on information included in the signal, a first transform in a first set of multiple transforms with a second transform in a second set of multiple transforms for the non-camera-captured image content;

inverse transform the block of transform coefficients using said second transform; and

decode at least a portion of the encoded image information based on an inverse transform of the block of transform coefficients,

wherein the second transform comprises one of a Haar transform, a Walsh-Hadamard transform, a Slant transform, or an identity transform.

25. The apparatus of claim 24 , wherein the transforms included in the first set of multiple transforms and the second set of multiple transforms are defined based on one or more index values corresponding to entries in a table of transforms.

26. The apparatus of claim 25 , wherein the one or more processors are further configured to:

obtain at least one of the one or more index values from a received encoded signal; and

apply one or more transforms included in the second set of multiple transforms corresponding to the one or more index values instead of the first transform.

27. The apparatus of claim 24 , wherein the one or more processors are further configured to:

determine the second transform for replacing the first transform based on at least one of a shape and/or size of the block and/or a prediction mode.

28. The apparatus of claim 27 , wherein for a non-square block different transforms are applied for a horizontal dimension and a vertical dimension, the shortest of the horizontal dimension or the vertical dimension having a transform applied which is more suitable for a less correlated residual.

29. The apparatus of claim 28 , wherein the Haar transform or the Walsh-Hadamard transform is applied to the shortest of the horizontal dimension or the vertical dimension.

30. The apparatus of claim 24 , wherein for different block sizes different transforms having differing suitability for differently correlated residuals are applied and wherein below a predetermined transform size a transform is applied which is more suitable for a highly correlated residual.

31. The apparatus of claim 30 , wherein below the predetermined transform size the Slant transform is applied and above the predetermined transform size the Haar transform or the Walsh-Hadamard transform is applied.

32. The apparatus of claim 24 , wherein the second transform has a basis function including a discontinuity and/or constant increase or decrease.