JPH0816626B2 - Method for obtaining spectral reflectance distribution of non-fluorescent object color - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for accurately obtaining a spectral reflectance distribution of a non-fluorescent object color such as paint, pigment and dye.

Conventionally, except when toning the actual color sample itself,
The automatic color matching is performed using the colorimetric coordinates or color symbols attached to the color sample as an index. If the color sample is not provided with colorimetric coordinates or color symbols, the color sample is physically measured and the colorimetric coordinates obtained thereby are used.

Currently devised color systems for non-fluorescent object colors, such as Munsell color system, Ostwald color system, DIN color system, N
In the CS color system and the like, colors are specified by colorimetric coordinates or color symbols, and the standard color chart is an array of the specified colors according to the color system. However, the standard color chart is
It does not completely reproduce the underlying color system. Therefore, in the automatic color matching, a duplicate color is manufactured by using the colorimetric coordinates of the color sample or the color symbol as an index. Therefore, the produced duplicate color and the color sample match each other under the illumination during the toning, but do not match under the illumination different from that during the toning. There is a so-called color matching condition. Therefore, the duplicated color is provided with an index indicating the degree of color matching with conditions for various light sources. As described above, the conventional automatic color matching technique has a low economic efficiency, and is unavoidable as a trial and error technique.

The present invention has been made in consideration of the above points, and uses not only colorimetric coordinates or color symbols but also a predetermined spectral reflectance distribution of a non-fluorescent object color using a predetermined four-component reflectance distribution. And the spectral reflectance distribution corresponding to arbitrary colorimetric coordinates or color symbols. The present invention will be described below.

 First, the principle of the present invention will be described.

Currently used standard color chart, namely Munsell Bo
The characteristics of the spectral reflectance distribution (4353 lines) of the ok of Color glossy plate, matte plate, and JIS Z-8721-compliant glossy plate standard color chart were extracted, and as a result, the color chart was red, yellow, green, blue, and It was confirmed that it can be classified into five purple lines, and that all color chips can be manufactured with a combination of several types of basic colors and a mixing ratio of basic colors.

Next, the spectral reflectance distributions classified for each color system were aggregated by overlapping adjacent hues for each color system and including all the lightness and saturation to form five color groups. It is based on the rule of thumb that the five basic colors are closely related to human color perception. Then, four component reflectance distributions were extracted for each color group. That is, a total of 20 component reflectance distributions were obtained for the five color groups. This will be described with reference to FIGS.

1 to 5 are diagrams showing the component reflectance distributions obtained in each color group, and FIG. 1 shows a reddish color, that is, in the Munsell symbol, from 2.5RP to 2.5 Munsell Hues step
Fig. 2 is a component reflectance distribution chart that can synthesize spectral reflectance distributions of non-fluorescent object colors up to 7.5YR. Fig. 2 shows yellow colors, that is, if written in Munsell symbols, it is 7.5YR from 2.5YR to 7.5 Munsell Hustep. It is a component reflectance distribution map which can synthesize | combine the spectral reflectance distribution of the non-fluorescent object color to GY. Similarly, Fig. 3 is greenish color, that is, if you write with Munsell symbol, from 2.5GY to 2.5BG in 7.5 Munsell Hue step, Fig. 4 is blue color, that is, if you write with Munsell symbol. , 7.5PB from 2.5BG to 2.5 Munsell Hustep
Up to Fig. 5 is purple color, that is, if written with Munsell symbol, it is 7.5R from 2.5PB to 2.5 Munsell Hustep.
It is a figure which respectively shows the component reflectance distribution map which can synthesize | combine the spectral reflectance distribution to P.

Each of FIGS. 1 to 5 is an average component reflectance distribution chart for each wavelength, each of FIGS. 1B to 5B is a first component reflectance distribution chart, and each of FIGS. Rate distribution map, each (d)
The figure is a third component reflectance distribution diagram, in which the horizontal axis represents wavelength (nm) and the vertical axis represents relative reflectance. In the following description, the function shown in each figure (a) is S ₀ (λ), the function shown in each figure (b) is S ₁ (λ), and the function shown in each figure (c) is S ₂ (λ). (Λ), the function shown in each (d) diagram is S
Expressed as ₃ (λ).

The principle of deriving the average component reflectance distribution and the first, second, and third component reflectance distributions will be described by taking the red color group as an example.

The spectral reflectance distribution of the standard color chart is measured at intervals of 10 nanometers for wavelengths of 400 to 700 nanometers. There are 31 points of measurement data for each color chart. That is, one color chart can be physically characterized in a 31-dimensional space. The invention is an idea to try to estimate the physical properties of a color chart with as few dimensions as possible.

The number of color chips belonging to the red color group is n.

The color space of the red color group is a group of 31 × n measurement points. Consider consolidating 31 × n groups into as few dimensions as possible.

From the 31 × n matrix, the average component reflectance distribution (corresponding to the symbol S ₀ (λ) of the present inventors) for each wavelength is obtained for wavelengths of 400 to 700 nanometers.

Next, the principal component analysis method is used for the dispersion covariance matrix between the wavelengths to calculate the reflectance distributions of the first, second and third components (symbols S ₁ (λ), S ₂ (λ) and Equivalent to S ₃ (λ).

Next, the spectral reflectance distribution ρ will be described with reference to FIGS.
A method of obtaining (λ) will be described.

Spectral reflectance distribution ρ (λ) of an arbitrary non-fluorescent object color
Is determined by the following formula (1).

ρ (λ) = S ₀ (λ) + k ₁ S ₁ (λ) + k ₂ S ₂ (λ) + k ₃ S ₃ (λ) (1) where λ indicates a wavelength of 400 nm to 700 nm, and the wavelength interval is Is 10 nm for convenience, and k ₁ , k ₂ , and k ₃ represent undetermined coefficients.

The undetermined coefficients k _{1 to} k ₃ can be determined as follows.

First, the colorimetric coordinates (X _C , Y _C , Z _C ) of the color sample of the non-fluorescent object color
If this is given, based on this value, Appendix 1 of Japanese Industrial Standards "Color display method-Display by three attributes JIS Z 8721"
(Achromatic color) or Appendix 2 (chromatic color) is referred to determine the color group, and the above functions S ₀ (λ), S ₁ (λ), S ₂ (λ), S
₃ Calculate the colorimetric coordinates (X, Y, Z) with respect to (λ).

Here, the colorimetric coordinates for S ₀ (λ) are (X ₀ , Y ₀ , Z ₀ ), and the colorimetric coordinates for S ₁ (λ), S ₂ (λ), S ₃ (λ) are (X _S1 , Y _S1 , Z _S1 ), (X _S2 , Y _S2 , Z _S2 ) and (X _S3 , Y _S3 , Z
_S3 ).

If these colorimetric coordinates are used to represent the formula (1), the following formula (2) is obtained.

Where E _C (λ) is the spectral distribution of the standard light C of the International Commission on Illumination (CIE), S ₀ (λ) is the average component reflectance distribution obtained by the present invention, and ((λ), (Λ),
(Λ)) represents a CIE standard colorimetric observer. λ = 1 is wavelength 400
Nanometer, λ = 2 is wavelength 410 nanometer and λ = 31
Corresponds to a wavelength of 700 nanometers. Δ (λ) is 10 nanometers.

Si (λ) is the component reflectance distribution obtained in the present invention,
i = 1 corresponds to the first component reflectance distribution S ₁ (λ), i = 2 corresponds to the second component reflectance distribution S ₂ (λ), and i = 3 corresponds to the third component reflectance distribution S ₃ (λ) To do.

If the inverse matrix of A is B, the undetermined coefficients k ₁ , k ₂ and k ₃ can be uniquely obtained by the following equation (3).

By substituting the undetermined coefficients k ₁ , k ₂ , and k ₃ thus obtained into the equation (1), a spectral reflectance distribution satisfying the colorimetric coordinates (X _C , Y _C , Z _C ) of the color sample can be synthesized. .

Next, how the spectral reflectance distribution obtained as described above can be a reference spectral reflectance distribution will be described by taking the Munsell display system, which is the most widely used at present, as an example.

The measurement coordinates defined by the Munsell color system are calculated according to the above equations (1) to (3), and the spectral reflectance distribution is synthesized for each color chart. Calculating the colorimetric coordinates by the spectral reflectance distribution synthesized for the color chart of each lightness specified in the Munsell color system, the International Commission on Illumination (CI
Place the colorimetric coordinates on the chromaticity diagram defined by E) and check the regularity. This study was carried out on the standard lights A, B, C and D daylight prescribed by the International Commission on Illumination, and various actual white practical light sources.

The regularity of the arrangement of colorimetric coordinates was verified under the theoretical light source and various practical white light sources in the spectral reflectance distribution synthesized as described above.

Examples of the results of this examination are shown in FIGS. 6 and 7 by the CIE (u ′, v ′) chromaticity coordinates for the theoretical light source and the practical light source. Fig. 6 is an equal-hue, equal-chroma locus under the CIE standard light "A", Fig. 7 is a locus similar to Fig. 6 under an actual fluorescent lamp, and u'on the horizontal axis is a numerical value. Represents the change from green to red in the color chart from the smallest to the largest, and v ′ on the vertical axis represents the blue to yellow in the color chart from the smallest numerical value to the largest. Represents the change of. Further, VALUE = 5 in both FIGS. 6 and 7.

Figure 7 shows the colorimetric coordinates under a three-wavelength fluorescent lamp that has a unique spectral distribution shape in a practical fluorescent lamp.
This indicates that the arrangement of grid points in the figure is smooth.

The contribution rates of the respective functions S ₁ (λ), S ₂ (λ) and S ₃ (λ) shown in (a) to (d) of FIGS. 1 to 5 are shown in Table 1 below. Like

As can be seen from Table 1, the component reflectance distribution according to the present invention has extremely high reproducibility.

Next, an embodiment of the present invention based on the above principle will be described with reference to FIG.

In FIG. 8, 1 is a light source, 2 is a measured object of a non-fluorescent object color, 3 is a color measurement unit, and the reflected light from the measured object 2 is input,
The colorimetric coordinates (X _C , Y _C , Z _C ) are obtained from this reflected light. Reference numeral 4 is a color chart group determination unit, based on the output of the color measurement unit 3, to which system the color of the DUT 2 belongs, among red, yellow, green, blue and purple.
Alternatively, it is determined whether or not the group is spread over two groups. Reference numeral 5 is a memory, and the figures shown in FIGS. 1 to 5 (a) to (d) have a total of 20 tables quantified in steps of wavelength 10 nm.
Individually prepared. The memory 5 may be analog type. An arithmetic unit 6 calculates the above-described equations (1) to (3) to obtain the spectral reflectance distribution ρ (λ). Reference numeral 7 denotes a control unit that controls each unit.

In the above structure, the light from the light source 1 is reflected by the surface of the DUT 2 and enters the colorimetric unit 3. The colorimetric coordinates (X _C , Y _C , Z _C ) are obtained here, and the color chart group is determined by the color chart group determination unit 4 in the next stage. This determines the table in the memory 5 to be used. Necessary calculations are performed by the calculation unit 6 to obtain the spectral reflectance distribution ρ (λ).

As described in detail above, according to the present invention, the feature extraction of the spectral reflectance distribution of the standard color chart is performed by dividing into five color chart groups of red, yellow, green, blue, and purple, and the average component of each color chart group is calculated. , The first component, the second component and the third component reflectance distribution functions S ₀ (λ), S ₁ (λ), S ₂ (λ), S ₃ (λ) are prepared in advance for automatic color matching. , If the colorimetric coordinates of the non-fluorescent object color of the measured object are obtained, the color chart group is determined from the values, and if the coefficients k ₁ , k ₂ , and k ₃ in the color chart group are obtained, then the spectral reflectance is The distribution ρ (λ) can be obtained from ρ (λ) = S ₀ (λ) + k ₁ S ₁ (λ) + k ₂ S ₂ (λ) + k ₃ S ₃ (λ).

Therefore, it is possible to synthesize the reference spectral reflectance distribution of the non-fluorescent object color, in the automatic color matching, it is possible to obtain an objective function that did not exist in the past, if this is used as an index, the automatic color matching is significantly improved, The toning precision can be improved, the speed can be increased, and the economy can be improved.

As described above, the present invention provides one new index in automatic color matching in response to the colorization of various industrial products, home appliances, daily necessities, etc., which have been rapidly developing in recent years, and its significance is extremely high. large.

[Brief description of drawings]

FIGS. 1 to 7 are diagrams for explaining the principle of the present invention, FIGS. 1 to 5 are diagrams showing component reflectance distributions created from standard color charts, and each (a) diagram is an average component. The reflectance distributions, each (b) figure shows the first component reflectance distribution, each (c) figure shows the second component reflectance distribution, and each (d) figure shows the third component reflectance distribution. Figures 6 and 7 show under the CIE standard light "A",
Diagram showing colorimetric coordinates under three-wavelength fluorescent lamp,
FIG. 8 is a block diagram showing an embodiment of the present invention. In the figure, 1 is a light source, 2 is an object to be measured, 3 is a color measurement unit, 4 is a color chart group determination unit, 5 is a memory, 6 is a calculation unit, and 7 is a control unit.

Claims (1)

[Claims] 1. A feature extraction of a spectral reflectance distribution of a standard color chart of a non-fluorescent object color is divided into five color chart groups based on Munsell hue values, and each of the color chart groups is analyzed by a principal component analysis method. Average component with respect to wavelength, first component, second component and third component reflectance distribution S ₀ (λ), S ₁ (λ), S ₂ (λ)
And S ₃ (λ) are created, the colorimetric coordinates of the non-fluorescent object color of the measured object are measured during automatic color matching, the color chart group is determined, and then the coefficient k ₁ in the color chart group, Find k ₂ and k ₃ ,
A method for obtaining a spectral reflectance distribution of a non-fluorescent object color, characterized by obtaining a spectral reflectance distribution ρ (λ) by the following equation using each function and each coefficient. ρ (λ) = S ₀ (λ) + k ₁ S ₁ (λ) + k ₂ S ₂ (λ) + k ₃ S ₃ (λ)