System and method for reshaping and adaptation of high dynamic range video data

1. A method for generating High Dynamic Range (HDR) and/or Wide Color Gamut (WCG) video data from an encoded video data stream, the method comprising:

extracting, by a non-HDR and/or non-WCG video decoder, a self-referential metadata structure signaled for a video data set in an encoded video data stream comprising encoded non-HDR and/or non-WCG video data;

wherein said signaled self-referential metadata structure includes reshaping parameters for a video data reshaping transfer function signaled in a supplemental enhancement information message of said encoded video data stream and/or video usability information message of said encoded video data stream;

wherein said signaled self-referential metadata structure includes data defining a plurality of segments of the video data reshaping transfer function;

wherein said signaled self-referential metadata structure includes a smoothness parameter for each of said plurality of segments, where the smoothness parameter selectively indicates one selected from the group comprising;

(i) a first value for said smoothness parameter indicating all polynomial coefficients are provided in said signaled self-referential metadata structure for one of said plurality of segments;

(ii) a second value for said smoothness parameter indicating said one of said plurality of segments is contiguous with an immediately prior one of said plurality of segments;

(iii) a third value for said smoothness parameter indicating a slope of said video data reshaping transfer function at said one of said pivot points is the same for both a current segment associated with said one of said pivot points and an immediately prior segment associated with said one of said pivot points, and indicating said one of said plurality of segments is contiguous with an immediately prior segment related to said one of said plurality of segments;

wherein said reshaping parameters are relevant to the video data set signaled at a picture level in the encoded video data stream;

decoding, by the non-HDR and/or non-WCG video decoder, the encoded non-HDR and/or non-WCG video data to produce decoded non-HDR and/or non-WCG video data;

determining the video data reshaping transfer function based on the extracted metadata structure;

generating reshaped HDR and/or WCG video data as output data by applying the decoded non-HDR and/or non-WCG video data to the video data reshaping transfer function.

2. The method of claim 1 wherein applying said decoded non-HDR and/or non-WCG video data to the video data reshaping transfer function is an inverse luma mapping function applied to a luma component and a scaling of chroma components using a luma-dependent scaling function, where said inverse luma mapping and said luma-dependent chroma scaling are based upon said signaled self-referential metadata structure in said supplemental enhancement information message of said encoded video data stream and/or said video usability information message of said encoded video stream.

3. The method of claim 1 wherein said encoded video data stream is HEVC compliant.

4. The method of claim 1 wherein said encoded video data stream is AVC compliant.

5. The method of claim 1 wherein said supplemental enhancement information message is included in a NAL unit of said encoded video data stream.

6. The method of claim 1 wherein said video usability information message is included in a sequence parameter set of said encoded video data stream.

7. The method of claim 1 wherein said reshaping function is a perceptual quantizer function.

8. The method of claim 1 wherein said reshaping function is a gamma function.

9. The method of claim 1 wherein said reshaping function includes tone mapping.

10. The method of claim 1 , further comprising:

generating a look-up-table (LUT) for the video data reshaping transfer function by applying values representing a range of decoded video data values to the transfer function and storing output values of the transfer function at locations in the LUT addressed by the applied values; wherein generating the reshaped HDR and/or WCG video data includes applying the decoded video data as address values to the LUT.

11. The method of claim 1 , wherein:

the extracted metadata structure includes data defining said plurality of segments of the transfer function, each segment being specified by first and second pivot points and a set of coefficient values, wherein the coefficient values in the set of coefficients is signaled in the metadata structure and a number of coefficient values in the set varies based on characteristics of the segment; and

the regenerating of the video data reshaping transfer function includes regenerating respective equations for each segment from the signaled coefficient values for each segment.

12. The method of claim 11 , wherein:

each segment is modeled as one of: a first order function having a two coefficient coefficient set, a second order function having a three coefficient coefficient set, or a third order function having a four coefficient coefficient set.

13. The method of claim 11 , wherein:

the data defining each segment of the plurality of segments further includes smoothness parameter indicating a number of coefficients for a current segment of the transfer function that are derived from an immediately previous segment of the transfer function; and

the coefficient sets extracted from the metadata structure for segments having smoothness parameter of 1 and 2 have respectively 1 and 2 fewer coefficient values than coefficient sets for segments modeled by equations having the same order and a smoothness degree of 0.

14. The method of claim 1 , wherein the metadata structure includes:

a first metadata structure associated with a supra-picture element of the data stream the first metadata structure including the data defining the video data reshaping transfer function; and

a second metadata structure associated with a picture or sub-picture element of the data stream, the second metadata structure defining application of the video data reshaping transfer function to at least one component of the decoded video data.

15. The method of claim 14 , wherein:

the decoded video data includes a luminance component and two chrominance components;

the first metadata structure includes data defining multiple video data reshaping transfer functions; and

the second metadata structure includes data identifying one of the multiple video data reshaping transfer functions to apply to each of the luminance component and chrominance components.

16. The method of claim 14 , wherein:

the decoded video data includes a luminance component and two chrominance components;

the first metadata structure includes data defining one video data reshaping transfer function;

the second metadata structure includes:

data identifying the one video data reshaping transfer function as being associated with the luminance component and the two chrominance components; and

data indicating respective input and output offset values for each chrominance component;

the regenerating of the video data reshaping transfer function uses data extracted from the first metadata structure; and

the generating of the reshaped HDR video data for the chrominance components includes applying the decoded chrominance components to the video data reshaping transfer function as modified by input and output offset values extracted from the second metadata structure.

17. The method of claim 14 , wherein:

the decoded video data includes a luminance component and two chrominance components;

the first metadata structure includes data defining first and second video data reshaping transfer functions;

the second metadata structure includes:

data identifying the first video data reshaping transfer function as being associated with the luminance component; and

data identifying the second video data reshaping transfer function as being associated with at least one of the two chrominance components;

the regenerating of the video data reshaping transfer function uses data extracted from the first metadata structure; and

the generating of the reshaped HDR or WCG video data uses data extracted from the second metadata structure to apply the decoded video data of the luminance component to the first video data reshaping transfer function and to apply the decoded video data of the at least one of the two chrominance components to the second video data reshaping transfer function.

18. The method of claim 17 , wherein:

the first metadata structure includes data defining a third video data reshaping transfer function;

the second metadata structure includes data identifying the second video data reshaping transfer function as being associated with a first one of the two chrominance components and data identifying the third video data reshaping transfer function as being associated with a second one of the two chrominance components; and

the generating of the reshaped HDR or WCG video data uses data extracted from the second metadata structure to apply the decoded video data of the first chrominance component to an LUT generated from the second video data reshaping transfer function and to apply the decoded video data of the second chrominance component to the third video data reshaping transfer function.

19. The method of claim 14 , wherein:

the encoded data stream is a H.265 HEVC raw byte sequence payload (RBSP);

the extracting of the metadata includes extracting the first metadata structure from a sequence parameter set (SPS) of the RBSP; and

the extracting of the metadata includes extracting the second metadata structure from a picture parameter set (PPS) of the RBSP.

20. The method of claim 14 , wherein:

the encoded data stream is an H.265 HEVC raw byte sequence payload (RBSP); and

the extracting of the metadata includes extracting the first and second metadata structures from a supplemental enhancement information (SEI) message of the H.265 HEVC RBSP.