Methods and systems for transform selection and management
Embodiments of the present invention comprise systems and methods for managing and combining layers in a multi-layer bitstream.
1. A method for automatically selecting a transform size, said method comprising:
a) a processing device determining a base layer transform size;
b) said processing device determining if said base layer transform size matches a first predefined transform size, and determining whether an enhancement layer includes a flag indicating a second predefined transform size;
c) said processing device selecting said second predefined transform size when said enhancement layer includes said flag, and otherwise:
i) said processing device selecting a transform of said first predefined transform size for said enhancement layer when said base layer transform size matches said first predefined transform size; and
ii) said processing device otherwise selecting a transform of a default transform size for said enhancement layer.
2. A method as described in claim 1 wherein said default transform size is used for parsing a bitstream regardless of the selection for a transform.
3. A method as described in claim 1 further comprising inverse transforming enhancement layer coefficients with said transform.
4. A method as described in claim 1 wherein said first predefined transform size is 8.times.8.
5. A method as described in claim 1 wherein said first predefined transform size is 16.times.16.
6. A method as described in claim 1 further comprising determining a prediction mode for said enhancement layer and performing steps a) through c) only when said prediction mode indicates that said enhancement layer is predicted from said base layer.
7. A method as described in claim 1 further comprising extracting a plurality of enhancement-layer coefficients formatted for said default transform size.
8. A method as described in claim 7 further comprising reformatting for said first predefined transform size said plurality of extracted enhancement-layer coefficients.
9. A method as described in claim 1 further comprising extracting a plurality of quantized enhancement-layer coefficients formatted for said default transform size.
10. A method as described in claim 9 further comprising reformatting for said predefined transform size said plurality of extracted quantized enhancement-layer coefficients thereby creating reformatted quantized enhancement-layer coefficients.
11. A method as described in claim 10 further comprising:
i) inverse quantizing a plurality of base layer quantized transform coefficients thereby creating base layer transform coefficients;
ii) scaling said base layer transform coefficients to match a characteristic of said enhancement-layer thereby creating scaled, base layer transform coefficients;
iii) inverse quantizing said reformatted quantized enhancement-layer coefficients thereby creating enhancement-layer transform coefficients; and
iv) combining said scaled, base layer transform coefficients with said enhancement-layer transform coefficients to form combined coefficients.
12. A method as described in claim 11 further comprising generating a combined bitstream comprising said combined coefficients.
13. A method as described in claim 12 wherein said combined bitstream further comprises an intra-prediction mode.
14. A method as described in claim 12 wherein said combined bitstream further comprises a motion vector.
15. A method as described in claim 11 further comprising inverse transforming said combined coefficients using said reconstruction transform thereby generating a spatial-domain residual value.
16. A method as described in claim 15 further comprising combining said spatial-domain residual value with said spatial-domain prediction value.
17. A method as described in claim 10 further comprising:
i) scaling a plurality of base layer quantized transform coefficients to match a characteristic of said enhancement-layer thereby creating scaled, base layer quantized transform coefficients; and
ii) combining said scaled, base layer quantized transform coefficients with said reformatted quantized enhancement-layer coefficients to form combined quantized coefficients.
18. A method as described in claim 17 further comprising inverse quantizing said combined quantized coefficients, thereby creating combined coefficients.
19. A method as described in claim 18 further comprising generating a combined bitstream comprising said combined coefficients.
20. A method as described in claim 19 wherein said combined bitstream further comprises an intra-prediction mode.
21. A method as described in claim 19 wherein said combined bitstream further comprises a motion vector.
22. A method as described in claim 18 further comprising inverse transforming said combined coefficients using said reconstruction transform thereby generating a spatial-domain residual value.
23. A method as described in claim 22 further comprising combining said spatial-domain residual value with said spatial-domain prediction value.