PROVIDING 3D LOOK-UP TABLE (LUT) ESTIMATION FOR COLOR GAMUT SCALABILITY
Systems and/or methods for estimating color conversion components. A video coding device may receive a picture associated with a first color space. The picture may comprise a first component at a first sampling location, a second component at a second sampling location and the second component at a third sampling location. The video coding device may apply a first interpolation filter to the second component at the second sampling location and the second component at the third sampling location to determine the second component at the first sampling location. The second component at the first sampling location may be associated with the first color space. The video coding device may apply a color conversion model to the first component at the first sampling location and to the second component at the first sampling location to translate the first component at the first sampling location to a second color space.
1 - 23 . (canceled)
24 . A video coding device, comprising:
a processor configured to:
receive a first video signal in a first color format and a second video signal in a second color format;
identify a color conversion model for mapping colors in the first color format to corresponding colors in the second color format, the color conversion model comprising a plurality of vertices;
identify a plurality of used vertices in the color conversion model that are referenced in converting the first video signal in the first color format to the second video signal in the second color format;
estimate a plurality of color conversion model parameters for the plurality of used vertices; and
estimate a plurality of color conversion model parameters for a plurality of unused vertices based on the plurality of color conversion model parameters for the used vertices.
25 . The video coding device of claim 24 , wherein the plurality of color conversion model parameters for the plurality of unused vertices are estimated via interpolation using the plurality of color conversion model parameters for the used vertices.
26 . The video coding device of claim 24 , wherein the processor is further configured to include an indication of the plurality of color conversion model parameters for the used vertices and the plurality of color conversion model parameters for the unused vertices in a video bitstream associated with the first video signal and the second video signal.
27 . The video coding device of claim 24 , wherein the processor is further configured to:
identify the plurality of unused vertices in the color conversion model that are not referenced in converting the first video signal in the first color format to the second video signal in the second color format;
construct a first color conversion weight parameter set that comprises weight parameters referencing the plurality of vertices in the color conversion model; and
generate a compact color conversion weight parameter set by removing weight parameters that reference the plurality of unused vertices, wherein the plurality of color conversion model parameters for the plurality of used vertices are estimated based on the compact color conversion weight parameter set.
28 . The video coding device of claim 27 , wherein the processor is further configured to:
calculate a compact auto-correlation matrix based on the compact color conversion weight parameter set, wherein the plurality of color conversion model parameters for the plurality of used vertices are estimated based on the compact auto-correlation matrix and the compact color conversion weight parameter set.
29 . The video coding device of claim 28 , wherein the processor is further configured to:
determine a balance factor for balancing between estimation accuracy and process stability based on the compact color conversion weight parameter set, wherein the compact auto-correlation matrix is calculated further based on the balance factor.
30 . The video coding device of claim 24 , wherein a vertex in the color conversion model is identified as a used vertex when the vertex is used in a color conversion interpolation process to map a color in the first color format to a corresponding color in the second color format.
31 . The video coding device of claim 24 , wherein the color conversion model is partitioned into a plurality of segments, a segment corresponding to a portion of a color space and a subset of the plurality of vertices.
32 . The video coding device of claim 31 , wherein the segment is an octant in the color space.
33 . The video coding device of claim 24 , wherein the color conversion model is a 3D look-up table.
34 . The video coding device of claim 24 , the plurality of color conversion model parameters for the plurality of used vertices are estimated using a least square estimation method.
35 . A video coding method, comprising:
receiving a first video signal in a first color format and a second video signal in a second color format;
identifying a color conversion model for mapping colors in the first color format to corresponding colors in the second color format, the color conversion model comprising a plurality of vertices;
identifying a plurality of used vertices in the color conversion model that are referenced in converting the first video signal in the first color format to the second video signal in the second color format;
estimating a plurality of color conversion model parameters for the plurality of used vertices; and
estimating a plurality of color conversion model parameters for a plurality of unused vertices based on the plurality of color conversion model parameters for the used vertices.
36 . The video coding method of claim 35 , wherein the plurality of color conversion model parameters for the plurality of unused vertices are estimated via interpolation using the plurality of color conversion model parameters for the used vertices.
37 . The video coding method of claim 35 , wherein a vertex in the color conversion model is identified as a used vertex when the vertex is used in a color conversion interpolation process to map a color in the first color format to a corresponding color in the second color format.
38 . The video coding method of claim 35 , further comprising:
identifying the plurality of unused vertices in the color conversion model that are not referenced in converting the first video signal in the first color format to the second video signal in the second color format;
constructing a first color conversion weight parameter set that comprises weight parameters referencing the plurality of vertices in the color conversion model; and
generating a compact color conversion weight parameter set by removing weight parameters that reference the plurality of unused vertices, wherein the plurality of color conversion model parameters for the plurality of used vertices are estimated based on the compact color conversion weight parameter set.
39 . The video coding method of claim 38 , further comprising:
calculating a compact auto-correlation matrix based on the compact color conversion weight parameter set, wherein the plurality of color conversion model parameters for the plurality of used vertices are estimated based on the compact auto-correlation matrix and the compact color conversion weight parameter set.
40 . The video coding method of claim 39 , further comprising:
determining a balance factor for balancing between estimation accuracy and process stability based on the compact color conversion weight parameter set, wherein the compact auto-correlation matrix is calculated further based on the balance factor.
41 . The video coding method of claim 35 , wherein the color conversion model is partitioned into a plurality of segments, a segment corresponding to a portion of a color space and a subset of the plurality of vertices.
42 . The video coding method of claim 35 , wherein the color conversion model is partitioned into a plurality of octants, an octant corresponding to a portion of a color space and a subset of the plurality of vertices.
43 . The video coding method of claim 35 , wherein the color conversion model is a 3D look-up table.