JPH0318179A - Color conversion method - Google Patents

Abstract

PURPOSE:To improve color reproducibility by dividing a 1st color space into a tetrahedron having n-point of apexes, converting an optional point into a map point in the tetrahedron in a 2nd color space, obtaining the coordinate and outputting it as a color signal. CONSTITUTION:An LUT 21 sets mapping images Xi,Yi, Zi zetai, xsii, phii, i=1-n from an X, Y, Z space into zetaxsiphi space to cubes V, W representing a color space. The XYZ space (cube V) is divided into a 1st tetrahedron (T) taking 4 points as its apex. Then an optional color is read from an original to make it corresponding to an optional point P in the XYZ space. The 1st tetrahedron Tp including the point P is selected, Let apexes of the tetrahedron Tp be A-D, then coefficients alpha, beta, gamma having a relation of vector AP=alphaAB+betaAC+gammaAD are obtained. A point Q of a tetrahedron Ua (apexes E-H) in the zetaxsiphi space corresponding to the tetrahedron Tp having a relation of vector EQ=alphaEF+betaEG +gammaEH is obtained. Thus, a map point Q in the tetrahedron U0 with respect to the point P is obtained. The coordinate is outputted as a color signal of a 2nd display system.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a color conversion method for converting a color signal of a certain color system to a color signal of another color system, and particularly relates to a color conversion method for converting a color signal of a certain color system into a color signal of another color system. This paper relates to a color conversion method that achieves this. [Prior Art] Generally, in a color image input device, photoelectric conversion is performed using a color image sensor or the like via a complementary color filter to obtain, for example, a complementary color signal. However, in order to convert this color signal into the primary color R (red), G (green), and B (blue) color signals, or to output it as a hard copy, Y (yellow) is added from RG Bffi. , M (magenta), C (cyan), etc., it is necessary to convert the color system. Figure 5 shows, for example, "Journal of the Institute of Image Electronics Engineers" (Volume 12, 4
1983), “Color Photoelectric Conversion” (p. 265~
This is a diagram III showing a general photoelectric conversion system described on page 268 of page 271). In the figure, a color original (1) is irradiated with light from a light source (2), and a rod lens array (3) is arranged on the irradiated surface of the color original (1). Rod lens array (
3) On the imaging plane, there is a CCD (Charge Coupon).
A line sensor chip (4) made of PLEDD device is disposed, and this line sensor chip (4) is arranged on a ceramic substrate (5). Figure 6 is an explanatory diagram showing the arrangement of the light receiving parts that perform photoelectric conversion on the line sensor chip (4). In the figure, the light receiving part (4H) corresponding to W (white) has no color filter, and
A yellow color filter is provided on the surface of the light receiving section (4Y) corresponding to (yellow), and a yellow color filter is provided on the surface of the light receiving section (4Y) corresponding to C (cyan).
A cyan color filter is provided on the surface of C>. Next, the operation of the photoelectric conversion system shown in FIGS. 5 and 6 will be explained. The color image on the color original (1) is illuminated by a light source (2), and is focused through a rod lens array (3) onto a line sensor chip (4) as an erect real image at the same size as the line sensor chip. (4) Each upper light receiving section (4-),
(4Y) and (4C) output electrical signals indicating each color component separated by the color filter as w, y, and c color signals.

Next, the thus obtained W. Y. Based on the color signal of R. G. A conventional color conversion method for obtaining a color signal of the B primary color system will be explained. The value of the RGB signal can be found, for example, in "Encyclopedia of Colors J" by Motoro Kawakami et al.
(1987), page 28, R= S S(λ) r(λ) dλ G= S S(λ) g(λ>dλ ・・・■
B= i S(λ) b(λ) dλ. However, S(λ) is the spectral energy distribution of the reflected light from the color original (1), and r(λ), g(λ
), b(λ) are color matching functions for each of RGB. Therefore, the spectral characteristics of the line sensor chip (4) are? −
y. For each color signal of c, W=c. [r(λ) 10g(λ)+b(λ)]Y=c2[
r(λ)+g(λ)]...■C = c
, [g (λ〉+b(λ〉1) However, C l■ C *
: If the constant is satisfied, R, G. The value of H is R = W −
1.12c G = Y +1. l2C-W...
・■B = W− Y. This primary conversion formula ■ allows color originals (
1> The above colors can be converted from WYC signals to primary color RGB signals. In addition, the color space is WYC space as mentioned above, RG
In addition to B space and YMC space, the two-dimensional color coordinates are 1 (
Luminance) L”a’b” space with added coordinates and L”u”v
``Although there are various spatial, etc., similar linear conversion formulas are used when converting to color signals of other color systems. [Problems to be solved by the invention] Conventional color conversion methods are as described above. , by performing linear conversion, color signals of other color systems, such as RGB signals, are obtained. However, the spectral characteristics of the line sensor chip (4) such as a color image sensor are different from each light receiving part (4Y) and (
4C) and the spectral sensitivity characteristics of the light-receiving element, the isochromatic interval number r(λ
), g(λ), and b(λ). Therefore, the spectral characteristics of the line sensor chip (4) are almost never expressed by the formula (2), and it is difficult to achieve highly accurate color reproducibility using the linear conversion formula (2). In order to solve such problems, for example,
As described in Japanese Patent Application No. 332535, a method can be considered in which each color space is divided into finite small spaces and a different linear transformation formula is applied to each small space. However, in this method, each small space is a cube and a linear transformation formula is used, so discontinuities occur at the boundaries of adjacent small spaces, which causes discontinuities when reproducing the original color at the boundaries. There was a problem that false contours were generated due to continuous parts. This invention was made to solve the above-mentioned problems, and aims to provide a color conversion method that achieves highly accurate color reproducibility without producing false contours. [Means for Solving the Problems] A color conversion method according to the present invention provides a first color system representing a first color system.
a step of setting a mapping for converting the color space from the color space to a second color space representing a second color system for each of the n points; dividing into a plurality of first tetrahedra that do not overlap each other and always include any point in the first color space;
a step of making an arbitrary color signal according to a first color system correspond to an arbitrary point in the first color space;
the step of selecting a tetrahedron in the second color space C corresponding to the first tetrahedron, and the step of calculating by linear calculation a mapping point corresponding to an arbitrary point on the second tetrahedron in a second color space C corresponding to the first tetrahedron. It is. [Operation] In this invention, a mapping from the first color space to the second color space is given in advance for n points, and a tetrahedral division algorithm is introduced to transform the first color space with n points as vertices. The mapped point is divided into a plurality of first tetrahedrons, and any point in the first tetrahedron is converted into a mapped point in the second tetrahedron in the second color space that corresponds to the mapped point by mapping. The color reproducibility is improved by finding the coordinates of by linear calculation and outputting them as color signals of the second color system. ``Example'' An example of the present invention will be described below with reference to the drawings. 1st
The figure is a flowchart diagram showing one embodiment of this invention.
The figure is a block diagram showing the circuit configuration for realizing this invention, Figure 3 is an explanatory diagram showing the first and second color spaces, and Figure 4 is an explanatory diagram showing the first and second color spaces.
The figure is an explanatory diagram showing the first and second tetrahedrons. The photoelectric conversion system applied to this invention is as shown in FIGS. 5 and 6. Now, as shown in Figure 3, any two
Consider two color spaces, and let XYZ space be the first color space, ζξψ
Let space be the second color space. Here, the XYZ space is WY
It can be thought that the ζξψ space corresponds to the C space, and the ζξψ space corresponds to the RGB space.The upper limit of the coordinates that can be taken within each color space is determined by the illuminance of the light source (2) and the line sensor chip (4) shown in Figure 5. It is given by the sensitivity etc. Therefore, each color space is defined as a first cube V and a second cube W by the upper limit values X, Y, Zs, ζLξwheel, and ψ1. For example, if the color signal from the line sensor chip (4) is AD converted into 8 bits, the possible range of the WYC signal output from the photoelectric conversion system in FIGS. 5 and 6 is 0.
~255. Therefore, any color original (1
), the point P in the XYZ space corresponding to that color signal is V ((X.Y.Z) l o=x=x m,05Y3
Ym, 0≦2≦Z It will be included in the coordinates (X, Y, Z) in the first cube V represented by Ml. Similarly,
The point Q in the ζξψ space corresponding to the point P in the XYZ space is W{(ζ,ξ.φ)10≦ζ≦ζtransport, 0≦ξ≦ξcon. 0
≦ψ≦ψtransport} in the second cube W
It will be included in (ζ.ξ, ψ). Therefore, as shown in FIG.
, the mapping of n points from XYZ space to ζξψ space (
X. ,Y. ．． Z. →ζ、． ξ1. ψ+: i=
l to n) (step S1). This mapping is not a simple linear transformation like the formula ■, nor is it a transformation according to an analytical function system, but can be obtained as follows. That is, n color patches are prepared as standard samples, and the spectral reflection characteristics of each color patch are measured using a spectrophotometer or the like. For example, if the data of the spectral energy distribution S(λ) is obtained from the distributed reflection characteristics of the i-th color patch, and the RGB values are obtained using the formula from this data, this RGB value will be calculated from the color chart in the ζξψ space. It corresponds to point Q, which indicates the position coordinates of . On the other hand, the same i-th color patch is read using the line sensor chip (4) as described above, and the obtained color signal is +X
Point P indicating the position coordinates of the color patch in YZ space. Suppose that By repeating the above operation and sequentially reading the color signals of the i (=1 to n> color patches), a mapping is obtained that converts P, (X., Y., Z.) → Q. (ζ1ξiψI) is obtained for n points, and step S1 ends.This mapping can also be used to convert Q.→P1.Also, at this time, the number n of types of color chips used is Several types including color and black and white are required as a minimum, and in order to ensure practical conversion accuracy,
Approximately 100~zoos level is required. Next, the XYZ space, that is, the first cube V, is divided into a plurality of first tetrahedrons T, each of which has four points among the n points as vertices (
Step S2). At this time, each tetrahedron T is defined so that it does not overlap with each other and always includes any point within the first cube V.

Furthermore, as shown in FIG. 4, if each term point of one of the plurality of first tetrahedra Tp is A to D, then according to the mapping obtained in step Sl, there are also points in the second cube W. , a second tetrahedron UQ having vertices at each of the four points E to H is defined corresponding to the first tetrahedron Tp. Next, assuming that an arbitrary color is read from the color original (1), a certain color signal according to the first color system is selected, and this is made to correspond to an arbitrary point P in the XYZ space ( Step S3). Then, any point P within the first cube V is
is included in the first tetrahedron T that includes any point P, as shown in FIG. (FuP)
(Step S4>) At this time, since the plurality of first tetrahedrons T are set not to overlap each other in step S2, the first tetrahedron Tp is uniquely determined. Next, the vertices of the first tetrahedron Tp are defined as A to D, and satisfying coefficients α, β, and γ are determined (step S5).

At this time, as described above, since the vertices E to H of the second tetrahedron U0 corresponding to the first tetrahedron T are included in the n-point mapping set in step S1, the vertices A to D (
The first tetrahedron T. ) are vertices E to H (second tetrahedron U.)
mapped to

Moreover, each tetrahedron T, and U. Assuming that the interior of is linear, any point P in XYZ space (xp, ylP, zp
) mapped from point Q (ζQ, ξq, ψQ
) is EQ-αEF+βEC+γEl+...
・■ will be satisfied. Therefore, ■ based on linear operations
From the equation, the second tetrahedron U. in the ζξψ space after transformation. The mapping point Q can be found for (step S6). The above operations are performed in the XYZ space and ζξψ as shown in Figure 2.
LUT (21) to (23) are mappings corresponding to n points, which are executed by a plurality of LUTs (lookup tables) <21) to (23) provided corresponding to each coordinate value in space. It is composed of a ROM that stores , and is designed to output XYZ values consisting of digital signals as ζξψ values. Therefore, L U T (21) to (23) output the mapped point Q obtained as described above as a point obtained by mapping an arbitrary point P to the ζξψ space (step S7). At this time, for the n points determined by the measurement in step S1, the output value ζξψ is obtained according to the measured value, so if the noise component is ignored, the color difference becomes O, and the color reproducibility is improved compared to the conventional linear conversion. Significant improvement. Also, assuming that the interior of each tetrahedron T and U is linear,
Since the mapping (transformation) point Q in the ζξψ space for each point P in the XYZ space is determined, continuity at the boundary surface of each tetrahedron is always ensured. Therefore, the converted color signal can be reproduced with high precision without generating false contours due to the color difference.

Note that the XYZ space and ζξψ are the first and second color spaces.
It goes without saying that the space is not limited to WYC space, RGB space, etc., but any color space will have the same effect. [Effects of the Invention] As described above, according to the present invention, the mapping for converting from the first color space representing the first color system to the second color space representing the second color system is performed at n points. Steps to configure for
dividing the first color space into a plurality of first tetrahedra that do not overlap each other and always include any point in the first color space, with each of the n points having four points as vertices; a step of associating an arbitrary color signal according to the first color system with an arbitrary point in the first color space; a step of selecting a first tetrahedron including the arbitrary point; and a step of calculating by a linear operation a mapping point corresponding to an arbitrary point on the second tetrahedron in the second color space corresponding to the tetrahedron, and introducing a tetrahedral division algorithm. Since any point within the tetrahedron in the color space is linearly converted into a corresponding point within the tetrahedron in the second color space through mapping, false contours are not generated and color reproducibility is highly accurate. This has the effect of providing a color conversion method that achieves this.

[Brief explanation of the drawing]

FIG. 1 is a flow chart diagram showing an embodiment of the present invention;
Fig. 2 is a block diagram showing an example of a circuit configuration for realizing the present invention, Fig. 3 is an explanatory diagram showing the first and second color spaces, and Fig. 4 shows the first and second four-sided book. 5 is a configuration diagram showing a general photoelectric conversion system used for color conversion, and FIG. 6 is an explanatory diagram showing a color filter arrangement of the light receiving part of the line sensor chip in FIG. 5. ■... Cube W in first color space... Cube T1 in second color space... First four sides #U. ...Second tetrahedron P
...Arbitrary point Q...Mapped point A-D. E~
H... Vertex S1... Step S2 of setting a mapping... Step S3 of dividing into a plurality of first tetrahedra
...Step S4 of making the color signal correspond to an arbitrary point...
・Steps 55 to S6 of selecting a tetrahedron containing an arbitrary point
. . . Step of determining mapped points In the drawings, the same reference numerals indicate the same or corresponding parts.

Claims (1)

[Scope of Claims] A color conversion method for converting an arbitrary color signal from a first color system to a second color system, comprising: converting an arbitrary color signal from a first color space indicating the first color system to the second color space; a step of setting a mapping for each of the n points to be converted into a second color space showing a color system of and dividing an arbitrary color signal according to the first color system into a plurality of first tetrahedra that necessarily include an arbitrary point in the first color space; selecting a first tetrahedron that includes the arbitrary point; and selecting a second tetrahedron in the second color space that corresponds to the first tetrahedron.
A color conversion method comprising the step of determining a mapping point corresponding to the arbitrary point in the tetrahedron by linear calculation.