COLOR QUANTIZATION BASED ON DESIRED UPPER BOUND FOR RELATIVE QUANTIZATION STEP
First Claim
1. A computer-implemented method for quantizing a first digital image, the method comprising:
- obtaining a first parameter representing a desired upper bound δ
Smax for relative quantization steps to be used when quantizing any color in at least a first range of colors, wherein any adjacent colors S′
;
S″
available for a quantized image correspond to relative quantization steps δ
S′
=∥
S′
−
S′
∥
/∥
S′
∥ and
δ
S″
=∥
S′
−
S″
∥
/∥
S″
∥
, where for any color S, ∥
S∥
is the square root of the sum of squares of tristimulus values of the color S in a predefined 70%-orthonormal linear color coordinate system (“
linear CCS”
); and
generating a quantized digital image using the first parameter.
1 Assignment
0 Petitions
Accused Products
Abstract
A computer system (710) receives a desired upper bound δSmax for a relative quantization step ∥S′−S″∥/∥S′∥ to be used when quantizing any color in some range of colors. Here S′ and S″ are adjacent colors in the set of colors to be made available for the quantized image, and ∥·∥ is a norm in a 70%-orthonormal linear color coordinate system, the norm being the square root of the sum of squares of the tristimulus values. The computer system determines (510) suitable quantization steps for the brightness coordinate (B) and the chromatic coordinates (e,f) in a non-linear color coordinate system, and quantizes (520) the brightness and chromatic coordinates accordingly.
15 Citations
34 Claims
-
1. A computer-implemented method for quantizing a first digital image, the method comprising:
-
obtaining a first parameter representing a desired upper bound δ
Smax for relative quantization steps to be used when quantizing any color in at least a first range of colors, wherein any adjacent colors S′
;
S″
available for a quantized image correspond to relative quantization steps δ
S′
=∥
S′
−
S′
∥
/∥
S′
∥ and
δ
S″
=∥
S′
−
S″
∥
/∥
S″
∥
, where for any color S, ∥
S∥
is the square root of the sum of squares of tristimulus values of the color S in a predefined 70%-orthonormal linear color coordinate system (“
linear CCS”
); andgenerating a quantized digital image using the first parameter. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28)
-
-
29. A computer-implemented method for image processing, the method comprising converting a digital image between a first digital representation of the image and a second digital representation of the image, wherein the image comprises a plurality of image portions, wherein an image at each said image portion has color coordinates S1, S2, S3 in a first color coordinate system (CCS), wherein
S1=√- {square root over (T12+T22+T32)}
S2=T2/S1
T1, T2, T3 are color coordinates in a predefined 70%-orthonormal linear CCS;wherein at least for each said image portion whose color coordinate S1 is in a predefined range, the first digital representation comprises color coordinates s1, s2, s3 such that;
s1=kB·
(1n(α
S1)+β
) rounded to an integer, where α and
β
are predefined constants,
S2=keƒ
·
S2/S1 rounded to an integer,
S3=keƒ
·
S3/S1 rounded to an integer,wherein keƒ
is about 3·
kB, or keƒ
is equal to the smallest power of 2 which is greater than or equal to about 3·
kB. - View Dependent Claims (30, 31)
- {square root over (T12+T22+T32)}
-
32. A computer-readable manufacture comprising computer-readable encoding of digital data representing an image at a plurality of image portions, wherein the image comprises a plurality of image portions, wherein an image at each said image portion has color coordinates S1, S2, S3 in a first color coordinate system (CCS), wherein
S1=√- {square root over (T12+T22+T32)}
S2=T2/S1
S3=T3/S1T1, T2, T3 are color coordinates in a predefined 70%-orthonormal linear CCS; wherein at least for each said image portion whose color coordinate S1 is in a predefined range, the first digital representation comprises color coordinates s1, s2, S3 such that;
s1=kB·
(1n(α
S1)+β
) rounded to an integer, where α and
β
are predefined constants,
S2=keƒ
·
S2/S1 rounded to an integer,
S3=keƒ
·
S3/S1 rounded to an integer,wherein keƒ
is about 3·
kB, or keƒ
is equal to the smallest power of 2 which is greater than or equal to about 3·
kB. - View Dependent Claims (33, 34)
- {square root over (T12+T22+T32)}
Specification