Methods and systems for scalable-to-non-scalable bit-stream rewriting
Aspects of the present invention relate to systems and methods for selectively determining a quantization parameter based on coding parameters. In some embodiments, a quantization parameter selection may be based on the presence of non-zero transform coefficients as well as other parameters. Some embodiments relate to rewriting a multi-layer bitstream to a single-layer bitstream with a quantization parameter selected with reference to transform coefficients. Some embodiments convert a scalable, multi-layer video sequence into a single-layer, non-scalable video sequence that is compatible with legacy devices.
1. A method for image bitstream processing, said method comprising:
extracting symbols from a multi-layer image bitstream, said symbols being related to an image portion;
generating combined transform coefficients from said symbols, wherein said combined transform coefficients are a combination of enhancement layer coefficients and base layer coefficients;
determining whether said combined transform coefficients comprise any non-zero coefficients;
when said combined transform coefficients comprise a non-zero coefficient, using a quantization parameter identified in said symbols for processing said image portion; and
when said combined transform coefficients do not comprise a non-zero coefficient, using a default quantization parameter, wherein said default quantization parameter is a quantization parameter used for a previous image portion, for processing said image portion.
2. A method as described in claim 1 wherein said processing said image portion comprises in-loop filtering.
3. A method as described in claim 1 wherein said processing said image portion comprises signaling a quantization offset parameter for a subsequent image portion.
4. A method as described in claim 1 further comprising scaling said base layer coefficients and combining said scaled base layer coefficients with said enhancement layer coefficients to form said combined transform coefficients.
5. A method as described in claim 1 wherein said image portion is an image block.
6. A method as described in claim 1 further comprising storing the quantization parameter selected for said processing as a default quantization parameter for a next image portion.
7. A method as described in claim 1 wherein said processing comprises storing said combined transform coefficients in a second bitstream.
8. A method as described in claim 1 further comprising determining image intensity values based on said combined transform coefficients and filtering said image intensity values.
9. A method for combining a bitstream base layer with a bitstream enhancement layer in a multi-layer bitstream, said method comprising:
a) generating base-layer transform coefficients from said multi-layer bitstream;
b) generating enhancement-layer residual coefficients from said multi-layer bitstream;
c) combining said base-layer coefficients with said enhancement-layer residual coefficients to produce combined-layer coefficients;
d) determining whether said combined-layer coefficients comprise any non-zero coefficients;
e) when said combined-layer coefficients comprise a non-zero coefficient, encoding said combined-layer with a quantization parameter and signaling said quantization parameter with a quantization parameter identifier; and
f) when said combined-layer coefficients do not comprise a non-zero coefficient, encoding said combined-layer with a default quantization parameter from a previous block.
10. A method as described in claim 9 further comprising scaling said base-layer transform coefficients.
11. A method as described in claim 9 further comprising using said quantization parameter or said default quantization parameter for in-loop filtering for a current portion of said bitstream.
12. A method as described in claim 9 further comprising using said quantization parameter or said default quantization parameter for signaling a quantization offset parameter for a subsequent bitstream portion.
13. A method as described in claim 10 wherein said scaling is based on quantization parameters of said base layer and said enhancement layer.
14. A method as described in claim 9 further comprising entropy coding said combined layer coefficients.
15. A method as described in claim 9 wherein said base layer and said enhancement layer are part of an SVC-compliant bitstream and said combined layer is part of an AVC-compliant bitstream.