Traitement en cours

Veuillez attendre...

Paramétrages

Paramétrages

Aller à Demande

1. US20180098094 - Inverse luma/chroma mappings with histogram transfer and approximation

Note: Texte fondé sur des processus automatiques de reconnaissance optique de caractères. Seule la version PDF a une valeur juridique

[ EN ]

Claims

1. A method comprising:
constructing a standard dynamic range (SDR) cumulative density function (CDF) based on an SDR histogram generated from a distribution of SDR codewords in one or more SDR images;
constructing a high dynamic range (HDR) CDF based on an HDR histogram generated from a distribution of HDR codewords in one or more HDR images that correspond to the one or more SDR images;
generating a histogram transfer function based on the SDR CDF and the HDR CDF;
transmitting the one or more SDR images along with backward reshaping metadata to one or more recipient devices;
wherein the backward reshaping metadata is generated at least in part on the histogram transfer function.
2. The method of claim 1, wherein at least one of the SDR CDF and the HDR CDF represents a one-dimensional CDF over a single color component of a color space.
3. The method of claim 1, wherein each of the SDR CDF and the HDR CDF represents a multi-dimensional CDF over two or more color components of a color space; further comprising:
constructing one or more marginal SDR CDFs by integrating the SDR CDF over one or more SDR color components of an SDR color space;
constructing one or more marginal HDR CDFs by integrating the HDR CDF over one or more HDR color components of an HDR color space;
performing one or more first searches with the one or more marginal SDR CDFs against the one or more marginal HDR CDFs, wherein each of the one or more first searches is a one-dimensional search for an optimal value of an HDR color component of an HDR codeword to which a given value of an SDR color component of an SDR codeword is to be mapped based on matching a CDF value of one of the one or more marginal SDR CDFs with a CDF value of a corresponding one of the one or more marginal HDR CDFs.
4. The method of claim 3, further comprising performing a second further search with the SDR CDF against the HDR CDF based at least in part on results of the one or more first searches, wherein the second search is:
a one-dimensional search over a specific HDR color component other than the one or more HDR color components of the HDR color space; or
a multi-dimensional search over a multi-dimensional local area in the HDR color space; and wherein the multi-dimensional local area in the HDR color space is defined based on the results of the one or more first searches.
5. The method of claim 1, wherein the backward reshaping metadata comprises a first backward reshaping function represented by the histogram transfer function; and wherein the first backward reshaping function is used to backward reshape luma components of SDR codewords into luma components of predicted HDR codewords.
6. The method of claim 5, wherein the backward reshaping metadata further comprises a second backward reshaping function that is determined based on an MMR-based histogram approximation method; and wherein the second backward reshaping function is used to backward reshape chroma components of SDR codewords into one or more chroma components of predicted HDR codewords.
7. The method claim 1, wherein the one or more target HDR images are derived from the one or more SDR images using an SDR-to-HDR conversion tool.
8. The method of claim 1, wherein the one or more target HDR images are generated from video data without performing any SDR-to-HDR conversion; and wherein the video data is also used to generate the one or more SDR images.
9. The method of claim 1, wherein the backward reshaping metadata comprising operational parameters specifying a set of polynomials that approximate the histogram transfer function.
10. The method of claim 1, wherein the backward reshaping metadata is carried in a video signal as a part of image metadata separate from the one or more SDR image.
11. The method of claim 1, further comprising performing color space conversion with one or both of the one or more SDR images or the one or more target HDR images.
12. The method of claim 1, wherein the video signal excludes the target HDR images or predicted HDR images approximating the target HDR images.
13. The method of claim 1, wherein at least one of the one or more SDR images or the one or more target HDR images are represented in one of: an IPT PQ (ICtCp) color space, an YCbCr color space, an RGB color space, a Rec. 2020 color space, a Rec. 709 color space, an extended dynamic range (EDR) color space, a gamma/HLG/PQ color space, or a standard dynamic range (SDR) color space.
14. The method of claim 1, wherein the one or more SDR images are of a bit depth of one of 8, 9, 10, or 11+ bits.
15. The method of claim 1, wherein the one or more target HDR images are of a bit depth of one of 8, 9, 10, 11, 12, 13, 14, or 15+ bits.
16. An apparatus comprising a processor and configured to perform the method recited in claim 1.
17. A non-transitory computer-readable storage medium having stored thereon computer-executable instruction for executing a method with one or more processors in accordance with claim 1.
18. A method comprising:
constructing, based on backward reshaping metadata received with one or more standard dynamic range (SDR) images, a backward reshaping function;
wherein the backward reshaping function is represented by a histogram transfer function generated based on an SDR cumulative density function (CDF) and a high dynamic range (HDR) CDF;
wherein the SDR CDF is constructed based on an SDR histogram generated from a distribution of SDR codewords in the one or more SDR images;
wherein the HDR CDF is constructed based on an HDR histogram generated from a distribution of HDR codewords in one or more target HDR images that correspond to the one or more SDR images;
applying the backward reshaping function to the one or more SDR images to generate one or more predicted HDR images that approximate the target HDR images;
causing the one or more predicted HDR images to be rendered with a display device.
19. The method of claim 18, wherein the backward reshaping metadata comprises a second backward reshaping function that is generated based on a multivariate multi-regression (MMR)-based histogram approximation method.
20. A non-transitory computer-readable storage medium having stored thereon computer-executable instruction for executing a method with one or more processors in accordance with claim 18.