JP3260293B2 - Color information exchange method and device characteristic adjustment method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technique relating to color reproduction and color adjustment which can be used for any input image device or output image device such as a camera, scanner, monitor and printer.

[0002]

2. Description of the Related Art In recent years, the colorization of all types of image equipment has been progressing, and color images have been developed for both office use and home use.
Opportunities to use color documents have increased. Along with this, the color reproduction technology of each device alone has been developed, and the performance has been remarkably improved.

However, with the progress of network technology and the increase in social use, an open system has emerged, and the integration of color reproduction and color adjustment technologies that have been growing in individual devices has emerged as the next problem. For example, NTSC signals have been used only for broadcast-type color information transmission. In this case, the color receiving side has sufficient color management according to the standard, and the color management need only be performed by the broadcasting station on the transmitting side. However, with the construction of a high-bandwidth network infrastructure, two-way color information transmission is being used, and each home and each office is the source of information, and the color that has been used by broadcasters with specialized knowledge until now Management must be implemented to some extent in each home and office. Specifically, the sender must convert the information he wants to send, such as a photograph, into an NTSC signal without deteriorating the color reproducibility. Furthermore, color information exchange in open systems, as well as conversion to NTSC signals, requires interconnection of all color signals.

Among the problems of color reproduction and color adjustment caused by the openness of color communication, "difference in color reproduction range between devices" and "difference in observation environment"
Is a major challenge.

First, the "difference in color reproduction range between devices" will be described. In general, CRT displays have a wider color reproduction range than printed matter and can display vivid colors. Therefore, when a CG (Computer Graphic) image is to be printed by a printer, there arises a problem that a color intended by a designer cannot be expressed on a printed material. on the other hand,
Printed matter has better gradation in dark areas, that is, low luminance parts, than CRTs, and when displaying a picture of the printed matter on the CRT, there is a problem in the expressibility of detail in shadow parts. Such a problem of the difference in the color reproduction range between the devices cannot be fundamentally solved due to the limitation that a certain device cannot physically present a color. However, in the above-described open system, color information is exchanged between devices without any restrictions. Therefore, some measure must be taken in such a manner as to minimize the deterioration of color reproducibility. For this reason, "color gamut mapping" for mapping a color exceeding the color reproduction range of the receiving device into the color reproduction range has been actively developed in recent years.

One of them is "Color gamut compression in CG images (II)" (Color Forum JAPAN'96 Abstract P21-P
There is a method described in 24). this is

(Equation 1) This is a method of defining a color difference formula and performing color gamut mapping with a color signal pair that minimizes the color difference. In (Equation 1), (L * ₁ , H * ₁ , C * ₁ ) are the colorimetric values of the colors of the CG image on the monitor, and the color gamut mapping is (L * ₁ , H * ₁ , C * ₁ ) that minimizes the color difference ΔE. * ₂ , H * ₂ , C * ₂ ) within the color gamut of the printer. When several sets of (Kl, Kh, Kc) were set and the mapping conditions were changed to compare the original image and the reproduced image, when Kc ≧ Kh ≧ Kl, a subjectively preferable color gamut mapping was obtained. It is reported that this has been achieved. The mapping method of Kc ≧ Kh ≧ Kl can be expressed as “compressing in the saturation direction while keeping the brightness as constant as possible and changing the hue to some extent” by directly citing the text of the reference.

[0007] Also, "Nikkei Electronics no.570,
1992.12.21 ”has a method of mapping brightness to the outer edge of the color reproduction range by maintaining only the brightness and hue as shown in FIG. 12, adjusting only the saturation, and the center of gravity of the color space as shown in FIG. A method is described in which all points are smoothly compressed toward a part and a mapping point is determined while avoiding a decrease in saturation.

All of the conventional methods described above focus on the degree of preservation of three attributes of lightness, hue, and saturation when determining the mapping direction. Fundamentally, it is not possible to reproduce images with exactly the same color between devices, so if you give priority to saving which attribute or degrade which attribute, the appearance of the reproduced image will be better The mapping method is determined based on the criterion such as whether or not. In the case of FIG. 12, emphasis is placed on preserving brightness and hue.
At the expense of reduced saturation. Therefore, in the above-described conventional example, a plurality of adjustment amounts of the attributes of all three colors are set, a reproduction image is created by all combinations thereof, and the adjustment amounts of the three attributes used for the reproduction image with the highest subjective evaluation are adopted as a mapping method. Steps are taken.

Next, "difference in observation environment" will be described. The colorimetry system specified by the International Commission on Illumination CIE has been used effectively in designing color reproduction systems. If the colorimetric values of the two colors match, it can be determined that the two colors are the same.
However, to date, all colorimetric systems recommended by the CIE have fixed illumination conditions, and two colorimetric values presented under different illuminations cannot be compared.
For example, it is not possible to evaluate the comparison between the colors of an original image illuminated by a fluorescent lamp and a reproduced image illuminated by an incandescent lamp by colorimetric values.

However, in real life, it is extremely difficult to find a scene having exactly the same “illumination environment” consisting of exactly the same illumination state, and furthermore, the observation state including the background state of the observation image and the observation distance. When observing the same CG image on the same display during the day and late at night, the amount of light reflected on the surface of the display is different, so that even if the emission characteristics of the phosphor alone of the display are constant, the contrast,
The colors and the like look completely different. The input / output characteristics of the monitor, including the effects of external light, changed. Further, the exchange of color information between remote locations via a network is a case where it is necessary to control colors especially between different observation environments. When the illumination light source is different between the sender and the receiver, a color pair that is perceived as the same color between the two is generally called a “corresponding color”. When the color of the sender is illuminated by the illumination light source of the receiver, predicting what color the color looks like is referred to as “corresponding color prediction”. A change in the input / output characteristics of a single device due to a change in the observation environment is also one form of the corresponding color prediction.

The corresponding color prediction is described by a change in the adaptation state of the visual system to the illumination ("Color Engineering" (Noboru Ota), Tokyo Denki University Press (1993) P184-P1).
91). von Kries considers that the tristimulus value of the perceived color changes linearly with the ratio of the tristimulus value of the illumination light source, and "the adaptation equation"
Suggested. Naya added the illuminance level and the reflectance of the background to the parameters of the adaptation equation, and applied the Kherson-Jud effect (when illuminating a gray scale with chromatic light, the light gray felt the hue of the illuminating light, while the dark gray felt its complementary color. When you illuminate a group of achromatic colors with different illuminances, you can see the hue), the light grays look whiter and the dark grays appear blacker at high illuminance, and the hunt effect The accuracy of the corresponding color prediction is improved by a non-linear adaptation equation that can predict, for example, the perceived saturation (colorfulness rises according to the illuminance when illuminated).

[0012]

However, some problems remain in the color gamut mapping method and the corresponding color prediction according to the above-mentioned conventional method.

First, in gamut mapping, since a unique relationship has not been established between the adjustment amounts of hue, saturation, and lightness and the color reproducibility, the best color can be obtained unless all possible adjustment amounts are tested. There is a problem that a reproduced image cannot be specified.

Further, the corresponding color prediction can be used only under limited observation conditions, has a large gap with respect to the actual use situation, and has a problem that it has not reached a practical level.

The present invention solves the above-mentioned problems of the prior art, and is capable of uniquely determining a color gamut mapping method and exchanging color information for practically predicting a corresponding color that absorbs a difference in viewing conditions. It is an object to provide a method and a device characteristic adjustment method.

[0016]

In gamut mapping, since the same color cannot be fundamentally presented between devices, color information is exchanged so that the overall impression of an image on each device is as small as possible. Is the goal. For example, if expressed in shades, "red flowers" on a CRT should be "red flowers" on printed matter, and the color categories of flowers such as "yellow flowers" and "orange flowers" are different. It is not preferable. Further, if "flower red is stronger than apple red" on a CRT, it is preferable that "flower red is stronger than apple red" on printed matter, and "flower red is stronger than apple red". "Weaker" is not preferred.

As described above, in order to make the overall impression of an image as different as possible in color gamut mapping, (a) the color category of the same pixel must be the same between devices (b) the entire image It is important that color inversion and discontinuity do not occur over the entire range.

Therefore, according to the present invention, the color space formed by the color signal data for each of the transmitting device and the receiving device is divided into a plurality of color categories by a set of colors perceived by an observer,
Color information is exchanged only between the same color categories.

That is, the drive signal of the pixel perceived as “red” by the transmitting device at the receiving device is selected only from the color signal data group perceived as “red” by the receiving device.

In the present invention, the received color vector is determined from the same color category based on the relative positional relationship between the transmitted color vector and the center of gravity of all the color categories in the transmitting device color space.

For example, the position of the transmitted color vector perceived as “red” is close to the center of gravity of the “orange” color category similar to “red”, and the position of the “green” color category that is a complementary color of “red” is When the center of gravity is farther, color signal data closer to “orange” should be selected for the received color vector rather than “green”.
If the distance between the transmitted color vector and the center of gravity of “orange” is の of the distance between the transmitted color vector and the center of gravity of “green” in the color space of the transmitting device, the reception selected in the receiving device space The distance that the color vector has with the center of gravity of “orange” should be の of the distance with the center of gravity of “green”.

As described above, by selecting the received color vector for each pixel so that the relative positional relationship between the transmitted color vector and the distance that the transmitted color vector has with the center of gravity of each color category is established in the color space of the receiving device, the "image Color information exchange that does not cause "color inversion and discontinuity" throughout can be realized.

In the present invention, the color category is modeled by a standard normal distribution. An average vector of a plurality of color signal data groups that the observer perceptually classified into each color category,
A variance-covariance vector is obtained, a standard normal distribution is defined for each category, and this is defined as the occurrence probability of a color vector from each color category. The color category of a certain pixel is the color category with the highest probability among the occurrence probabilities of each color category.

The present invention also uses the Mahalanobis distance obtained from the average vector and the variance-covariance vector for the distance that the transmitted color vector has with the centroid of each color category and the distance that the received color vector has with the centroid of each color category.

The relative positional relationship between the transmitted color signal and the center of gravity of each color category is obtained by calculating the center of gravity of all color categories, which is given as the average of the center of gravity of all color categories, and calculating from the center of gravity of all color categories to the transmitted color vector. The Mahalanobis distance is described as a normalized transmission color distance obtained by dividing the Mahalanobis distance by the Mahalanobis distance from the center of gravity of all color categories to the center of gravity of each color category.

The relative positional relationship of the received color signal with the center of gravity of each color category is obtained by calculating the center of gravity of all color categories given as the average of the center of gravity of all color categories, The Mahalanobis distance is described as a normalized reception color distance obtained by dividing the Mahalanobis distance by the Mahalanobis distance from the center of gravity of all color categories to the center of gravity of each color category.

The mapping point of the received color vector is determined such that the magnitude relation of the normalized received color distance for all color categories is equal to the normalized transmitted color distance for all color categories.

As described above, according to the present invention, the color category of an arbitrary pixel can be determined in the transmitting device and the receiving device by modeling the color category from the observer's response using the standard normal distribution. By maintaining the relative positional relationship of the center of gravity of the category on the receiving device side, "uniqueness" that cannot be realized by the conventional color gamut mapping method is secured.

The color information exchange method can also be applied to the corresponding color prediction. Conventional color prediction can be performed only under the condition that the observation environment such as lighting is completely controlled, but the present invention performs the classification under the arbitrary observation environment by classifying the color category by the observer. Color prediction can be performed.

If the transmitting device and the receiving device are replaced with the original state and the current state of a single device having changed input / output characteristics, the above color information exchange method can be applied to the device characteristic adjusting method.

As described above, according to the present invention, a color gamut mapping method can be uniquely determined, and the color reproducibility of an image between devices having different viewing conditions can be improved, physical fluctuation or deterioration of devices, or Adjustment of characteristics of a single device whose input / output characteristics have changed due to changes in observation conditions can be realized.

[0032]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A first aspect of the present invention relates to a transmitting device.
The representative color in each color space for each chair and receiving device
And the perceived color for the representative color is determined in advance.
According to the response of the subject who responded with the color name,
Color signal data, and classify the color signal data.
Each area of each color space divided based on the group is classified into a color category.
As a glue area, a color category area to which the color signal data of the transmitting device belongs is determined, and the color signal data is mapped to a color category area of the same color category in the color space of the receiving device, and an overall impression of the image is obtained. Color information can be exchanged so as not to differ from each other, and the color reproducibility of an image can be improved.

The second aspect of the present invention, from the color signal data group receiving device transmitting color vectors of the color information is classified belonging color category area <br/> same color category area originating device transmits Monodea of selecting a reception color vector
Thus, color information can be exchanged between the same color category areas , and color information can be exchanged so that the overall impression of the image does not differ.

According to a third aspect of the present invention, the positional relationship between the transmission color vector and the center of gravity of each color category area of the transmission device is determined by the relationship between the reception color vector and the center of gravity of each color category area of the reception device. position is also intended to select the received color vector as stored in relation hue inverted over the entire image, it is possible to exchange color information so discontinuities have occurred.

According to a fourth aspect of the present invention, the color signal data group classified into each color category region has a color category region in which the transmission color vector shows the highest occurrence probability among all color category regions of the transmission device. The color category with the highest occurrence probability is determined by the probability density function given by the standard normal distribution formed by the mean vector and the variance-covariance vector.
Re region is intended to determine the color category areas belonging the the originating color vector, color originating color vector belongs categories
The re-region can be determined in a versatile form that does not depend on the type of device or viewing conditions.

According to a fifth aspect of the present invention, as the positional relationship between the transmission color vector and the center of gravity of each color category region of the transmission device , the transmission is performed from the center of gravity of all color categories given as the average of the center of gravity of all color category regions. is intended to use a normalized outgoing color distance divided by the Mahalanobis distance of the Mahalanobis distances to the color vector from the full color category centroid to each color category centroid, without breaking the continuity over the entire image, the gradation inversion of The mapping point of the received color vector can be determined without generating it.

According to a sixth aspect of the present invention, as the positional relationship between the received color vector and the center of gravity of each color category area of the receiving device , the reception is performed from the center of gravity of all color categories given as the average of the center of gravity of all color category areas. is intended to use a normalized received color distance divided by the Mahalanobis distance of the Mahalanobis distances to the color vector from the full color category centroid to each color category centroid, without breaking the continuity over the entire image, the gradation inversion of The mapping point of the received color vector can be determined without generating it.

[0038] A seventh aspect of the invention, in each of the source state representing the original state of the current state and the device that represents the current properties of the device input and output characteristics are changed by changing the viewing environment
The representative color is presented to the subject, and the perceived color for the representative color is determined.
Each of the above is determined by the response of the subject responding with a predetermined basic color name.
The color signal data corresponding to the representative color is classified, and the classification is performed.
Of each color space divided based on the color signal data group
Each area is used as a color category area to exchange device characteristic adjustment information between the same color category areas , so that the overall impression of the image does not differ between the original state and the current state in a single device. The color information can be exchanged, and it has the effect of uniquely adjusting the characteristics of a single device whose input / output characteristics have changed due to physical fluctuation or deterioration of the device or changes in observation conditions.

The eighth mode of the present invention is the same as the color category area to which the original state color vector, which is the color information of the original state, belongs.
From the color signal data group of the current state of being classified into one color category area it is intended to pick the current state color vector, the same color mosquito
Color information can be exchanged between categorized areas , and color information can be exchanged so that the overall impression of an image does not differ.

According to a ninth aspect of the present invention, the positional relationship that the original state color vector has with the center of gravity of each color category area in the original state is the positional relationship that the current state color vector has with the center of gravity of each color category area in the current state. as also stored in, which determines the current state color vector, color inversion over the entire image, it is possible to exchange color information so discontinuities have occurred.

The tenth aspect of the present invention, a color category region showing the highest occurrence probability in the source state color vectors the full color category area of the original state, the color signal data groups classified into each color category areas The color category with the highest probability of occurrence, determined by the probability density function given by the standard normal distribution formed by the mean vector and the variance-covariance vector
Is intended to determine the region to the color category region where the source state color vector belongs, the color former state color vector belongs catheter
The glue area can be determined in a versatile form that does not depend on observation conditions.

In an eleventh aspect of the present invention, the positional relationship between the original state color vector and the center of gravity of each color category region in the original state is calculated from the center of gravity of all color categories given as the average of the center of gravity of all color category regions. is intended to use the Mahalanobis distance to the source state color vector from the full color category centroid normalized source state color distance divided by the Mahalanobis distance to each color category centroid, without breaking the continuity over the entire image, the gradation Can be determined without inversion of the current state.

According to a twelfth aspect of the present invention, as a positional relationship between the current state color vector and the center of gravity of each color category area ,
The normalized current state color obtained by dividing the Mahalanobis distance from the center of gravity of all color categories to the current state color vector given by the average of the center of gravity of all color category areas by the Mahalanobis distance from the center of gravity of all color categories to the center of gravity of each color category. Since the distance is used , the mapping point of the current state color vector can be determined without breaking the continuity over the entire image and without inverting the gradation.

According to the thirteenth and sixteenth aspects of the present invention, the color categories are white, black, red, green, yellow, blue, purple, thigh,
Ash, brown or white, black, red, g
reen, yellow, blue, purple, pink, orange,
gray and brown, so 11 basic color names found by Berlin and Kay (reference: "Basic Color Ter
ms. Their Universality and Evolution ", Unive
By using rsity of California Press, Berkley, 1969), there is an effect that classification into color category regions can be executed efficiently without duplication.

In the fourteenth and seventeenth aspects of the present invention, the color signal data is an input color to an input device, an output value from an input device, an input value to an output device, or an output color from an output device. In a certain method, the color signal data can be described using a colorimetric value or a device signal.

The fifteenth and eighteenth aspects of the present invention are the methods in which the input color or the output color is CIELAB or CIELUV in a uniform color space, and CIELAB and CIELUV are linear in the human color difference sense scale and the space scale in the color space. Since they have a relationship, they have an effect of improving the accuracy of color information exchange. According to a nineteenth aspect of the present invention,
Multiple perceived colors by sending and receiving devices
The presented subject is assigned to multiple predetermined color categories.
Probability that the color signal data group of each classified perceived color is a population
The transmitting device and the receiving device according to a density function
Color space corresponding to the plurality of perceived colors
Area, and exchange color information between the same color category areas.
That is, The twentieth aspect of the present invention relates to a transmission device.
Color signal data of the chair and each color category of the transmitting device
Of each Mahalanobis distance with the color signal data group of
The color category with the minimum Mahalanobis distance
Determining a color category to which the color signal data belongs;
The color signal of the same color category as the judgment result in the device
Color signal data to be used as the received color
It is characterized by the following.

(Embodiment 1) In Embodiment 1, in a system for exchanging color information between devices, a group of colors perceived by an observer in a color space formed by color signal data for each of a transmitting device and a receiving device. To generate a probability density function having a population of color signal data groups classified into the same color category, and generating a transmission color vector, which is color information transmitted by the transmission device, from each color category. The color category to which the outgoing color vector belongs is determined by the occurrence probability that the outgoing color vector belongs, and from among the color signal data group of the receiving device generated from the same color category as the color category to which the outgoing color vector belongs, A color whose positional relationship with the center of gravity is such that the transmitted color vector is closest to the positional relationship with the center of gravity of each color category of the transmitting device. Exchanging color information picked out No. data as received color vector. Hereinafter, this color information exchange method will be described in detail.

FIG. 1 shows one embodiment of the present invention, C.
A procedure for displaying an image on an RT display on a liquid crystal display will be described.

First, the color display range of the CRT display and the liquid crystal display will be described. Fig. 2 (a) is a CRT
It is an example of a color display range (201) of a display and a color display range (202) of a liquid crystal display. CRT display data is NTSC standard (red primary color; (x, y) = (0.6
7, 0.33), green primary color; (x, y) = (0.21, 0.71), blue primary color; (x, y) = (0.14, 0.08)). The data on the LCD is described in “Nikkei Electronics no.57
0, 1992.12.21 ", p. 94, based on FIG. Overall CR
The T display has a wider color display range, and requires color gamut mapping when displaying an image on a CRT display on a liquid crystal display. Also, if the white point of both devices is different and adapted to the white of each device,
Compatibility of colorimetric values is not maintained, and corresponding color prediction is required.
A specific method of the present invention for solving these problems will be described.

FIG. 2 (b) shows the region of the color indicated by Kelly in FIG.
The color display range of the device shown in FIG. K
The color region indicated by elly divides colors in the entire visible light range into 23 color regions. From FIG. 2B, for example, a portion classified as “green” on both displays is extracted as shown in FIG. 3A. Reference numeral 301 denotes a color display range of a CRT display, and 302 denotes a color display range of a liquid crystal display. When the display ranges of both displays are divided as shown in FIG. 3B (303 is the same as 301, 304 is the same as 302), color gamut mapping is performed between black circles, white circles, black squares, and white squares of both devices. When executed so as to connect the colors, the continuity of the color change is maintained, and a natural correspondence is established. The present invention realizes natural color gamut mapping over the entire image by performing mapping as shown in FIG. 3B for each color region. Hereinafter, the mapping control method as shown in FIG. 3B will be described in detail.

FIG. 4 shows a color discrimination ellipse examined by MacAdam (references: “Basics of Color Engineering”, p.137, Mitsuo Ikeda
Author). 2 ° bisection (48cd / m2) with standard light source C
(24cd / m2) is a color discrimination ellipse determined by the presentation of a stimulus that surrounds it, indicating that all the colors in the ellipse look the same (however, in order to make the result easier to see, the size of the ellipse is 10 times the actual size). Doubled). from this result,
A group of colors, for example, a group of colors such as "green", "yellow green", and "blue green" used by Kelly can be expected to have an elliptical shape. The boundary of the color region obtained by Kelly can be interpreted as the intersection of the respective ellipses as shown in FIG. In FIG. 5, reference numeral 501 denotes a set of colors perceived as “green”, 502 denotes a set of colors perceived as “yellow green”,
Represents a group of colors perceived as "blue green", and 50
Reference numeral 4 denotes a boundary between “green” and “yellowish green” perception.
Represents the boundary between “green” and “bluish green” perception, and 506 represents the color display limit of the CRT display.

Based on the above concept, the present invention classifies the color space of both the CRT display and the liquid crystal display into a plurality of color categories according to perceived colors, and determines mapping points only between the same color categories. Therefore, first, both color spaces of the CRT display and the liquid crystal display must be divided into a plurality of color category regions by a perceived color like a color region indicated by kelly shown in FIG. The method requires versatility that does not depend on the type of device, observation conditions, and the like. In other words, the color region indicated by kelly is a result of a certain limited visual environment and an object, and it must be possible to describe this under arbitrary conditions.

Therefore, the representative color in the color space is presented to the subject, and classification into a plurality of color categories is performed, and a probability density function is created with the color signal data group classified into the same color category as a population. . The color category to which the transmission color (CRT display) or the reception color (liquid crystal display) belongs is the color category having the highest probability among the probabilities that the transmission color or the reception color occurs from each color category. A multidimensional normal distribution such as (Equation 2) is used for the probability density function. This corresponds to an ellipse of color cohesion, and the result selected according to the occurrence probability corresponds to the Kelly diagram of FIG. 2B.

[0054]

(Equation 2) Here, | Σ | is the determinant of the variance-covariance vector Σ, Σ
^-1 represents the inverse matrix of the variance-covariance vector Σ, respectively.
D ² is called the Mahalanobis distance means the distance from the center of gravity in consideration of the spread of the distribution. P represents the number of dimensions. Although the description has been made on the two-dimensional plane up to FIG. 5, since the color space is a three-dimensional space of (R, G, B), it is assumed that P = 3 in the present embodiment. X is a transmission color vector and is represented by (Equation 3).

(Equation 3) If the average vector μ is (Equation 4),

(Equation 4) The variance-covariance vector Σ is given by (Equation 5).

(Equation 5) Here, n represents the number of elements of the color signal data group. When (Equation 2) becomes the maximum, the Mahalanobis distance D ² becomes the minimum, so that the transmission color or the reception color becomes the Mahalanobis distance D.
² belongs to the smallest color category.

The subject experiment for obtaining the mean vector μ and the variance-covariance vector Σ is performed as shown in FIG. A test stimulus color patch (602) is presented on a CRT display or a liquid crystal display (601), and the subject (60
3) sees this and performs color naming using a plurality of basic color names (604 indicates a response operation for inputting the response of the subject to the computer). For the test stimulus, representative colors distributed at appropriate positions in the color space are used. The conditions for setting the representative color are as follows: All the set basic color names appear in the response from the subject. 2. Choose representative colors as evenly as possible in the color space. It is in.

Next, the following 11 basic color names found by Berlin and Kay are used as the basic color names (reference: "Ba
sic Color Terms. Their Universality and Evol
ution ", Univ. of California Press, Berkley,
1969).

1. white (shiro) 2. black (kuro) red (aka) green (midori) yellow (ki) 6. blue (ao) 7. purple (murasaki) pink (momo) 9. orange (daidai) 10. gray (hai) 11. brown (cha) These 11 basic color names were derived from the research of nearly 100 languages and the fact that there were 11 categorical basic color names common to all developed languages.
Crawford summarized the definitions of basic color names as follows (references: "Defining basic color terms", Anthrop
ol Linguist, 24, 338-343, 1982).

1. Be included in everyone's vocabulary. 2. Be used stably regardless of the person, regardless of when using it. 3. That the meaning is not included in other words. 4. It must not be used only for a specific object. Uchikawa also derived 11 basic categorical colors in Japanese (references: "Color Categorical Perception and Memory", 7th Color Engineering Conference, 1990).

From the above background, in the present embodiment, the above 11
Eleven color categories are set using the basic color names. However, the present invention is not limited to the present embodiment, and the color category setting method is arbitrary.

As described above, the colors of the CRT display and the colors of the liquid crystal display are classified by the probability density functions of the 11 color categories, and the gamut mapping is performed while maintaining the relative positional relationship between the same color categories according to the procedure shown in FIG. Execute Hereinafter, the color gamut mapping will be described in detail with reference to FIG.

An outgoing color vector C, which is the color of the CRT display, is represented by (Formula 6) (101 in FIG. 1).

(Equation 6) Which color category this outgoing color vector C belongs to,
Of Mahalanobis distance D ² which eleven color category has an originating color vector C, it determined to belong to the color category that gives the smallest D ^2. A color category j formed by the subject's response that responds to the perceived color of the representative color with 11 basic color names
Is represented by (Equation 7).

(Equation 7) Also the average vector alpha _j color category j and (8).

(Equation 8) Therefore, the variance-covariance vector S _j of the color category j _changes from (Equation 5) to (Equation 9).

(Equation 9) As described above, the Mahalanobis distance D ² _j of the color category j is given by (Equation 10).

(Equation 10) Among the 11 Mahalanobis distances obtained from (Equation 10), the color category Q that gives the minimum value is detected (1 in FIG. 1).
02). Since this color category Q is a color category to which the transmitted color vector belongs, the color data group Q ′ (103 in FIG. 1) classified into the color category Q on the liquid crystal display is a candidate for a mapping point. 'And expressed as (Equation 11).

[Equation 11] Here, m represents the total number of color data classified into the color category Q of the liquid crystal display. The candidate color vector Q 'is obtained by calculating the probability that all the colors of the light emitted from the liquid crystal display will be generated from each color category. That is, of all the emission colors of the liquid crystal display, the color data group that minimizes the Mahalanobis distance of the color category Q is the candidate color vector Q ′.
Becomes Since the candidate color vector Q 'is selected according to the color category to which the transmission color vector belongs, it is necessary to make a database of the candidate color vectors Q' for each of the 11 color categories of the liquid crystal display in advance.

As described above, by determining the candidate color vector Q ′ in (Equation 11), it is guaranteed that the perceptual category of the color displayed on the liquid crystal display is the same as that of the CRT.

Next, an algorithm for extracting the received color vector C 'from the candidate color vectors Q' will be described with reference to FIGS. For convenience of notation, it is expressed in two dimensions.

The rules for mapping the transmission color vector to the reception color vector are as shown in FIG. 3 (b), and colors that match categorically in the same color category area are mapped. For example, a black circle belonging to a green area (303) of the CRT display in FIG. 3B can be categorically described as "a color at the boundary between green and yellowish green and of which the color saturation is the highest". In order to perform mapping while maintaining such a categorical meaning, a search is made for “a color at the boundary between green and yellowish green and of which the saturation is the highest” on the liquid crystal display. This is exactly the black circle that belongs to the green area (304) of the liquid crystal display. Similarly, in both displays, the black square is “the color of the border between green and bluish green, and the color with the highest saturation”, and the open circle is “the color of the border between green and yellowish green, And the white square is the color of the border between green and bluish green and the color with the lowest saturation, and both points are categorically matched. .

In this case, it can be interpreted that a point obtained by averaging four color data of a black circle, a black square, a white circle, and a white square is categorically identical between the two displays. Further, when the four color data are expanded to all color data groups constituting the color category, if the number of data is sufficient, it can be interpreted that these averaged points are categorically identical between the two displays.

Therefore, as shown in FIG. 7, the average vectors are mapped between the same color categories. In FIG. 7, a point G (701) represents the center of gravity given as an average value of the color data group classified as green on the CRT display, and a point YG (702) represents a color data group classified as yellowish green on the CRT display. And a point BG (703) represents a center of gravity given as an average value of a group of color data classified as bluish green on a CRT display. On the other hand, the point G '(704) represents the center of gravity given as the average value of the color data group classified as green on the liquid crystal display, and the point YG' (70)
5) represents the center of gravity given as the average value of the color data group classified as yellowish green on the liquid crystal display, and point B
G ′ (706) represents the center of gravity given as the average value of the color data group classified as bluish green on the liquid crystal display. As described above, the centroids of the 11 color categories are mapped between the two displays.

Next, the center of gravity of the three color categories shown in FIG. 7 will be considered. FIG. 8A is the same as FIG.
A point O (801) is added to the RT display, and a point O '(802) is added to the liquid crystal display. Point O is the center of gravity G, Y of the three color categories of the CRT display.
The center of gravity of the triangle G-YG-BG formed by G and BG. Similarly, the point O 'is a triangle G'- formed by the centers of gravity G', YG ', and BG' of the three color categories of the liquid crystal display.
It is the center of gravity of YG'-BG '. As discussed in FIG.
Since the two centroids are points mapped to each other between the two displays, it can be said that the points O and O ′ also have a mapping relationship.

The mapping between the centers of gravity has been described based on FIG. 3B. However, as shown in FIG. 8B, no matter where the centers of gravity of the transmitting device and the receiving device are located, By mapping the centroids in the same color category with each other, a categorical relationship can be preserved. By the way, the point O (803) and the point O '(804) in FIG. 8B are points to be mapped, and at this time, the line segments LO, M
It can be said that the lengths of O, NO, L'O ', M'O', and N'O 'have the same relationship in the sense of establishing a categorical match. Therefore, in order to handle the color gamut mapping that preserves the categorical relationship, it is convenient to perform normalization based on the distance from each color category connected to the points O and O ′ in FIG. 8B.

It is now assumed that the transmitted color C has occurred at the position of the point O (803). At this time, the distance between the outgoing color C and the 11 color categories is represented by (Equation 12).

(Equation 12) Here, D ² _Cj represents the Mahalanobis distance from the center of gravity of the color category j to the transmission color C. If the distance of 11 color categories from point O is (Equation 13),

(Equation 13) Since the transmission color C is on the point O, it becomes (Equation 14) (Equation 15).

[Equation 14]

(Equation 15) Then, in the receiving device, the received color vector C ′ and the center of gravity of the 11 color categories have a Mahalanobis distance represented by (Equation 16),

(Equation 16) When the point O ′ has the center of gravity of 11 color categories and the Mahalanobis distance represented by (Equation 17),

[Equation 17] (Equation 18) as in (Equation 15)

(Equation 18) Is satisfied, the received color vector C ′ must be a vector
Must be identical to O '. Therefore, the reception color C ′ is
It is determined to be on O ′ (804). Where point O and point
The center of gravity of the eleven color categories such as O 'is called the all-color category center of gravity, and its vector is called the all-color category average vector. Further, as in (Equation 15), the Mahalanobis distance (D ² cj in (Equation 15)) from the center of gravity of the color category to the transmission color is the Mahalanobis distance ((Equation 15)) from the center of gravity of the same color category to the center of gravity of all color categories. ) Is D ² oj)
Is referred to as a normalized transmission color distance. The same applies to the receiving device.

Based on the above concept, the general case of FIG. 8C is extended. First, consider the left side of FIG. The point P (805) is the position of the transmitted color, and the distances from the centers of gravity L, M, and N of the three color categories are respectively

[Equation 19]

(Equation 20)

(Equation 21) It is. Where P is the transmitted color vector, α _L is the average vector of color category L, α _M is the average vector of color category M, α _N is the average vector of color category N, and S ⁻¹ _L is the variance-covariance of color category L The inverse matrix of the matrix, S ^-1 _M indicates the inverse matrix of the variance-covariance matrix of the color category M, and S ^-1 _N indicates the inverse matrix of the variance-covariance matrix of the color category N. The distance from the center of gravity L, M, N of the three color categories to the center of gravity O of all color categories is respectively

(Equation 22)

(Equation 23)

(Equation 24) It is. Therefore, the normalized transmission color distance K _PL of the point P from the color category L is

(Equation 25) And the normalized transmission color distance K _PM of the point P from the viewpoint of the color category M is

(Equation 26) And the normalized transmission color distance K _PN of the point P as viewed from the color category N is

[Equation 27] Becomes

Next, consider the receiving device side on the right side of FIG. The point P ′ (806) is the position of the receiving color, L ′ is the center of gravity of the same color category as the color category L of the transmitting device, and M ′ is the center of gravity of the same color category as the color category M of the transmitting device. , N ′ are the centroids of the same color category as the color category N of the transmitting device. L 'at point P',
The distance from M ', N'

[Equation 28]

(Equation 29)

[Equation 30] It is. Here, P 'is the received color vector, α _L ' is the average vector of color category L, α _M 'is the average vector of color category M, α _N ' is the average vector of color category N, and S _L '
^-1 is the inverse of the variance-covariance matrix of the color category L, S _M ' ^-1
Denotes an inverse matrix of the variance-covariance matrix of the color category M, and SN′-1 denotes an inverse matrix of the variance-covariance matrix of the color category N. The distance from the center of gravity L ', M', N 'of three color categories to the center of gravity O' of all color categories is respectively

(Equation 31)

(Equation 32)

[Equation 33] It is. Therefore, the point P ′ viewed from the center of gravity L ′ of the color category L
The normalized outgoing color distance K _PL '

(Equation 34) From the color category M, the normalized transmission color distance K _PM 'of the point P is

(Equation 35) And the normalized transmission color distance K _PN ′ of the point P as viewed from the color category N is

[Equation 36] Becomes

As defined in FIG. 8B, the establishment of categorical matching means that the normalized transmitted color distance and the normalized received color distance of the same color category match. Therefore, in the case of FIG. 8 (c), the received color vector is represented by (Equation 37) for the color category L.

(37) About color category M (Equation 38)

(38) About color category N (Equation 39)

[Equation 39] Must be satisfied.

As described above, the operation for extracting the received color vector C ′ from the candidate color vectors Q ′ is

(Equation 40) For all 11 color categories is equivalent to searching for a color vector that can be satisfied with the smallest possible error from the candidate color vectors Q ′. Here, Kcj indicates the normalized transmission distance when the transmission color vector C is viewed from the color category j, and Kcj 'indicates the normalized reception distance when the reception color vector C' is viewed from the color category j. Therefore, (Equation 41) is used for the error evaluation formula.

[Equation 41] However, in the color category j (Equation 42)

(Equation 42) In the condition of

[Equation 43] Is replaced by (Equation 44).

[Equation 44] Among the candidate color vectors Q ′, a color vector that minimizes (Equation 41) is defined as a received color vector C ′.

FIG. 1 shows a portion for executing the algorithm for extracting the received color vector C ′ from the candidate color vector Q ′.
It explains in. This operation is divided into the following four steps. Step 1. The average vector of all color categories is calculated on both the CRT display and the liquid crystal display (1 in FIG. 1).
04, 105). Step 2. In the CRT display, the average vector and the variance-covariance vector of the color category Q (106 in FIG. 1),
Using the transmitted color vector (101 in FIG. 1) and the average vector of all color categories (104 in FIG. 1), the normalized transmitted color distance (FIG. 1)
107) is calculated. Step 3. In a liquid crystal display, normalization is performed using an average vector and a variance-covariance vector (108 in FIG. 1) of a color category Q, a candidate color vector Q ′ (103 in FIG. 1), and an average vector of all color categories (105 in FIG. 1). The receiving color distance (109 in FIG. 1) is calculated. Step 4.1. CRT for each of the 11 color categories
The normalized transmission color distance and the normalized reception color distance are compared between the display and the liquid crystal display (110 in FIG. 1), and the color vector with the smallest distance error is selected from the candidate color vectors Q ′, and this is selected from the liquid crystal display. A drive signal C ′ for the display is set (111 in FIG. 1).

In step 1, the average vector O of all color categories of the CRT display is

[Equation 45] Given by n is the number of color categories, and in this embodiment, n = 11. In addition, the average vector O 'for all color categories of the LCD is

[Equation 46] Given by n is the number of color categories, and in this embodiment, n = 11.

In step 2, the normalized transmission color distance Kcj is

[Equation 47] Given by Here, Dcj represents the Mahalanobis distance from the center of gravity of the color category viewed from the color category j to the transmission color C, and Doj represents the Mahalanobis distance from the center of gravity of the color category viewed from the color category j to the center of gravity of all color categories.

In step 3, the normalized received color distance Kcj '
Is

[Equation 48] Given by Here, Dcj ′ represents the Mahalanobis distance from the center of gravity of the color category viewed from color category j to the received color C ′, and Dco ′ represents the Mahalanobis distance from the center of gravity of the color category viewed from color category j to the center of gravity of all color categories. Express.

In step 4, the errors between the normalized transmitted color distance and the normalized received color distance are evaluated using the following equations (41) to (44).

From the above, the color (r,
g, b) is mapped to the color (r ', g', b ') on the LCD monitor, since the two pairs have the same color category and the same relative positional relationship within the color category. Color gamut mapping can be performed with natural colors. Furthermore, the above method has uniqueness not found in the conventional method, and it is possible to effectively implement the design of a color reproduction system of an image device and software and the color management at each user site.

When the CRT display and the liquid crystal display are considered as two color charts observed under different illumination light sources, they can be used for corresponding color prediction. It can be easily understood that the characteristics can be used for adjusting the characteristics of the device when viewed as two characteristics of a single device.

That is, when the current characteristics of a single device whose input / output characteristics have changed due to a change in the state are restored to the original characteristics, the observer sets the color space formed by the color signal data for each of the current state and the original state. The perceived color group is divided into a plurality of color categories, and a probability density function with a population of color signal data groups classified into the same color category is created, and an original state color vector which is color information of the original state Determines the color category to which the original state color vector belongs according to the occurrence probability of occurrence from each color category, and from the current state color signal data group generated from the same color category as the color category to which the original state color vector belongs. Color signal data such that the positional relationship between the current state color category and the center of gravity of each color category is closest to the positional relationship that the original state color vector has with the center of gravity of each color category in the original state. Is selected as the current state color vector.

(Embodiment 2) Next, a procedure for outputting a color of a CRT display to a four-color printer will be described as Embodiment 2 with reference to FIG.

The color (r, g, b) of the CRT display is set to 4
When converting to color (c, m, y, k) of a color printer,
As shown in FIG. 0, it can be seen that color gamut mapping is required from the difference in the color display range of both devices. FIG.
Is an example of the color display range of the NTSC standard and the color reproduction range of the inkjet printer, and is data actually measured by the present inventors. FIG. 10A is a diagram in which the colorimetric values of the representative colors of the CRT display and the ink jet printer are projected on the a * -b * plane of the CIELAB space, and FIG. FIG. 10C is a diagram in which colorimetric values are projected on an L * -a * plane in a CIELAB space, and FIG.
FIG. 4 is a diagram in which colorimetric values of representative colors of a display and an ink jet printer are projected on an L * -b * plane in a CIELAB space.

In FIGS. 10A to 10C, white squares indicate primary colors (C, M, Y) and secondary colors (C + M,
M + Y, Y + C), which are representative colors of gray. The white triangle is NT
Primary color of CRT display (R, G,
B) Secondary colors (R + B, R + G, G + B) and gray representative colors. Further, the black dot is the representative color 72 of the ink jet printer.
Nine colors are selected from the entire color display range of the ink jet printer. Reference numeral 1001 in FIG. 10 represents the outer shape of the color display range of the CRT display.
02 represents the outer shape of the color display range of the ink jet printer.

The color gamut mapping in this embodiment is a mapping between a three-dimensional space and a four-dimensional space. Since the number of dimensions is different, the method used in “Embodiment 1” cannot be directly applied. Therefore, the number of dimensions of the printer is reduced to 3 by the method shown in FIG.
Perform mapping by dropping into dimensions. First, a representative color of the printer is selected from the CMY three-dimensional space (11 in FIG. 11).
01). The selected color vector E becomes an element of the color signal data group of each color category, and is used to describe the probability density function of the color category.

On the other hand, the color stimulus (1102 in FIG. 11) presented to the test subject generates black and removes undercolor in the color vector E (FIG. 11).
1103)

[Equation 49] The printer is created by driving the printer with the color vector EP (1104 in FIG. 11). In (Equation 49), min (c,
m, y) represents a function for detecting the minimum value of c, m, y.
ρ, τc, τm, τy, φ are arbitrary constants, respectively.
(Equation 49) is merely an example of black generation and undercolor removal, and any black generation method and undercolor removal method can be applied to the present invention.

By the above operation, the probability density function formed in the printer space can be described in three dimensions.
The procedure of the color gamut mapping shown in FIG. 9 is basically the same as that of FIG. However, after the reception color vector C ′ is obtained,
As shown by 901 in FIG. 9, black generation and under color removal are performed.
Then, black generation and under color removal in 901 in FIG.
Must be the same as 103. An output 902 of the 901 is a mapping point.

As described above, it can be understood from the present embodiment that color gamut mapping, corresponding color prediction, and device characteristic adjustment can be performed even with a four-color printer having a different color space from the color space of the CRT display.

The color information exchange method and the device characteristic adjustment method of the “first embodiment” and the “second embodiment” have been described using device-dependent color signals such as RGB and CMYK. By modeling the relationship between device signals and colorimetric values, a similar embodiment can be implemented in the colorimetric space. For example, in the NTSC standard

[Equation 50]

(Equation 51) Thus, conversion between RGB signals and XYZ tristimulus values is possible. For input devices such as cameras,
The captured RGB signals can be converted to XYZ tristimulus values, and C
In an output device such as an RT display, the color XYZ of the color displayed by the driven RGB signal is expressed by (Equation 51).
Can be converted to tristimulus values. If it can be converted to XYZ tristimulus values
Since it is possible to convert to other colorimetric value spaces such as CIELAB and CIELUV, it is easy to understand that the color space can be arbitrarily selected.
In addition to finding a function that connects the device signal and the colorimetric value as in (Equation 50) and (Equation 51), all conversion information is directly tabulated, conversion information is obtained for representative color data, and all the conversion information is obtained by interpolation. If the relationship between the device signal and the colorimetric value is determined by compensating the conversion information, color information exchange and device characteristic adjustment can be performed in the colorimetric value space.

When the mapping based on the device signal as in the first and second embodiments is compared with the mapping based on the colorimetric values, it is understood that the first and second embodiments have a feature that the colorimetric operation is not required.

[0091]

As described above, according to the present invention, the color gamut mapping can be determined with uniqueness not found in the conventional method, and the color reproducibility of an image can be improved between devices having different viewing conditions. An excellent color information exchange method can be realized.

Further, according to the present invention, an excellent device characteristic adjusting method which can determine a device characteristic adjusting method with uniqueness not found in the conventional method and can realize improvement of color reproducibility of images having different observation conditions and the like. Can be realized.

[Brief description of the drawings]

FIG. 1 is a diagram showing a procedure for converting a color of a CRT display to a color of a liquid crystal display according to the first embodiment of the present invention.

2A is a diagram showing an example of a color display range of a CRT display and a liquid crystal display. FIG. 2B is a diagram showing a color display range of a CRT display and a liquid crystal display superimposed on a color region indicated by Kelly.

FIG. 3 (a) is a diagram showing a “green” region of a CRT display and a liquid crystal display cut out in FIG. 2 (b). (B) A color gamut that realizes natural hue, continuity preservation, and no inversion. Diagram explaining the concept of the mapping method

FIG. 4 is a diagram showing a color discrimination ellipse obtained by MacAdam.

FIG. 5 is a view for explaining a relationship between an intersection line of an ellipse and a boundary line of a Kelly color region.

FIG. 6 is a diagram for explaining a subject experiment method of collecting color signal data of each color category.

FIG. 7 is a view for explaining a method of mapping the center of gravity of the same color category between the CRT display and the liquid crystal display.

FIG. 8A illustrates a method of mapping the center of gravity of a triangle formed from the centers of gravity of three color categories. FIG. 8B illustrates a general method of mapping the center of gravity of a triangle formed from the centers of gravity of three color categories. (C) A diagram illustrating a method for mapping an arbitrary outgoing color using the normalized outgoing color distance and the normalized receiving color distance

FIG. 9 is a diagram showing a procedure for converting a color of a CRT display to a color of a four-color printer according to the second embodiment of the present invention.

FIG. 10 is a diagram illustrating an example of a color display range of a CRT display and a four-color printer.

FIG. 11 is a view for explaining a method of handling a four-dimensional color space of a four-color printer as a three-dimensional color space;

FIG. 12 is a diagram showing a mapping method for determining a mapping point while keeping brightness and hue constant and decreasing saturation.

FIG. 13 is a diagram showing a mapping method for determining a matching point while smoothly compressing all points aiming at a center of gravity of a color space and avoiding a decrease in saturation.

[Explanation of symbols]

101 Outgoing color vector C 102 Color category identification processing to which outgoing color vector C belongs 103 Candidate color vector 104 All color category average vector of CRT display 105 All color category average vector 106 Liquid crystal display average vector and covariance vector supply operation 107 Normal Transmission color distance calculation unit 108 Average vector and covariance vector of liquid crystal display 109 Normalized reception color distance calculation unit 110 Error evaluation unit 111 Receiving color vector C ′ 201, 203 Color display range of CRT display (NT
SC standard) 202, 204 Color display range of liquid crystal display 301, 301 Range corresponding to Kelly's green area of CRT display 302, 304 Range corresponding to Kelly's green area of liquid crystal display

Continuation of the front page (56) References JP-A-9-266537 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H04N 1/46 H04N 1/60

Claims (20)

(57) [Claims] 1. Each color for each transmitting device and receiving device
The representative color in the space is presented to the subject, and the
Based on the response of the subject who responded to the perceived color with a predetermined basic color name
The color signal data corresponding to each of the representative colors is classified.
Each of the above-mentioned divided based on the classified color signal data group
Using each area of the color space as a color category area, a color category area to which the color signal data of the transmitting device belongs is determined, and the color signal data is mapped to a color category area of the same color category in the color space of the receiving device. Color information exchange method. 2. A receiving color vector is selected from a color signal data group of a receiving device classified into a color category region of the same color category as a color category region to which a transmitting color vector of color information transmitted by a transmitting device belongs. The color information exchange method according to claim 1, wherein 3. The receiving means so that the positional relationship between the transmitting color vector and the center of gravity of each color category region of the transmitting device is preserved also in the positional relationship between the receiving color vector and the center of gravity of each color category region of the receiving device. 3. The color information exchange method according to claim 2, wherein a color vector is selected. 4. An average vector and a variance-covariance vector of a color signal data group classified into each color category area, wherein a color category area in which a transmission color vector has the highest occurrence probability among all color category areas of the transmission device is determined. 4. A color category region having the highest occurrence probability is determined as a color category region to which the transmitted color vector belongs, obtained by a probability density function given by a standard normal distribution formed by: The color information exchange method described in any of them. 5. The Mahalanobis distance from the center of gravity of all color categories, which is given as the average of the center of gravity of all color category areas, to the transmission color vector, as the positional relationship between the transmission color vector and the center of gravity of each color category area of the transmission device. 4. The color information exchange method according to claim 3, wherein a normalized transmission color distance divided by a Mahalanobis distance from the all-color category center of gravity to each color category center of gravity is used. 6. The Mahalanobis distance from the center of gravity of all color categories, which is given as the average of the centers of gravity of all color category areas, to the received color vector as a positional relationship between the received color vector and the center of gravity of each color category area of the receiving device. 6. The color information exchange method according to claim 3, wherein a normalized received color distance divided by a Mahalanobis distance from the all-color category center of gravity to each color category center of gravity is used. 7. A representative color is presented to a subject in each of a current state representing a current characteristic of a device whose input / output characteristics change due to a change in an observation environment and an original state representing an original state of the device.
Then, the perceived color for the representative color is determined by a predetermined basic color name.
A color signal corresponding to each of the representative colors is obtained according to the response of the subject.
Signal data and classify them based on the classified color signal data group.
Each area of each color space is divided into color category areas.
A device characteristic adjustment method characterized by exchanging device characteristic adjustment information between regions of the same color category as regions . 8. Selecting a current state color vector from a current state color signal data group classified into a color category area of the same color category as a color category area to which an original state color vector, which is color information of the original state, belongs. The device characteristic adjusting method according to claim 7, wherein 9. The positional relationship that the original state color vector has with the center of gravity of each color category area in the original state is also preserved in the positional relationship that the current state color vector has with the center of gravity of each color category area in the current state. 9. The device characteristic adjusting method according to claim 8, wherein a current state color vector is determined. 10. A color category area in which an original state color vector has the highest probability of occurrence among all color category areas in an original state is defined as an average vector and a variance covariance of a color signal data group classified into each color category area. The color category area having the highest occurrence probability is determined as a color category area to which the original state color vector belongs, obtained by a probability density function given by a standard normal distribution formed by the vectors. 10. The method for adjusting device characteristics according to any one of claims 9 to 9. 11. A Mahalanobis from an all-color category centroid point to an original state color vector given as an average of centroid points of all color category areas as a positional relationship between the original state color vector and the centroid point of each color category area in the original state. 10. The device characteristic adjusting method according to claim 9, wherein a normalized original state color distance obtained by dividing a distance by a Mahalanobis distance from the center of gravity of each color category to the center of gravity of each color category is used. 12. The Mahalanobis distance from the center of gravity of all color categories, which is given as the average of the center of gravity of all color category areas, to the current state color vector, as a positional relationship between the current state color vector and the center of gravity of each color category area. 12. The device characteristic adjustment method according to claim 9, wherein a normalized current state color distance divided by a Mahalanobis distance from the all-color category centroid to each color category centroid is used. 13. The color signal data of the transmitting device and the receiving device is obtained by subject experiments (white, black, red, green, yellow, blue,
Purple, thigh, orange, ash, brown) or (white, black,
red, green, yellow, blue, purple, pin
7. The color information exchange method according to claim 1, wherein color categories classified into basic colors (k, orange, gray, and brown) are used. 14. The image processing apparatus according to claim 1, wherein the color signal data is an input color to an input device, an output value from the input device, an input value to the output device, or an output color from the output device. The color information exchange method described in any of them. 15. An input color or an output color in a uniform color space.
15. The color information exchange method according to claim 14, wherein the color information is CIELAB or CLELUV. 16. The color signal data of the original state and the current state of a single device are obtained by subject experiments (white, black, red, green, yellow,
Blue, purple, peach, orange, ash, brown) or (white, bla
ck, red, green, yellow, blue, purple,
13. The device characteristic adjusting method according to claim 7, wherein a color category classified into basic colors (pink, orange, gray, and brown) is used. 17. The image processing apparatus according to claim 7, wherein the color signal data is an input color to the input device, an output value from the input device, an input value to the output device, or an output color from the output device. The device characteristic adjusting method according to any one of the above. 18. An input color or an output color in a uniform color space.
The device characteristic adjusting method according to claim 17, wherein the device characteristic is CIELAB or CLELUV. 19. A probability density function in which a subject presented with a plurality of perceived colors by a transmitting device and a receiving device has a population of color signal data groups of each perceived color classified into a plurality of predetermined color categories, A color information exchange method, wherein each color space of the transmitting device and the receiving device is divided into color category regions corresponding to the plurality of perceived colors, and color information is exchanged between the same color category regions. 20. Calculate and compare each Mahalanobis distance between the color signal data of the transmission device and the color signal data group of each color category of the transmission device, and determine the color category having the minimum Mahalanobis distance as the color signal data. 20. The color information exchange method according to claim 19, wherein a color category to which the received device belongs is determined, and color signal data to be a received color is determined from a color signal data group of the same color category as the determination result in the receiving device.