JP2003256823A - Image processing method and device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processing method and device capable of realizing highly precise color reproduction even when an observation environment for an input image and the observation environment of psychological evaluation test and an observation environment for an output image are different from each other. <P>SOLUTION: This image processing method is realized by an image processor configured of a rgb value storing part 1, a converting part 2 for converting the rgb value into an XYZ value, a first color sight model processing part 3, a converting part for converting the XYZ value into an L*C<SB>ab</SB>*h<SB>ab</SB>* value, a fuzzy image processing part 5, a converting part 6 for converting the L*C<SB>ab</SB>*h<SB>ab</SB>* value into the XYZ value, a second color sight model processing part 7, a converting part 8 for converting the XYZ value into the rgb value, a converting part 9 for converting the rgb value into a CMYK value, an observation environment parameter storing part 10 of an input image, an observation environment parameter storing part 11 of a psychological evaluation test, and an observation environment parameter storing part 12 of an output image. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing method for color copying and the like and an apparatus for performing the same.

[0002]

2. Description of the Related Art In a conventional color hard copy system, the tristimulus values of the original color and the reproduced color do not match,
In addition to color reproduction that matches the appearance of colors (corresponding color reproduction), we also considered color reproduction (preferred color reproduction) that reproduces memory colors such as skin color, blue sky color, and grass color as people would like. Image processing is being performed.

As a preferable color reproduction method of a skin color in a color copy, a quantification theory I is used for analysis of a psychological evaluation experiment result of an SD (Semantic Differential) scale, and then a skin color area is extracted and a color is extracted using a fuzzy theory. A method of making adjustments has been proposed. The execution procedure of this color reproduction method will be described with reference to FIG. First, the color original image is read from the scanner unit and the rgb value of the color image is calculated. Next, from the rgb value, the XYZ value, L ^* a ^* b ^* value, and L ^* a ^* b ^* standardized by CIE (International Commission on Illumination) ^.
The values are sequentially converted into L ^* _Cab ^* _hab ^* values which are the cylindrical coordinate system of the color space. _{^{L * C a b * h ab}} * value is a value indicating brightness, saturation, and hue angle in the L ^* a ^* b ^* color space.

Next, in the fuzzy image processing unit, the L
^A preferable color reproduction process is performed on the ^* _Cab ^* _hab ^* values. Then, the L ^* _Cab ^* _hab ^* value after this processing is converted again into the L ^* a ^* b ^* value, the XYZ value, and the rgb value. The rgb value is input to the color printer unit as a CMYK color signal, and a color print image is output.

According to this method, different L ^* C _ab ^* h _ab ^* is set in advance ^.
The SD scale is obtained by conducting a psychological evaluation experiment for a plurality of skin color patches represented by values, and the SD scale evaluation value is quantified using the L ^* _Cab ^* _hab ^* value of the patch as an explanatory variable. L ^* that gives a favorable impression when analyzed by theory I
Extract the range of C _ab ^* h _ab ^* values. Furthermore, a membership function is set for each skin color distribution region of each L ^* _Cab ^* _hab ^* value from this extraction result, and fuzzy rules are created in consideration of the characteristics of each region. Fuzzy reasoning
Input the pixels L ^* C _ab ^* h _ab ^* value as the input to the output at the pixel L ^* C _ab ^* h _ab ^* value and color adjustment amount ΔL ^* ΔC _ab ^*
It is the sum of Δh _ab ^* . However, ΔL ^* , ΔC _ab ^* , Δh
_ab ^* is expressed by the following equation.

[0006]

[Equation 1]

Μ _Li (L ^* ), μ _Ci (C _ab ^* ), μ _hi (h _ab
^*) Is L ^* to the i-th rule, respectively, C _ab ^*, the goodness-of-fit of the ^{_{h ab *, ΔL * i,}} L by ^{_{ΔC ab * i, Δh ab *}} i Each of the i-th rule ^*, C _ab ^*, h _ab ^* output value. In addition,
It is the fuzzy image processing unit shown in FIG. 16 that performs color adjustment by this fuzzy inference.

By going through such a procedure, it is possible to obtain an image that gives a more favorable impression than the original image in skin color. Further, by using this method for a memory color other than the skin color, it is possible to obtain an image that gives a more preferable impression than the original image even for the memory color other than the skin color.

[0009]

However, the above-mentioned prior art methods do not sufficiently consider the observation environment, which is an important factor for color appearance. Therefore, when the observation environment of the input image and the observation environment in which the psychological evaluation experiment is performed are different, there is a possibility that the accuracy of preferable color reproduction may be reduced.

When the observation environment of the input image and the observation environment in which the psychological evaluation experiment is performed are different, generally, even in the same image, the color appearance is different between the observation environments. Accordingly, although made under these different viewing environments require chromatic adaptation transform to equalize the appearance of color, the are based in a conventional manner L ^* a ^*
It is known that the b ^* color space does not describe an accurate chromatic adaptation transformation, and the description estimation accuracy of the color appearance is reduced particularly in a hue change and a high saturation region. Therefore, in the conventional method,
When the observation environment is different, since the fuzzy image processing is executed after the conversion with the low estimation accuracy of the color appearance description is executed,
It is considered that the accuracy of preferable color reproduction is reduced. The conventional method presupposes that a preferable color reproduction is an observation environment for an input image, an observation environment in which a psychological evaluation experiment is performed,
It is limited to the case where the three viewing environments of the output image match.

The present invention has been made in view of the above problems of the prior art, and its purpose is to provide an observation environment for an input image, an observation environment for a psychological evaluation experiment,
Even if the observation environment for the output image is different,
An object of the present invention is to provide an image processing method and apparatus capable of highly accurate color reproduction.

[0012]

In order to achieve the above object, an image processing method of the present invention comprises a step of acquiring rgb signal values of three primary colors of red, green and blue from an original image, and the rgb signal value CIEXYZ color system XYZ value conversion step, and CIE recommends the XYZ value as a uniform perceptual color space L ^* a ^* b ^* L ^* _Cab ^* _hab ^* which is a cylindrical coordinate system of the color space ^.
A step of converting into a value, a step of performing a fuzzy process on the L ^* C _ab ^* h _ab ^* value, and a L ^* C _ab ^* h _ab after the fuzzy process
^{* The} step of converting the value into the XYZ value of the CIEXYZ color system, the step of converting the XYZ value into the rgb value, and the step of converting the r
an image processing method including the step of reproducing an image using gb values, the first color appearance model processing step of giving a predetermined parameter to the XYZ values converted from the rgb values, and the fuzzy processing. A second color appearance model processing step of imparting predetermined parameters to the XYZ values converted from the subsequent L ^* C _ab ^* h _ab ^* values.

The parameters provided in the first color appearance model processing step are parameters that suggest the observation environment of the input image and parameters that suggest the observation environment in which the psychological evaluation experiment was performed. The parameters given in the appearance model processing step are the parameters suggesting the observation environment in which the psychological evaluation experiment was performed and the parameters suggesting the observation environment of the output image.

Further, the image processing method of the present invention comprises a step of obtaining rgb signal values of three primary colors of red, green and blue from an original image and a step of converting the rgb signal values into XYZ values of the CIEXYZ color system. And a color appearance model order conversion step of calculating the lightness, saturation, and hue angle values of the CIECAM97 _S model by assigning predetermined parameters to the XYZ values, and predetermined values for the lightness, saturation, and hue angle values. The process of applying the fuzzy process with the parameter of
CIE is performed by assigning certain parameters to the values of saturation and hue angle.
A color appearance model inverse conversion step of converting into an XYZ value of the XYZ color system, and a step of converting this XYZ value into an rgb value,
It is also possible to perform the process of reproducing an image using this rgb value.

The parameters given in the color appearance model order conversion step are parameters that suggest the observation environment of the input image, and the parameters given in the step of performing the fuzzy processing are the observations obtained through a psychological evaluation experiment. The parameter that suggests the environment and that is given in the color appearance model inverse conversion step is a parameter that suggests the observation environment of the output image.

The image processing apparatus of the present invention uses the
Means for acquiring rgb signal values of the three primary colors of green and blue;
Converts rgb signal value to XYZ value of CIEXYZ color system
And the XYZ values as CIE's uniform perceptual color space.
Recommended L^*a^*b^*L, which is the cylindrical coordinate system of the color space^*C_ab
^*h_ab ^*A means to convert to a value and this L^*C_ab ^*h_ab ^*Against value
Means for performing fuzzy processing and L after fuzzy processing^*C
_ab ^*h_ab ^*Convert values to XYZ values in the CIEXYZ color system
Means and means for converting this XYZ value to an rgb value,
And means for reproducing the image using the rgb value of
In the image processing device, the converted rgb value is converted.
The first parameter for giving a predetermined parameter to the XYZ value
Color appearance model processing means and L after the fuzzy processing^*
C_ab ^*h_ab ^*The specified parameters are converted to XYZ values.
And a second color appearance model processing means for imparting
It was

The parameters provided by the first color appearance model processing means are parameters that suggest the observation environment of the input image and parameters that suggest the observation environment in which the psychological evaluation experiment was conducted, and the second color. The parameters given by the appearance model processing means are a parameter indicating the observation environment in which the psychological evaluation experiment is performed and a parameter indicating the observation environment of the output image.

The image processing apparatus of the present invention further comprises means for obtaining rgb signal values of the three primary colors of red, green and blue from the original image and means for converting the rgb signal values into XYZ values of the CIEXYZ color system. And a color appearance model order conversion means for calculating the lightness, saturation, and hue angle values of the CIECAM97 _S model by adding predetermined parameters to the XYZ values, and predetermined values for the lightness, saturation, and hue angle values. The method of applying the fuzzy processing with the parameter of
CIE is performed by assigning certain parameters to the values of saturation and hue angle.
Color appearance model inverse conversion means for converting into XYZ values of the XYZ color system, and means for converting this XYZ value into rgb values,
And a means for reproducing an image using this rgb value.

The parameter given by the color appearance model order conversion means is a parameter indicating the observation environment of the input image, and the parameter given by the means for performing the fuzzy processing is the observation obtained by conducting a psychological evaluation experiment. The parameter that suggests the environment and is given by the color appearance model inverse transforming means is a parameter that suggests the observation environment of the output image.

[0020]

BEST MODE FOR CARRYING OUT THE INVENTION The present invention will now be described in detail based on the illustrated embodiment. <First Embodiment> FIG. 1 is a block diagram showing the schematic arrangement of an image processing apparatus according to the present embodiment. The image processing apparatus according to the present embodiment includes an rgb value storage unit 1, an rgb value to XYZ value conversion unit 2, a first color appearance model processing unit 3, and an XYZ value to an L ^* _Cab ^* _hab ^* value. Conversion unit 4, fuzzy image processing unit 5, L ^* C _ab ^* h _ab ^* value to XYZ value conversion unit 6, second color appearance model processing unit 7, XYZ value to rgb value conversion unit 8, rgb value to CMYK value conversion unit 9, input image observation environment parameter storage unit 10,
It includes an observation environment parameter storage unit 11 for a psychological evaluation experiment and an observation environment parameter storage unit 12 for an output image.

In the image processing apparatus according to the present embodiment, the color original image 13 which is the input image is read by the color scanner unit 14, and all pixel information (r (red), g (green) forming the color original image 13 is read. ), B (blue)) as the rgb value and r
Stored in the gb value storage unit 1. The conversion unit 2 for converting the rgb value into the XYZ value is the r stored in the rgb value storage unit 1.
The gb value is converted into an XYZ value of the CIEXYZ color system (here, for convenience, (X _in , Y _in , Z _in )).

Input image observation environment parameter storage unit 10
Is the white point (X _in ^(W) , Y _in
^(W) , Z _in ^(W) ) and its observation environment parameters (brightness of adaptation field, environmental effect constant, etc.) are stored. Further, in the observation environment parameter storage unit 11 of the psychological evaluation experiment, the white point (X _ex ^(W) ,
Y _ex ^{( W)} and Z _ex ^(W) ) and their observation environment parameters (brightness of the adaptation visual field, environmental effect constants, etc.) are stored.

Then, the first color appearance model processing unit 3
Is the XYZ value of the input image output from the rgb value-to-XYZ value conversion unit 2, and the white point and its environment in the observation environment of the input image stored in the observation environment parameter storage unit 10 of the input image. Using the parameters, the white point in the observation environment in which the psychological evaluation experiment is performed and the observation environment parameters stored in the observation environment parameter storage unit 12 of the psychological evaluation experiment, the processing target pixel in the observation environment of the input image is used. It calculates the XYZ value _{(X in, Y in, Z} in) and XYZ values of the pixels which gives the appearance of the same color under observation environment subjected to psychological evaluation experiment _{(X ex, Y ex, Z} ex) a. That is, the color of a pixel viewed under the observation environment of the input image (hereinafter referred to as the first observation environment) and the color of the pixel viewed under the observation environment in which the psychological evaluation experiment is performed (hereinafter referred to as the second observation environment) , Will be perceived as the same color. The first color appearance model processing unit 3 performs processing based on the CIECAM97 _S standard recommended as the latest color appearance model by CIE (International Commission on Illumination).

Next, examples of images observed under the first and second observation environments are shown. 2A to 2C respectively show a B plane image, a G plane image, and an R plane image observed under the first observation environment, and FIG. 2D shows these images observed under the first observation environment. The composite image of is shown. 3 (e) to 3 (g) respectively show a B plane image, a G plane image, and an R plane image observed under the second observation environment, and FIG. 3 (h) shows these observed under the second observation environment. The synthetic image of the image is shown. Figure 2
The images shown in (a) to (d) and the images shown in (e) to (h) of FIG. 3 are perceived as images that give the same color appearance when viewed under respective observation environments. . As described above, the images shown in FIGS. 2A to 2D and FIGS. 3E to 3H give a completely different color appearance when observed under the same observation environment. Nevertheless, it can be seen that when each of them is observed under the first observation environment and the second observation environment, the same color appearance is given. In the first observation environment, the XYZ values of the white point are 95.0, respectively.
5,100.00,108.88, the luminance value of the adaptation visual field is 318.31, the environmental effect constant is 0.69, the color induction factor is 1.0, the brightness contrast factor is 1.0, and the adaptation degree factor is 1. 0.0, and the background relative brightness is 20. Also,
In the second observation environment, the XYZ values of the white point are 10 each.
9.85, 100.00, 35.58, the luminance value of the adaptation visual field is 31.83, the environmental effect constant is 0.69, the color inducing factor is 1.0, the brightness contrast factor is 1.0, and the degree of adaptation is a factor. Is 1.0 and the background relative brightness is 20.

The conversion unit 4 for converting the XYZ values to the L ^* _Cab ^* _hab ^* values is the X output from the first color appearance model processing unit 3.
The YZ values (X _ex , Y _ex , Z _ex ) are converted into L ^* C _ab ^* h _ab ^* values (L _ex , C _ex , h _ex ) which is a cylindrical coordinate system in the L ^* a ^* b ^* color space.

In the fuzzy image processing section 5, different L
^The SD scale is obtained by performing a psychological evaluation experiment for a plurality of skin color patches represented by ^* C _ab ^* h _ab ^* values, and the SD scale evaluation value is the objective variable, and the L ^* C _ab ^* h _ab ^* value of the patch is As an explanatory variable, the quantification theory type I is analyzed, and a range of L ^* _Cab ^* _hab ^* values that gives a favorable impression is extracted. Then, the fuzzy image processing unit 5 generates fuzzy data based on the range of L ^* _Cab ^* _hab ^* values that gives a favorable impression extracted previously, stores the fuzzy data, and stores the fuzzy data by fuzzy inference. Outputs the color adjustment amount.

Here, it is stored in the fuzzy image processing section 5.
The fuzzy data will be described. First, quantification theory
L that gives a favorable impression as a result of analysis of type I^*C
_ab ^*h _ab ^*From the range of values, L^*, C_ab ^*, H_ab ^*each
Set the membership function for the skin color distribution area of values
It For example, h_ab ^*Men representing the preferred skin color for the value
The barship function is set as shown in the graph of FIG.
Similarly, L^*Value, C_ab ^*A value that represents the preferred skin color for the value.
Also set the membership function. Finer color adjustment possible
L for the ability^*, C_ab ^*, H_ab ^*Each skin color region of the value
There are 3 categories, Small, Middle and Large respectively.
And set up a total of 27 membership functions. example
If h_ab ^*Condition part method for the values Small, Middle, Large
Set the membership function as shown in the graph in Figure 5.
It Similarly, L^*Value, C_ab ^*Condition part membership for value
Also set up function.

Next, these 27 membership functions
Based on 27 different fuzzy rules for each L^*, C_ab ^*, H
_ab ^*Create for each value of. For example, h_ab ^*Value for value
The rules are created as shown in the table in FIG. Similarly,
L^*Value, C_ab ^*Also create a fuzzy rule for the value. This
Here, L, which is one of the 27 fuzzy rules
_Small ^*, C_{ab Small} ^*, H_{ab Large} ^*As the first rule
And explain. Target pixel (L_ex, C_ex, H_ex)It
Each fuzzy set L₁= L_Small ^*, C₁=
C_{ab S mall} ^*, H₁= H_{ab Large} ^*Goodness of fit (fa
Membership value) μ_L1(L_ex), Μ_C1(C_ex),
μ_h1(H_ex) Is the graph shown in FIGS.
Ask for sea urchin. For example, to the first rule of the pixel to be processed
Goodness of fit μ₁Is μ_L1(L_{e x}), Μ_C1(C_ex), Μ_h1(H
_ex), It is the smallest of the values shown in FIGS. 7 (a) to 7 (c).
In rough₁= Μ_L1(L_ex). Also, h_ab ^*Against value
From the first rule to do Δh_{ab 1} ^*Can be -7
You see (see Figure 6). Exactly like L^*, C_ab ^*For each value of
ΔL₁ ^*, ΔC_{ab 1} ^*Can also be calculated. that's all
Pixel to be processed according to the first rule by the procedure like
Of μ₁, ΔL₁ ^*, ΔC_{ab 1} ^*, Δh_{ab 1} ^*Can be calculated, and
Similarly, for the j (j = 2, ..., 27) th rule
Μ of processing target pixel_j, ΔL_j ^*, ΔC_{ab j} ^*, Δh_{ab j} ^*Also
It can be calculated.

By the above procedure, the color adjustment amount Δ
L^*, ΔC_ab ^*, Δh_ab ^*Can be calculated and the output value (L_ex′,
C_ex′, H_ex′) ≡ (L_ex+ ΔL^*, C_ex+ ΔC_ab ^*, H
_ex+ Δh _ab ^*) Is obtained. This processing target pixel
(L_ex′, C_ex′, H_ex′) Values are observations of psychological evaluation experiments
L in consideration of preferable color reproduction under environment^*C_ab ^*h_ab ^*Value and
Has become. That is, the fuzzy image processing unit 5
L^*C_ab ^*h_ab ^*Takes a value as input and is stored
Fuzzy inference based on the fuzzy rules
Input L at that pixel^*C_ab ^*h_ab ^*Value and color adjustment amount ΔL^*
ΔC_ab ^*Δh_ab ^*Output the sum of and.

L^*C_ab ^*h_ab ^*Value to XYZ value converter
6 is the (L_ex′, C_ex′, H_ex′)
Value is XYZ value (X_ex′, Y_ex′, Z_ex′)
This is L from XYZ value^*C_ab ^*h_ab ^*The value conversion unit 4
This corresponds to the inverse conversion of the conversion performed. Second color appearance model
The processing unit 7 is the observation environment parameter storage unit 1 of the psychological evaluation experiment.
Observation environment of each parameter and output image stored in 1
Each parameter stored in the parameter storage unit 12
(White point (X_out ^(W)，
Y _out ^(W), Z_out ^(W)) And its observation environment parameters (adaptation)
Brightness of the field of view, environmental effect constants, etc.)
Pixel value to be processed (X_ex′,
Y_ex′, Z_ex′) And the output image have the same color in the viewing environment.
XYZ value (X_out, Y_out, Z_out)
calculate. Processing target in observation environment where psychological evaluation experiment was performed
Pixel value (X_ex′, Y_ex′, Z_ex′) Color and view of output image
XYZ values (X
_out, Y_out, Z_out) Colors are different in each observation environment
Perceived as the same color. This second color appearance model processing unit
Even in 7, CIECAM97_SBased on the standard of
Do the work.

The conversion unit 8 for converting XYZ values to rgb values is r
The inverse conversion of the conversion performed by the conversion unit 2 from the gb value to the XYZ value is performed. The conversion unit 9 from the rgb value to the CMYK value uses the XYZ
R of the output image, which is output from the value-to-rgb-value conversion unit 8 in consideration of preferable color reproduction under the observation environment of the output image.
Convert gb value to CMYK value. Color printer unit 15
Receives the CMYK values of all the pixels of the output image in consideration of preferable color reproduction under the observation environment of the output image, and outputs the color print image 16. Although the image processing focusing on one pixel has been described above, the image processing apparatus according to the present embodiment performs the same processing on other pixels.

Next, the configurations of the first color appearance model processing unit 3 and the second color appearance model processing unit 7, which are features of the image processing apparatus of the present embodiment, will be described in detail. First,
A schematic configuration of the first color appearance model processing unit 3 will be described.
FIG. 8 is a block diagram showing a schematic configuration of the first color appearance model processing unit 3. First color appearance model processing unit 3
Is composed of a color appearance model order conversion unit 31, a JCh value storage unit 32 that stores values of lightness J, saturation C, and hue angle h, and a color appearance model inverse conversion unit 33.

The color appearance model order conversion unit 31 uses the first observation environment parameter 34 to output the XYZ output value from the conversion unit 2 from the rgb value to the XYZ value (hereinafter, the XYZ value 30 in the first observation environment 30). From) CIEC in this pixel
The brightness J, saturation C, and hue angle h under the AM97 _S standard are calculated. The JCh value storage unit 32 stores the lightness J, the saturation C,
The hue angle h is temporarily stored. The color appearance model inverse conversion unit 33 uses the second observation environment parameter 35 to determine the JCh
The lightness J, the saturation C, and the hue angle h stored in the value storage unit 32 are converted into XYZ values 36 in the second observation environment. The XYZ values 36 in the second observation environment obtained here give the same color appearance in each observation environment as the XYZ values 30 in the first observation environment. As the first observation environment parameter 34, each parameter stored in the observation environment parameter storage unit 10 of the input image shown in FIG. 1 is used. As the second observation environment parameter 35, each parameter stored in the observation environment parameter storage unit 11 of the psychological evaluation experiment is used.

Next, the detailed structure of the color appearance model order conversion unit 31 will be described with reference to FIG. The color appearance model order conversion unit 31 includes a Bradford cone response conversion unit 311 and a modified vo.
n Kries transformation unit 312, Bradford cone response to Hunt-Poin
A ter-Estevez cone response conversion unit 313, a first post-adaptation cone response conversion unit 314, and a JCh value calculation unit 315 that calculates the values of lightness J, saturation C, and hue angle h. It

The Bradford cone response conversion unit 311 converts the XYZ values (X ₁ , Y ₁ , Z ₁ ) in the first observation environment into Bradford cone responses (R ₁ , G ₁ , B ₁ ). . The modified von Kries conversion unit 312 converts the Bradford cone response (R ₁ , G ₁ , B ₁ ) into a modified von Kries value (R _1c , G _1c , B _1c ) using the first observation environment parameter 34α. To do. It should be noted that the first observation environment parameter 34α is stored in the observation environment parameter storage unit 10 for the input image shown in FIG. 1, and the luminance value L _A (cd of the adaptation visual field for the image under the observation environment of the input image is L _A (cd /
m ² ) and the factor F of the degree of adaptation are used.

From Bradford cone response to Hunt-Pointer-Estev
The ez cone response conversion unit 313 converts the modified von Kries values (R _1c , G _1c , B _1c ) into the Hunt-Pointer-Estevez cone response (R
₁ ′, G ₁ ′, B ₁ ′). The first adaptation post-cone cone conversion unit 314 uses the first observation environment parameter 34α to generate the Hunt-Pointer-Estevez cone response (R ₁ ′,
G ₁ ′, B ₁ ′) is a post-adaptation cone response (R
_1a ′, G _1a ′, B _1a ′).

The JCh value calculation section 315 uses the first observation environment parameter.
Adaptation under the first observation environment using the parameter 34β
Dorsal cone response (R_1a′, G_1a′, B_1a′) To CIECAM
97_SJ, which represents the lightness J, saturation C, and hue angle h under the standard
The Ch value (J, C, h) is calculated. The first observation environment
The parameter 34β is the observation ring of the input image shown in FIG.
The input image stored in the boundary parameter storage unit 10
Background relative brightness Y under observation environment_b, Environmental effect constant
c, color induction factor N_c, Brightness contrast factor F_LLAnd this
Background inducing factor n, which is a parameter dependent on these, background
And color brightness inducing factor N _bb, N_cb, Exponential nonlinear factor
z is used. The J calculated by the JCh value calculator 315
The Ch value (J, C, h) is stored in the JCh value storage unit shown in FIG.
Stored in 32.

Next, the detailed structure of the color appearance model inverse conversion unit 33 will be described with reference to FIG. The color appearance model inverse conversion unit 33 converts the second adaptation post-cone cone response unit 33.
1, Hunt-Pointer-Estevez cone response conversion unit 332,
Constant-multiply Bradford cone response converter 333, and correction
The von Kries inverse transform unit 334 is used.

Second Adaptation Postcone Cone Response Conversion Unit 331
Using the second observation environment parameter 35β, the JCh value (J, C, h) stored in the JCh value storage unit 32 shown in FIG. R _2a ′, G _2a ′, B _2a ′). As the second observation environment parameter 35β, the background relative luminance Y _b , the environmental effect constant c, and the color induction in the observation environment of the psychological evaluation experiment stored in the observation environment parameter storage unit 11 of the psychological evaluation experiment illustrated in FIG. A factor N _c , a brightness contrast factor F _LL ,
A background inducing factor n, background and color lightness inducing factors N _bb and N _cb , and an exponential nonlinear factor z which are parameters dependent on these are used.

The conversion unit 332 for converting the Hunt-Pointer-Estevez cone response to the Hunt-Pointer-Estevez cone response uses the second observation environment parameter 35α to Hunt-convert the post-adaptation cone response (R _2a ′, G _2a ′, B _2a ′). Poin
Convert to ter-Estevez cone response (R ₂ ′, G ₂ ′, B ₂ ′). As the second observation environment parameter 35α, the luminance value L _A (cd / m of the adaptation visual field for the image under the observation environment of the psychological evaluation experiment, which is stored in the observation environment parameter storage unit 11 of the psychological evaluation experiment of FIG. ² ) and the factor F of the degree of adaptation are used.
The conversion unit 333 for the constant multiple Bradford cone response is the Hunt
-Convert the Pointer-Estevez cone response (R ₂ ′, G ₂ ′, B ₂ ′) into a constant multiple Bradford cone response (R _2c Y ₂ , G _2c Y ₂ , B _2c Y ₂ ). However, Y ₂ is a Y value defined by CIE finally calculated by the color appearance model inverse conversion unit 33, and is a constant for each input pixel. Modified von Kries inverse transformation unit 33
4 is the constant multiple Bradford cone response (R _2c Y ₂ , G _2c Y ₂ , B _2c Y ₂ ) using the second observation environment parameter 35α, and the pixel value (X ₁ , Y ₁ , Z ₁ ) and pixel values (X ₂ , Y ₂ , Z ₂ ) that appear to be the same color under the second observation environment.

On the other hand, the configuration of the second color appearance model processing unit 7 is the same as the configuration of the first color appearance model processing unit 3 shown in FIG. However, as the first observation environment parameter 34 input to the color appearance model order conversion unit 31, each parameter stored in the observation environment parameter storage unit 11 of the psychological evaluation experiment shown in FIG. 1 is used. Further, as the second observation environment parameter 35, each parameter stored in the observation environment parameter storage unit 12 of the output image shown in FIG. 1 is used.

The image processing method using the apparatus of this embodiment will be described in detail below. In the device shown in FIG.
The color original image 13 that is the input image is read by the color scanner unit 14, and all pixel information that forms the color original image 13 is stored in the rgb value storage unit 1 as an rgb value. When the rgb value-to-XYZ value conversion unit 2 receives the rgb value of one pixel from the rgb value storage unit 1, the rgb color information (r ₁ , g ₁ ,
Matrix M is applied to b ₁ ), and C is used in the first observation environment.
Convert to XYZ values (X ₁ , Y ₁ , Z ₁ ) of the IEXYZ color system. However, this matrix M is calculated from the rgb value of the NTSC system by X.
It is a matrix showing conversion to YZ value.

[0043]

[Equation 2]

In the first color appearance model processing unit 3, first, in the Bradford cone response conversion unit 311 which constitutes the color appearance model order conversion unit 31, the XYZ values (X _1, Y _1, Z ₁₎ to act on the matrix M _B respect, converts it into Bradford cone response _{(R 1, G 1, B} 1). However, the matrix M _B is a conversion matrix from the XYZ values to the Bradford cone response. Similarly, the cone response (R _1W , G _1W , B _1W ) for the white point of the light source is also calculated.

[0045]

[Equation 3]

Next, the modified von Kries conversion unit 312 uses the first observation environment parameter 34α to modify the Bradford cone response (R ₁ , G ₁ , B ₁ ) into modified von Kries values (R _1c , G _1c). ， B
_1c ). The first observation environment parameter 34α
Is a luminance value L _1A (cd / m ² ) of the adaptation visual field for the image in the first observation environment and a factor F ₁ of the degree of adaptation.

[0047]

[Equation 4]

Bradford cone response from Hunt-Pointer-Estev
ez conversion unit 313 into cone responses are as follows, the inverse matrix is of the matrix M _B matrix M _B ^-1 and the matrix M _H the modified von
Sequentially act on the Kries value (R _1c , G _1c , B _1c ) and use it as the Hun
Convert to t-Pointer-Estevez cone response (R ₁ ′, G ₁ ′, B ₁ ′).

[0049]

[Equation 5]

First Adapted Postcone Cone Response Conversion Unit 314
Is a first observation environment parameter 34α, using the parameter F _1L obtained from the luminance value L _1A of the adaptation visual field for the image under the first observation environment, the Hunt-Pointer-Estevez cone response (R ₁ ′, G ₁ ′, B ₁ ′) is converted to the post-adaptation cone response (R _1a ′, G _1a ′, B _1a ′) under the first observation environment.

[0051]

[Equation 6]

The JCh value calculation unit 315 uses the first observation environment parameter 34β in the following procedure to adjust the post-adaptation cone response (R _1a ′, G _1a ′, B _1a ′) under the first observation environment. From the above, the lightness J, the saturation C, and the hue angle h under the CIECAM97 _S standard are calculated. The first observation environment parameter 34
β is the background relative luminance Y _1b of the first observation environment, the environmental effect constant c ₁ , the color inducing factor N _1c , and the brightness contrast factor F _1LL ,
A background induction factor n ₁ , which is a parameter dependent on these,
The background and color lightness inducing factors N _1bb and N _{1c b} and the exponential non-linear factor z ₁ are stored, and their relations are as follows. However, Y _1W is the Y value defined by CIE for the white point of the light source in the first observation environment.

[0053]

[Equation 7]

First, the hue angle h is given by the equation (20). a
₁ and b ₁ are a red-green response and a yellow-blue response, respectively.
a ₁ and b ₁ are equations (21) and (2) by using the post-adaptation cone response (R _1a ′, G _1a ′, B _1a ′) under the first observation environment.
2)

[0055]

[Equation 8]

The lightness J is given by equation (23). However, A ₁ is an achromatic response, and A _1W is an achromatic response to the white point of the light source in the first observation environment. Also, the saturation C
Becomes equation (24). Becomes However, s ₁ is the degree of saturation, and e is an eccentricity factor determined by the value of the hue angle h.

[0057]

[Equation 9]

The JCh value calculated in this way is stored in the JCh value storage unit 32 shown in FIG. Next, in the color appearance model inverse conversion unit 33 shown in FIG.
From the Ch value storage unit 32, the JCh value of one pixel (J,
When C, h) is received, the second adaptation post-cone cone response conversion unit 331 shown in FIG. 10 uses the second observation environment parameter 35β to perform the JCh value (J,
C, h) is converted into a post-adaptation cone response (R _2a ′, G _2a ′, B _2a ′) under the second observation environment. The second observation environment parameter 35β includes background relative luminance Y _2b of the second observation environment, environmental effect constant c ₂ , color inducing factor N _2c , lightness contrast factor F _2LL, and background inducing factor n ₂ dependent on them. The background and color lightness inducing factors N _2bb and N _2cb and the exponential non-linear factor z ₂ are stored, and the relationship between them is as follows. However, Y _{2 W} is the Y value defined by CIE for the white point of the light source in the second observation environment.

[0059]

[Equation 10]

First, from the expressions (28) and (29), the lightness J and the saturation C are used to calculate the achromatic color response A ₂ and the saturation s _{2 in} the second observation environment.
To calculate. However, A _2W is an achromatic response to the white point of the light source in the second observation environment. Next, equation (30),
From (31), the red-green color response a ₂ and the yellow-blue color response b ₂ are calculated from the saturation s ₂ and the hue angle h.

[0061]

[Equation 11]

Then, the second adaptation post-cone cone conversion unit 331 uses the achromatic color response A ₂ , the red-green color response a ₂ and the yellow-blue color response b ₂ under the second observation environment. The post-adaptation cone response (R _2a ′, G _2a ′, B _2a ′) of is calculated.

[0063]

[Equation 12]

Next, to the Hunt-Pointer-Estevez cone response
The conversion unit 332 uses the second observation environment parameter 35α (second
Luminance value L of the adaptation field for the image under the observation environment_2A(Cd
/ m²) And the factor F of the degree of adaptation₂) Using the second observation ring
Conformal post-cone cone response (R _2a′, G_2a′, B_2a′) To Hun
t-Pointer-Estevez cone response (R₂′, G₂′, B₂Change to ′)
Replace.

[0065]

[Equation 13]

Conversion unit 333 to Bradford cone response of constant multiple
Is the Hunt-Pointer-Estevez cone response (R ₂ ′, G ₂ ′,
B ₂ ') a constant multiple Bradford cone response _{(R 2c Y 2, G 2c} Y 2, B 2c Y
₂ ) Convert to.

[0067]

[Equation 14]

However, Y ₂ is a color appearance model inverse conversion unit 33.
The Y value finally specified in CIE is defined by CIE, and is a constant for each input pixel. This equation is an inverse conversion of the conversion performed by the conversion unit 313 from the Bradford cone response shown in FIG. 9 to the Hunt-Pointer-Estevez cone response.

The modified von Kries inverse transformation unit 334 uses the second observation environment parameter 35α according to the following procedure.
Bradford cone response (R _2c Y ₂ , G _2c Y ₂ , B _2c Y ₂ )
Is converted into a pixel value (X ₁ , Y ₁ , Z ₁ ) in the _first observation environment and a pixel value (X ₂ , Y ₂ , Z ₂ ) that looks the same in the second observation environment. First, Y _2C is calculated from the equation (40).

[0070]

[Equation 15]

This Y _2C can be expressed by the following equations (42) by using the Bradford cone response (R _2c Y ₂ , G _2c Y ₂ , B _2c Y ₂ ) multiplied by the constant.
It can be expressed as (44). However, (R2W, G2W, B2W) is the second
Is the cone response to the white point of the light source under the observation environment,
Equations (45) and (46) are given.

[0072]

[Equation 16]

Further, since it is the expression (47), the pixel value (X ₂ , Y ₂ , Z ₂ ) that looks like the same color under the first observation environment and the second observation environment is calculated by the expression ( 48) and output as the XYZ value 36 in the second observation environment. This X
The YZ value 36 is a pixel value in which the color of a pixel under the observation environment of the input image and the color of the same color under the observation environment of the psychological evaluation experiment.

[0074]

[Equation 17]

Next, from the XYZ values shown in FIG. 1, L ^* C _ab ^*
The conversion unit 4 for converting into the _ab ^* value converts the XYZ values output from the first color appearance model processing unit 3 into the expressions (49) to (51), and further converts the XYZ values into the L ^* a ^* b ^* color space. C _ab ^* , h _ab ^* values which are the cylindrical coordinate system of are calculated. However, (X _W , Y _W ,
Z _W ) is the XYZ value for the white point of the light source in the observation environment of the psychological evaluation experiment.

[0076]

[Equation 18]

The fuzzy image processing unit 5 is as described above.
L that gives a favorable impression to^*C_ab ^*h _ab ^*Change the range of values
It is stored as fuzzy data and fuzzy reasoning is used.
To output the color adjustment amount. Fuzzy inference uses L for each pixel.
^*C_ab ^*h_ab ^*Take the value as input and follow the steps above
Based on the fuzzy rules created and stored by
Input L in pixel^*C_ab ^*h_ab ^*Value and color adjustment amount ΔL^*ΔC
_ab ^*Δh_ab ^*Output the sum of and. Color adjustment amount ΔL^*, Δ
C_ab ^*, Δh_ab ^*Is calculated by equations (1) to (4).
However, μ_Li(L^*), Μ_Ci(C_ab ^*), Μ_hi(H_ab ^*)
Is L for each i-th (i is a natural number) rule^*，
C_ab ^*, H_ab ^*Is the goodness of fit of ΔL_i ^*, ΔC_{ab i} ^*, Δ
h_{ab i} ^*Is L according to the i-th rule^*, C_ab ^*, H
_ab ^*It is an output value.

[0078]

[Formula 19]

The L ^* C _ab ^* h _ab ^* value conversion unit 6 converts the L ^* C _ab ^* h _ab ^* value, which is the output from the fuzzy image processing unit 5, into the psychological evaluation experiment. The image is converted into XYZ in consideration of preferable color reproduction under the observation environment. This corresponds to the inverse conversion of the conversion performed by the conversion unit 4 from the XYZ value to the L ^* _Cab ^* _hab ^* value.

[0080]

[Equation 20]

The second color appearance model processing unit 7
Stored in the observation environment parameter storage unit 11 of the price experiment
Parameter storage area for each parameter and output image
L using each parameter stored in 12^*C_ab ^*
h_ab ^*Output from the value to XYZ value conversion unit 6,
Considering favorable color reproduction under the observation environment of psychological evaluation experiments
Psychological evaluation experiment by receiving XYZ values of images as input values
Image and output image considering preferable color reproduction under other viewing environment
XYZ values of images that give the same color appearance under the image observation environment
Is output. The second color appearance model processing unit 7 is shown in FIG.
In the first color appearance model processing unit 3 shown, the first view
For the observation environment parameter and the second observation environment parameter
Observing environment parameters of the psychological evaluation experiment and observing the output image
In the first observation environment shown in FIG.
XYZ value of 30 is L^*C_ab ^*h_ab ^*Value to XYZ value
The XYZ output values from the conversion unit 6 correspond. Note this
As will be understood, the operation of the second color appearance model processing unit 7 is
The appearance of the color 1 is exactly the same as the operation of the model processing unit 3.
Therefore, the description is omitted.

The conversion unit 8 for converting XYZ values to rgb values uses r
The inverse conversion of the conversion performed by the conversion unit 2 from the gb value to the XYZ value is performed. Here, the XYZ values of the image in consideration of preferable color reproduction under the observation environment of the output image are converted into rgb values.

[0083]

[Equation 21]

The conversion unit 9 from the rgb value to the CMYK value is
The rgb value of the output image output from the conversion unit 8 for converting the XYZ value to the rgb value in consideration of the preferable color reproduction under the observation environment of the output image is converted into the CMYK value. Then, the color printer unit 15 receives the CMYK values of all the pixels of the output image in consideration of preferable color reproduction under the observation environment of the output image, and outputs the color print image 16.

With the image processing apparatus of this embodiment, even if the observation environment for the input image and the observation environment in which the psychological evaluation experiment is performed are different, a good image color can be obtained by the above procedure. It can be reproduced. Further, since the latest color appearance model is taken into consideration, it is possible to perform not only the preferable color reproduction but also the image processing in which the corresponding color reproduction is also taken into consideration. That is, by sufficiently considering the observation environment, it is possible to further improve the accuracy of the preferable color reproduction of the conventional technique, and at the same time, perform the color reproduction in consideration of the preferable color reproduction and the corresponding color reproduction.

<Second Embodiment> FIG. 11 is a block diagram showing the schematic arrangement of an image processing apparatus according to the present embodiment. The image processing apparatus according to the present embodiment includes an rgb value storage unit 1, a conversion unit 2 for converting rgb values to XYZ values, a color appearance model order conversion unit 41, a first JCh value storage unit 42, and a fuzzy image processing unit 43. , A second JCh value storage unit 44, a color appearance model inverse conversion unit 45, an XYZ value to rgb value conversion unit 8, an rgb value to CMYK value conversion unit 9, and an input image observation environment parameter storage unit 10. , A psychological evaluation experiment observation environment parameter storage unit 11 and an output image observation environment parameter storage unit 12. rgb value storage unit 1, rgb value to XYZ value conversion unit 2, XYZ value to rgb value conversion unit 8, rgb value to CMYK value conversion unit 9, input image observation environment parameter storage unit 10, psychology As the observation environment parameter storage unit 11 of the evaluation experiment and the observation environment parameter storage unit 12 of the output image, the same devices as those of the device shown in the first embodiment are used. Hereinafter, description of the same components as those of the device according to the first embodiment will be omitted.

First, the detailed structure of the color appearance model order conversion unit 41 will be described with reference to FIG. The color appearance model order conversion unit 41 corrects the Bradford cone response conversion unit 411.
von Kries transform 412, Bradford cone response to Hunt-Poi
An nter-Estevez cone response conversion unit 413, a first post-adaptation cone response conversion unit 414, and a JCh value calculation unit 415 that calculates the values of lightness J, saturation C, and hue angle h. It

The Bradford cone response conversion unit 411 uses r
XYZ output from the conversion unit 2 from gb value to XYZ value
Convert the values (X ₁ , Y ₁ , Z ₁ ) to the Bradford cone response (R ₁ , G ₁ , B ₁ ). The modified von Kries transformation unit 412 uses the observation environment parameter 416α of the input image to modify the Bradford cone response (R ₁ , G ₁ , B ₁ ) into a modified von Kries value (R _1c , G _1c , B).
_1c ). It should be noted that the observation environment parameter 416α of the input image is stored in the observation environment parameter storage unit 10 of the input image shown in FIG. 11, and the luminance value L _A (cd of the adaptive visual field for the image under the observation environment of the input image is L _A (cd / m ² ) and a factor F of the degree of adaptation.

Bradford cone response from Hunt-Pointer-Estev
The ez cone response conversion unit 413 converts the modified von Kries values (R _1c , G _1c , B _1c ) into the Hunt-Pointer-Estevez cone response (R
₁ ′, G ₁ ′, B ₁ ′). The first adaptation post-cone cone conversion unit 414 uses the observation environment parameter 4 of the input image.
16α is used to calculate the Hunt-Pointer-Estevez cone response (R ₁ ′, G ₁ ′, B ₁ ′) after adaptation of the input image under the observation environment (R _1a ′, G _1a ′, B _1a ′).

The JCh value calculation unit 415 uses the observation environment parameter 416β of the input image to calculate the CIECAM97 from the post-adaptation cone response (R _1a ′, G _1a ′, B _1a ′) under the observation environment of the input image. A JCh value (J, C, h) representing the lightness J, the saturation C, and the hue angle h under the _S standard is calculated. Note that the observation environment parameter 416β of the input image includes
1. The background relative luminance Y _b , the environmental effect constant c, the color induction factor N _c , and the brightness contrast factor F _LL stored in the input image observation environment parameter storage unit 10 shown in FIG. The background inducing factor n, which is a parameter dependent on these, the lightness inducing factor N _bb of the background and color,
N _cb , an exponential nonlinear factor z is used. The JCh value (J, C, h) calculated by the JCh value calculation unit 415 is
It is stored in the first JCh value storage unit 42 shown in FIG.

Next, in the fuzzy image processing section 43, a psychological evaluation experiment is performed for a plurality of skin color patches represented by different JCh values in advance, an SD scale is obtained, and the evaluation value of this SD scale is used as the objective variable and JCh of the patch. JCh that gives a favorable impression by analyzing with quantification theory type I with the value as an explanatory variable
Extract a range of values. Here, unlike the embodiment 1 of the prior art and embodiments, analyzing the psychological evaluation experiment results with respect to JCh values under standard _{^{L * C ab * h ab *}} CIECAM97 S of the observation environment due consideration than the value, preferably Extract the range of JCh values that give an impression. In order to calculate the JCh value of the skin color patch used in the psychological evaluation experiment, the observation environment parameter of the psychological evaluation experiment is necessary.

This parameter is the luminance value L _A (cd / m ² ) of the adaptation visual field for the image under the observation environment of the psychological evaluation experiment,
Adaptability factor F, background relative brightness Y _b , environmental effect constant c, color inducing factor N _c , lightness contrast factor _FLL, and background inducing factor n, which is a parameter dependent on these, background and color lightness inducing factor N _bb , N _cb and exponential non-linear factors, which are stored in the observation environment parameter storage unit 11 of the psychological evaluation experiment shown in FIG. 11. Then, the fuzzy image processing unit 43 generates fuzzy data based on the previously extracted JCh value range that gives a desirable impression, stores the fuzzy data, and outputs the color adjustment amount by fuzzy inference.

The fuzzy data stored in the fuzzy image processing section 43 will be described below. First, JC gives a favorable impression as an analysis result of quantification theory type I
From the range of h values, the membership functions for the skin color distribution regions of the respective values of J, C, and h are set. For example, h
A membership function representing a preferable skin color with respect to the value is set as shown in the graph of FIG. Similarly, J value, C
We also set up a membership function that represents the preferred skin color for the value. To enable finer color adjustment, JC
Each skin color area of h value is 3 of Small, Middle, Large
Classify into one category and set the membership function. For example, the conditional part membership functions for Small, Middle and Large of the h value are set as shown in the graph of FIG. Similarly, the conditional part membership function for the J value and the C value is also set. In addition, a fuzzy rule is created in consideration of the characteristics of each skin color area. FIG. 15 is a table showing fuzzy rules for h values created in the same manner as the fuzzy rules for h _ab ^* values in the first embodiment shown in FIG. Fuzzy rules for J value and C value are created in the same manner.

In the fuzzy image processing section 43, J of each pixel is
The Ch value is input, fuzzy inference is executed based on the previously stored fuzzy rule, and the sum of the JCh value and the color adjustment amount ΔJΔCΔh at that pixel is output. This output value is stored in the second JCh value storage unit 44.

Next, the detailed structure of the color appearance model inverse conversion section 45 will be described with reference to FIG. The color appearance model inverse transforming unit 45 transforms the second adapted post-cone cone response 45.
1. Hunt-Pointer-Estevez Cone response conversion unit 452,
Constant multiple Bradford cone response converter 453, and correction
The von Kries inverse transformation unit 454 is configured.

Second Adaptation Posterior Cone Response Conversion Unit 451
Using the observation environment parameter 455β of the output image,
The JCh value (J, C, h) output from the second JCh value storage unit 44 shown in FIG. 11 is used to adjust the cone response (R _2a ′, G _2a ′, B) of the output image under the observation environment. _2a ′).
The output image viewing environment parameter 455β includes background relative luminance Y _b , environment effect constant c, and color inducing factor stored in the output image viewing environment parameter storage unit 12 illustrated in FIG. 11 in the output image viewing environment. N _c , the brightness contrast factor F _LL, and the background inducing factor n, which are parameters dependent on these, the background and color lightness inducing factors N _bb and N _cb , and the exponential nonlinear factor z are used.

The Hunt-Pointer-Estevez cone response conversion unit 452 uses the observation environment parameter 455α of the output image to convert the post-adaptation cone response (R _2a ′, G _2a ′, B _2a ′) to H.
unt-Pointer-Estevez Cone response (R ₂ ′, G ₂ ′, B ₂ ′). The output image viewing environment parameter 455α is stored in the output image viewing environment parameter storage unit 12 shown in FIG. 11, and the luminance value L _A (cd / cd of the adaptation visual field for the image under the viewing environment of the output image is stored. m ² ) and the factor F of the degree of adaptation are used.

Conversion unit 453 to constant-multiplied Bradford cone response
Is the Hunt-Pointer-Estevez cone response (R ₂ ′, G ₂ ′,
B ₂ ') a constant multiple Bradford cone response _{(R 2c Y 2, G 2c} Y 2, B 2c Y
₂ ) Convert to. However, Y ₂ is a Y value defined by CIE finally calculated by the color appearance model inverse conversion unit 45, and is a constant for each input pixel. Modified von Kries
The inverse conversion unit 454 uses the observation environment parameter 45 of the output image.
5α, the Bradford cone response (R _2c Y ₂ , G
_2c Y ₂ , B _2c Y ₂ ) as the pixel value (X
_It is converted to pixel values (X ₂ , Y ₂ , Z ₂ ) that appear to be the same color under the observation environment of the output image and the color of ₁ , Y ₁ , Z ₁ ).

The image processing method using the apparatus of this embodiment will be described below. In the apparatus shown in FIG. 11, from the step of reading the color original image 13 which is the input image by the color scanner unit 14, the conversion unit 2 from the rgb value to the XYZ value converts the XYZ values of the CIEXYZ color system in the observation environment of the input image ( The process of converting into X ₁ , Y ₁ , Z ₁ ) is the same as in the case of the first embodiment.

The color appearance model order conversion unit 41 uses the observation environment parameters of the input image to convert the XYZ values in the observation environment of the input image output from the conversion unit 2 from the rgb value to the XYZ value into the brightness J, JCh representing saturation C and hue angle h
Convert to a value (see FIG. 12). The operation of the color appearance model order conversion unit 41 is almost the same as the operation of the color appearance model order conversion unit 31 of the device of the first embodiment shown in FIG. The difference is that instead of the first observation environment parameters 34α and 34β of FIG. 9, the observation environment parameters 416α and 416 of the input image are used.
This is the point where β is used. The JCh value output from the color appearance model order conversion unit 41 is stored in the first JCh value storage unit 42.

As described above, the fuzzy image processing unit 43 stores the range of JCh values that give a favorable impression as fuzzy data, and outputs the color adjustment amount by fuzzy inference. Figure 1 shows the conditional part membership function for h values, the conditional part membership function for small, middle and large h values, and the fuzzy rule for h values.
3, 14, and 15 respectively. Also,
Data for J value and C value are similarly set and stored. The fuzzy inference in the fuzzy image processing unit 43 is
The JCh value output from the first JCh value storage unit 42 is
It is performed by calculating the sum of the JCh value and the color adjustment amount ΔJΔCΔh in the pixel based on the stored fuzzy rule. Fuzzy image processing unit 43
The output value from is stored in the second JCh value storage unit 44.

Here, the color adjustment amounts ΔJ, ΔC, Δh are calculated by the equations (58) to (61). However, μ
_Ji (J), μ _Ci (C), and μ _hi (h) are the fitness of JCh to the i-th (i is a natural number) rule, and Δ
J _i , ΔC _i , and Δh _i are JC according to the i-th rule, respectively.
h output value.

[0103]

[Equation 22]

The color appearance model inverse conversion unit 45 uses the observation environment parameter of the output image to determine the J value of the image output from the second JCh value storage unit 44 in consideration of the preferable color reproduction.
The Ch value is converted into an XYZ value that gives the same color appearance under the observation environment of the image and the output image in consideration of preferable color reproduction (see FIG. 16). The operation of the color appearance model inverse conversion unit 45 is substantially the same as the operation of the color appearance model inverse conversion unit 33 in the device of the first embodiment shown in FIG. The difference is that instead of the second observation environment parameters 35α and 35β of FIG. 10, the observation environment parameters 455α and 455 of the output image are used.
This is the point where β is used. The subsequent steps are the same as those in the first embodiment.
The description is omitted because it is the same as.

As described above, in the device of this embodiment,
Psychological evaluation experiments were performed using CIECAM97 _S with lightness J, saturation C,
By performing on the skin color patch represented by the hue angle h,
Perform fuzzy image processing on JCh values. Therefore, the device is further simplified as compared with the device according to the first embodiment. In addition, the processing time is shortened.

The embodiment of the present invention has been described above with reference to the drawings. However, it goes without saying that the present invention is not limited to the items shown in this embodiment, and changes and improvements thereof can be made based on the description of the claims.
For example, the present invention is intended for color copying, but is not limited to copying as long as it is an apparatus that performs image processing. In the present invention, the CIECA currently recommended by CIE as the latest color appearance model.
As described above using the M97 _S , the fuzzy image processing unit executes the processing for the relative attributes of lightness, saturation, and hue angle, so that CIE can be performed with fewer viewing environment parameters.
It is also possible to use a ZLAB color appearance model that outputs almost the same results as CAM97 _S. Future CIECAM
When a color appearance model with a higher precision than 97 _S is proposed, it is possible to apply the model to the color appearance model processing unit of the present invention.

[0107]

As described above, according to the present invention, the observation environment for the input image, the observation environment for the psychological evaluation experiment,
Even if the observation environment for the output image is different,
An image processing method and device capable of highly accurate color reproduction can be provided.

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration of an image processing apparatus according to a first embodiment.

FIGS. 2A to 2D are diagrams showing images observed in a first observation environment.

3 (e) to 3 (h) are diagrams showing images observed under a second observation environment.

FIG. 4 is a graph showing a membership function representing a preferable skin color with respect to h _ab ^* value.

FIG. 5 is a graph showing a membership function of a conditional part for Small, Middle, and Large of h _ab ^* values.

FIG. 6 shows a table of fuzzy rules for h _ab ^* values.

FIG. 7A is a graph showing a method for obtaining a goodness of fit μ _L1 (L _ex ), (b) is a graph showing a method for obtaining a goodness of fit μ _C1 (C _ex ), and (c) is a goodness of fit μ _h1. It is a graph which shows the method of calculating _| requiring ( _hex ).

FIG. 8 is a block diagram showing a schematic configuration of a first color appearance model processing unit 3.

9 is a block diagram showing a configuration of a color appearance model order conversion unit 31. FIG.

FIG. 10 is a block diagram showing the configuration of a color appearance model inverse conversion unit 33.

FIG. 11 is a block diagram showing a schematic configuration of an image processing apparatus according to a second embodiment.

FIG. 12 is a block diagram showing a configuration of a color appearance model order conversion unit 41.

FIG. 13 is a graph showing a membership function indicating a preferable skin color with respect to an h value.

FIG. 14 is a graph showing a membership function of a conditional part for Small, Middle, and Large of h value.

FIG. 15 is a diagram showing a table of fuzzy rules for h values.

16 is a block diagram showing the configuration of a color appearance model inverse conversion unit 45. FIG.

FIG. 17 is a diagram illustrating an execution procedure of a conventional color reproduction method.

[Explanation of symbols]

1 rgb value storage section 2 rgb value to XYZ value conversion section 3 First color appearance model processing section 4 XYZ value to L ^* C _ab ^* h _ab ^* value conversion section 5, 43 Fuzzy image processing section 6 L ^* C _ab ^* h _ab ^* value to XYZ value converter 7 Second color appearance model processor 8 XYZ to rgb value converter 9 rgb to CMYK value converter 10 Observation environment parameter storage 11 Observation environment parameter storage 12 for psychological evaluation experiment Observation environment parameter storage 13 for output image Color original image 14 Color scanner 15 Color printer 16 Color print image 30 XYZ values in the first observation environment 31 , 41 Color appearance model forward conversion unit 32 JCh value storage unit 33, 45 Color appearance model inverse conversion unit 34, 34α, 34β First observation environment parameter 35, 35α, 35β Second observation Environment parameter 36 XYZ value in second observation environment 42 First JCh value storage unit 44 Second JCh value storage unit 311, 411 Bradford Cone response conversion unit 312, 412 Modified von Kries conversion unit 313, 413 Bradford Cone response from Hunt-Pointer-E
stevez Cone response conversion units 314, 414 First adaptation after cone response conversion units 315, 415 JCh value calculation units 331, 451 Second adaptation after cone response conversion units 332, 452 Hunt- Pointer-Estevez Cone response conversion unit 333, 453 Constant multiple Bradford cone response conversion unit 334, 454 Modified von Kries inverse conversion unit 416α, 416β Input image observation environment parameter 455α, 455β Output image observation environment parameter

   ─────────────────────────────────────────────────── ─── Continued front page    F-term (reference) 5B057 AA11 BA02 CA01 CA08 CA12                       CA16 CB01 CB08 CB12 CB16                       CC01 CE17 CE18 CH08                 5C077 LL19 MP08 PP32 PP35 PP36                       PP37 PQ12 TT02                 5C079 HB01 HB03 HB05 HB06 HB11                       HB12 LB01 MA11 NA03 PA03

Claims (8)

[Claims] 1. An rgb of three primary colors of red, green and blue from an original image.
A step of obtaining a signal value and the rgb signal value by CIEX
Converting the XYZ values into the XYZ values of the YZ color system, and converting the XYZ values into L ^* _Cab ^* _hab ^* values, which is a cylindrical coordinate system of the L ^* a ^* b ^* color space recommended by the CIE as a uniform perceptual color space. a step of converting, a step of performing a fuzzy processing with respect to the L ^* C _ab ^* h _ab ^* value, the L ^* C _ab ^* h _ab ^* value after fuzzy processing CIEX
An image processing method comprising: a step of converting into an XYZ value of a YZ color system; a step of converting the XYZ value into an rgb value; and a step of reproducing an image using the rgb value, wherein the rgb value The first color appearance model processing step of giving a predetermined parameter to the XYZ values converted from the above, and the X converted from the L ^* _Cab ^* _hab ^* values after the fuzzy processing.
And a second color appearance model processing step of giving a predetermined parameter to the YZ value, the image processing method. 2. The parameters provided in the first color appearance model processing step are parameters that suggest an observation environment of an input image and parameters that suggest an observation environment in which a psychological evaluation experiment is performed. 2. The image according to claim 1, wherein the parameters given in the color appearance model processing step are a parameter indicating an observation environment in which a psychological evaluation experiment is performed and a parameter indicating an observation environment of the output image. Processing method. 3. rgb of three primary colors of red, green and blue from the original image
A step of obtaining a signal value and the rgb signal value by CIEX
A step of converting into an XYZ value of the YZ color system, a step of converting a color appearance model into a step of giving a predetermined parameter to the XYZ value to calculate values of lightness, saturation and hue angle of the CIECAM97 _S model; A step of performing a fuzzy process by giving predetermined parameters to the values of lightness, saturation and hue angle, and giving a predetermined parameter to the values of the lightness, saturation and hue angle which have been subjected to the fuzzy process, by the CIEXYZ color system Color appearance model inverse conversion step of converting the XYZ values to rgb.
An image processing method comprising: a step of converting the value into a value; and a step of reproducing an image using the rgb value. 4. The parameter provided in the color appearance model order conversion step is a parameter indicating the observation environment of the input image, and the parameter provided in the fuzzy processing step is a psychological evaluation experiment. The image processing method according to claim 3, wherein the observation environment is a parameter indicating the observation environment, and the parameter added in the color appearance model inverse conversion step is a parameter indicating the observation environment of the output image. . 5. The rgb of three primary colors of red, green and blue from the original image.
A means for obtaining a signal value and the rgb signal value by CIEX
Means for converting into XYZ values of the YZ color system, and the XYZ values into L ^* _Cab ^* _hab ^* values, which are cylindrical coordinate systems of the L ^* a ^* b ^* color space recommended by the CIE as a uniform perceptual color space. means for converting, and means for performing a fuzzy processing with respect to the L ^* C _ab ^* h _ab ^* value, the L ^* C _ab ^* h _ab ^* value after fuzzy processing CIEX
In an image processing apparatus configured to include means for converting into an XYZ value of a YZ color system, means for converting the XYZ value into an rgb value, and means for reproducing an image using the rgb value, the rgb First color appearance model processing means for giving a predetermined parameter to the XYZ values converted from the values, and the X converted from the L ^* _Cab ^* _hab ^* values after the fuzzy processing.
An image processing apparatus, comprising: a second color appearance model processing means for giving a predetermined parameter to the YZ value. 6. The parameters provided by the first color appearance model processing means are a parameter indicating an observation environment of an input image and a parameter indicating an observation environment in which a psychological evaluation experiment is performed, 6. The image according to claim 5, wherein the parameters given by the color appearance model processing means are a parameter indicating an observation environment in which a psychological evaluation experiment is performed and a parameter indicating an observation environment of the output image. Processing equipment. 7. The rgb of three primary colors of red, green and blue from the original image.
A means for obtaining a signal value and the rgb signal value by CIEX
A means for converting into XYZ values of the YZ color system, a color appearance model order converting means for giving a predetermined parameter to the XYZ values to calculate the values of lightness, saturation and hue angle of the CIECAM97 _S model, A means for performing fuzzy processing by giving predetermined parameters to the values of lightness, saturation and hue angle, and a CIEXYZ color system by giving predetermined parameters to the values of lightness, saturation and hue angle subjected to the fuzzy processing. Color appearance model inverse conversion means for converting the XYZ values to rgb.
An image processing apparatus comprising: a unit for converting into a value and a unit for reproducing an image using the rgb value. 8. The parameter provided by the color appearance model order conversion means is a parameter indicating the observation environment of the input image, and the parameter provided by the fuzzy processing means is a psychological evaluation experiment. The image processing apparatus according to claim 7, wherein the parameter that indicates the viewing environment is a parameter that indicates the viewing environment of the output image. .