JP2001358960A - Color conversion coefficient introducing method, color conversion method, recording medium and color converter - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a conversion coefficient introducing method to stably conduct color conversion at all times without deteriorating the color reproducibility by taking information or the like of a spectral sensitivity distribution and a lighting light source of an input device such as a digital still camera or a scanner and information or the like of a spectral sensitivity distribution and a spectral intensity distribution of an illumination light source of an output device into account and to provide a color conversion method, a recording medium and a color converter. SOLUTION: In the case of color conversion of a 1st image signal characterized in a 1st spectral sensitivity distribution into a 2nd image signal characterized in a 2nd spectral sensitivity distribution, a conversion coefficient is introduced from the 1st spectral sensitivity distribution and the 2nd spectral sensitivity distribution or the conversion coefficient is introduced from the 1st spectral sensitivity distribution, the spectral intensity distribution furthermore characterizing the 1st image signal, the 2nd spectral sensitivity distribution, and the spectral intensity distribution of the 2nd lighting light furthermore characterizing the 2nd image signal in order to solve the tasks above.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of deriving a color conversion coefficient for converting a color image obtained by a scanner or a digital still camera into an image signal suitable for an image display monitor or the like, and in particular, to a method of deriving a color using a computer. Conversion coefficient derivation method,
The present invention belongs to the technical field of a color conversion method, a recording medium storing a program for executing the color conversion method by a computer, and a color conversion apparatus using the same.

[0002]

2. Description of the Related Art Today, personal computers, digital still cameras, scanners, and the like have been improved in performance and have become widespread, so that digital images captured by a digital still camera or scanner can be captured by a personal computer and displayed on a monitor. Or on a home television monitor or the like. In addition, there are many opportunities to display the same digital image on a display monitor, which is a disparate display medium. For example, a digital image captured by a scanner or a digital image captured by a digital still camera can be converted to a high-definition television (HDTV). There are many occasions of displaying on a television (NTSC) display monitor or an NTSC display monitor of the NTSC format, etc., and the mutual entry of the same image signal to different types of image display media is progressing.

[0003] Accordingly, it is expected that it will be difficult to specify or limit an image display medium for displaying a digital image in the future. For example, in a digital still camera currently on the market, an HDTV display monitor is preset as an output image display medium, but in the future, an NTSC display monitor or an sRGB color signal format sRG will be used.
It is expected that the number of cases of displaying an image on the B display monitor and viewing the image will increase. In the field of desktop publishing, a method of creating a print manuscript from a digital image captured by a scanner or the like via a lithographic film is currently generally used. Attempts have also begun to create halftone dots on printing plates directly with laser light using output devices, and it is expected that CTP output devices will be used as one of digital image output devices in the future, and digital image data A wider output destination can be selected.

By the way, such an image display medium or CT
When an image signal is input to an output device such as a P output device,
The image signal is characterized by a predetermined spectral intensity distribution and spectral sensitivity distribution to perform image display and printing, etc., and the spectral intensity distribution and spectral sensitivity distribution of this output device are obtained by an input device such as a digital still camera or a scanner. The spectral intensity distribution and the spectral sensitivity distribution of the illumination light at the time of acquiring the image do not always coincide with each other and often differ. Therefore, an image signal obtained by an input device such as a digital still camera is
It is necessary to perform color conversion on an image signal suitable for the output device.
Further, the spectral intensity distribution and the spectral sensitivity distribution of the illumination light at the time of photographing in the input device are also variously different, and even if the image signals of the same subject are image signals characterized by the spectral intensity distribution and the spectral sensitivity distribution, , Varies depending on the input device.

FIG. 4A shows a spectral sensitivity distribution of a digital still camera for obtaining an image signal, and FIG. 4B shows a spectral sensitivity distribution of a scanner (StatusA) for obtaining an image signal. It can be seen that the sensitivity distribution differs greatly depending on the type of input device such as a digital still camera or a scanner (StatusA). FIG. 4C shows the spectral sensitivity distribution of the HDTV display monitor, and FIG. 4C shows the spectral sensitivity distribution of the NTSC display monitor. The spectral sensitivity distribution greatly differs depending on the type of output device. I understand.

Considering that image signals will be more and more likely to enter an output device such as a heterogeneous image display medium in the future,
It is important to stably perform color conversion suitable for an output device.

[0007]

By the way, a subject is actually photographed with a digital still camera, and this photographed image is
The flow of an image signal until an image is displayed as a monitor image on a DTV display monitor will be specifically described. As shown in FIG. 5, an image was captured using a CCD image sensor having the spectral sensitivity shown in FIG. The device image signal of the subject is corrected in white balance to become a white balance corrected device image signal, and the signal is subjected to color conversion to obtain a white balance corrected monitor image signal. That is, a monitor image is obtained through the flow of [1] and [2] in FIG.

Although the flow of such an image signal is general, an intermediate color signal of a physiological primary color (LMS) having a spectral sensitivity distribution as shown in FIG. Then, the device image signal is color-converted into this intermediate signal, and the white balance corrected monitor image signal is finally obtained from this intermediate signal, that is, [4] → [5] →
[7] or [4] → [6] → [8] is assumed. Also,
It is converted into a white balance-processed intermediate signal according to the flow [3] in FIG. 5, and then converted into a white balance corrected monitor image signal according to the flow [7].
A flow of [3] → [7] is also assumed. In any case,
[2], [3], [4], [6] and [7] in FIG.
Needs to perform color conversion.

By the way, as shown in the flow of [2] in FIG. 5, it is obtained from an input device such as a digital still camera.
A method of color-converting a digital image signal (white balance corrected device image signal) subjected to white balance correction into an image signal (white balance corrected monitor image signal) suitable for an output device such as an image display monitor is generally used. Using a chart of a predetermined color sample, for example, a Macbeth chart composed of 24 color patches as an image of a subject, a conversion coefficient for color conversion is obtained from this image signal, and this conversion coefficient is used as a coefficient of a color conversion matrix. Is used to perform color conversion. That is, in the conventional method, as shown in FIG. 6, under a predetermined illumination light source, the value of the input image signal of the chart obtained in the input device is equal to the value of the output image signal defined as the image signal of the chart of the output device. A conversion coefficient that best approximates the value, that is, minimizes the mean square error between the input image signal and the output image signal, is obtained, and color conversion is performed using this coefficient as a conversion coefficient of a color conversion matrix. For example, when the input image signal is an image signal composed of R, G, and B, three rows and three columns are arranged so that the output image of the subject chart set in the output device is color-converted into R, G, and B image signals. A conversion coefficient of a color conversion matrix is obtained, and color conversion is performed on various image signals using the conversion coefficient.

Here, the input image signal is obtained by photographing a chart as an object in the input device using an illumination light source which emits illumination light having a predetermined spectral intensity distribution and a spectral sensitivity distribution of the input device. An image signal, that is, an image signal characterized by a spectral intensity distribution and a spectral sensitivity distribution of the illumination light of the input device, and the output image signal is displayed as an image on an output device having a predetermined spectral sensitivity distribution. An image signal predetermined so that the colorimetric value of the chart of the subject matches the colorimetric value of the chart of the actual subject under the illumination light source determined by the output device, that is, the spectral intensity distribution and spectral distribution of the illumination light of the output device. It is an image signal characterized by a sensitivity distribution.

However, the conversion coefficient of the color conversion matrix obtained by the above method greatly depends on the subject and the illumination light source, and there is a problem that color reproduction is not stable. For example, a chart used as a subject is made of a combination of specific pigments. Therefore, even if the chromaticity values are distributed over a wide range, the spectral waveforms are only combinations of the spectral waveforms of the pigments. The waveform has only a very limited spectral waveform. Therefore, as the difference between the distribution of the spectral waveform between the actual object and the object such as the chart used to obtain the conversion coefficient of the color conversion becomes larger, the color There has been a problem that color reproducibility when converting and displaying an image is greatly deteriorated.

[0012] The illumination light source may be used at the time of strobe light emission or in studio photography such as a television or a photo studio.
The illumination light source can be specified. For example, in the case of studio photography, it may be possible to specify the tungsten light source.However, in general, it is often impossible to specify the illumination light source, and the illumination used when calculating the conversion coefficient of color conversion using the actual illumination light source and the chart. As the difference from the light source increases, there is a problem that the color reproducibility when displaying an image by performing color conversion is greatly deteriorated. Therefore, color conversion suitable for the output device cannot always be stably performed on the image signal. Such a problem has a similar problem in color conversion using an intermediate signal as in the flow of [3], [4], [6], and [7] in FIG.

In order to solve the above-mentioned problems of the prior art, the present invention provides a method for outputting a color image signal as an output image signal, the spectral sensitivity distribution of an input device such as a digital still camera or a scanner for obtaining the image signal. Considering information on the illumination light source or information on the spectral sensitivity distribution of the intermediate signal, information on the spectral sensitivity distribution of the output device or the spectral intensity distribution of the illumination light source, or information on the spectral sensitivity distribution of the intermediate signal, A color conversion coefficient deriving method, a color conversion method, a recording medium, and a recording medium that derive a color conversion coefficient that can always perform stable color conversion without deteriorating color reproducibility even for an illumination light source or a subject. It is an object to provide a color conversion device.

[0014]

SUMMARY OF THE INVENTION The present invention provides a method for converting a first image signal characterized by a first spectral sensitivity distribution into a second image signal characterized by a second spectral sensitivity distribution. A method of deriving a conversion coefficient, wherein the conversion coefficient is derived from the first spectral sensitivity distribution and the second spectral sensitivity distribution, or the first spectral sensitivity distribution and the first spectral sensitivity distribution
Deriving a conversion coefficient from the spectral intensity distribution of the first illumination light, the second spectral sensitivity distribution further characterizing the image signal, and the spectral intensity distribution of the second illumination light further characterizing the second image signal. A color conversion coefficient deriving method is provided.

Here, the first spectral sensitivity distribution is represented by D
_{When j} (λ) (j = 1 to n) and the second spectral sensitivity distribution is E _i (λ) (i = 1 to m), the conversion coefficient is d ₁ shown in the following equation (1). It is preferably a coefficient k _ij that minimizes (D, E _i ) (i = 1 to m).

(Equation 7) However, E _i ^* in equation (1) is represented by equation (2) below.

(Equation 8) Furthermore, it is preferable to derive the coefficient k _ij under the constraint that the value of E _i ^* (λ) in the above equation (1) is non-negative. That is, it is preferable that the first image signal or the second image signal is non-negative in the same wavelength region as the wavelength region of the spectral sensitivity distribution.

A color conversion coefficient deriving method for deriving a conversion coefficient for performing color conversion from an input image signal characterized by an input spectral sensitivity distribution to an output image signal characterized by an output spectral sensitivity distribution, A conversion coefficient for color conversion from an input image signal to an intermediate signal characterized by a predetermined spectral sensitivity distribution is a _lj (l = 1 to k, j = 1 to n), and a white balance for white balance adjustment on the intermediate signal Adjustment factor is b _ll
(L = 1 to k) and a conversion coefficient for color conversion from the intermediate signal to the output image signal is c _il (i = 1 to m, l =
1 to k), conversion coefficients f _ij (i = 1 to m, j = 1 to n) for color conversion from the input image signal to the output image signal
_Derives at least one of the conversion coefficient a _lj and the conversion coefficient c _il using the color conversion coefficient deriving method,
It is preferably determined by synthesis according to the following formula (3).

(Equation 9) At this time, the spectral sensitivity distribution characterizing the intermediate signal is preferably non-negative. That is, it is preferable that the input image signal and the output image signal be non-negative in the same wavelength range as the wavelength range of the spectral sensitivity distribution.

The spectral intensity distribution of the first illumination light is P (λ), and the first spectral sensitivity distribution is D _j (λ) (j =
1 to n), the spectral intensity distribution of the second illumination light is Q
(Λ), the second spectral sensitivity distribution is represented by E _i (λ) (i = 1
To m), the conversion coefficient is d shown in the following equation (4).
_It is preferably a coefficient k _ij that minimizes ₂ (D, E _i ) (i = 1 to m).

(Equation 10) However, E _i ^* in Expression (4) is represented by Expression (5) below.

[Equation 11] Here, P (λ) · E _i ^* (λ) in the above equation (4)
It is preferable to derive the coefficient k _ij under the constraint that the value of is non-negative. That is, it is preferable that the first image signal or the second image signal is non-negative in the same wavelength region as the wavelength region of the spectral sensitivity distribution.

The spectral intensity distribution of the input illumination light and the input component
From the input image signal characterized by the light sensitivity distribution, the output illumination
Characterized by bright light spectral intensity distribution and output spectral sensitivity distribution
Derive conversion coefficients for performing color conversion on output image signals
A color conversion coefficient deriving method that determines a predetermined color conversion coefficient from an input image signal.
With the spectral intensity distribution of the illumination light source and the predetermined spectral sensitivity distribution
A is a conversion coefficient for color conversion into an intermediate signal to be characterized
_lj(L = 1 to k, j = 1 to n), a white background
The white balance adjustment coefficient for performing lance adjustment is b_ll(L = 1 ~
k) and color from the intermediate signal to the output image signal
The conversion coefficient to be converted is c _il(I = 1 to m, l = 1 to k) and
Color conversion from the input image signal to the output image signal
Conversion factor f_ij(I = 1 to m, j = 1 to n)
Coefficient a_ljAnd the conversion coefficient c_ilAt least one of
Derived using the above color conversion coefficient deriving method, the following equation (6)
Preferred by being synthesized by
No.

[Equation 12] Here, the spectral sensitivity distribution characterizing the intermediate signal is preferably non-negative. That is, it is preferable that the input image signal and the output image signal be non-negative in the same wavelength range as the wavelength range of the spectral sensitivity distribution.

Further, the present invention provides a color conversion method for a color image signal, wherein the color conversion is performed using the conversion coefficient derived by the above-described color conversion coefficient deriving method.

Further, according to the present invention, the first image signal is converted to the second image signal.
When performing color conversion to the image signal of the first, a first spectral sensitivity distribution characterizing the first image signal and a second spectral sensitivity distribution characterizing the second image signal are obtained, or
A first spectral sensitivity distribution characterizing the first image signal, a spectral intensity distribution of the first illumination light, and a second spectral sensitivity distribution and a spectral intensity distribution of the second illumination light characterizing the second image signal. Acquiring the spectral characteristic information to be acquired;
A conversion coefficient for color-converting the first image signal is derived from the spectral sensitivity distribution and the second spectral sensitivity distribution, or the first spectral sensitivity distribution and the spectral intensity of the first illumination light are derived. A derivation procedure for deriving a conversion coefficient from the distribution, the second spectral sensitivity distribution, and the spectral intensity distribution of the second illumination light; and performing a second color conversion on the first image signal using the conversion coefficient. A computer-readable recording medium storing a program for causing a computer to execute a color conversion procedure for obtaining an image signal of the above, and performing color conversion of the first image signal to the second image signal. I will provide a.

Further, according to the present invention, a first image signal characterized by a first spectral sensitivity distribution is obtained by using a conversion coefficient.
A color conversion device for converting a color image signal into a second image signal which is characterized by a second spectral sensitivity distribution and outputting the image signal, wherein the first spectral sensitivity distribution and the second spectral sensitivity distribution are obtained. Alternatively, in addition to these spectral sensitivity distributions, the first spectral sensitivity distribution and the spectral intensity distribution of the first illumination light, which further characterizes the first image signal, and the second spectral sensitivity distribution.
A spectral characteristic information acquiring unit that acquires a spectral intensity distribution of second illumination light that further characterizes the second image signal together with the spectral sensitivity distribution of the second image signal; and converts the first image signal into the second image signal. A conversion coefficient is derived from the first spectral sensitivity distribution and the second spectral sensitivity distribution, or the first spectral sensitivity distribution, the spectral intensity distribution of the first illumination light, and the second spectral sensitivity A conversion coefficient deriving unit for deriving a conversion coefficient from the distribution and the spectral intensity distribution of the second illumination light; and converting the conversion coefficient derived by the conversion coefficient deriving unit from the first image signal to the second image. A color conversion device for a color image signal, comprising: a color conversion processing unit that performs color conversion into a signal.

[0022]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, a color conversion apparatus, a color conversion method and a recording medium according to the present invention will be described in detail based on preferred embodiments shown in the accompanying drawings.

FIG. 1 shows an example of a color conversion apparatus 10 as an example of a color conversion apparatus according to the present invention, and an image output as an example in which an input device such as a digital still camera or a scanner or an output device such as an image display monitor is connected. 1 shows a system 12. The image output system 12 includes a digital still camera 14 as an input device for obtaining a digital image signal (hereinafter, referred to as an image signal) by performing photographing and the like, a scanner 16 for reading a document for a printing plate, and an image for obtaining a desired image. An image processing device 18 that performs image processing on a signal, a color conversion device 10 according to the present invention that converts an image signal processed by the image processing device 18 into an image signal suitable for an output device, and a color-converted image signal NTSC display monitor 20 and HDTV for image display
Display monitor 22 and CTP output device 2 for creating a printing plate
And a control device 26 for controlling and controlling the operations of the scanner 14, the digital still camera 16, the image processing device 18, and the color conversion device 10 and transferring various information to each device. The media drive device 30 is connected to the color conversion device 10. The scanner 14 and the input device of the digital still camera 16 and the NTSC display monitor 20
And the like can be connected to the image processing device 18 and the color conversion device 10 as appropriate, and the system of the present embodiment uses the input device of the digital still camera 14 and the scanner 16 and the NTSC display monitor for easy explanation. 20 and HD
A plurality of output devices of the TV display monitor 22 and the CTP output device 24 are connected.

The digital still camera 14 photoelectrically captures an image using a known image pickup device such as a CCD image pickup device, and converts an RGB image signal, that is, a color image signal composed of an R signal, a G signal and a B signal. Output. FIG. 4 (a)
2 shows the spectral sensitivity distribution of the digital still camera 14. The scanner 16 is an image reading apparatus having a spectral sensitivity distribution shown in FIG. 4B, and is of a transmission type or a reflection type which projects light on a print document and photoelectrically reads transmitted light or reflected light to obtain an image signal. A type scanner which outputs XYZ signals of the XYZ color system, that is, X, Y and Z signals, as image signals. In this embodiment, the white balance is adjusted for the image signal, the R signal, the G signal, and the B signal, and the X, Y, and Z signals output from the digital still camera 14 and the scanner 16.

The image processing unit 18 is a unit for performing gray balance and brightness correction of images, correction of gradation, correction of distortion caused by a photographing lens, electronic zooming, and the like. The image signal obtained in step 16 is subjected to known image processing and output to the color conversion device 10.

The color conversion device 10 is an example of the color conversion device of the present invention, and performs color conversion on an image signal (first image signal) of an image processed by the image processing section 18 to display a desired image. An output image signal (second
Image signal), the details of which will be described later.

The NTSC display monitor 20 has a spectral sensitivity distribution as shown in FIG. 4 (d), the illumination light and the light D ₆₅ of CIE standard, a display monitor for use in home television image display monitor, etc. Is an image display medium that receives a transmission signal converted into a YIQ analog signal and displays an image. HDTV display monitor 22, as shown in FIG. 4 (c), has a spectral sensitivity distribution different from the spectral sensitivity distribution of the NTSC display monitor 20 shown in FIG. 4 (d), the light D ₆₅ of CIE standard illumination light A display monitor used for displaying images such as BS digital broadcasts.
An image display medium that receives a GB color image signal as a transmission signal and displays an image. The CTP output device 24 is a device for directly exposing a printing plate using a laser beam to create a desired printing plate, and is a device for controlling exposure of the laser beam using an XYZ stimulus signal.

The control device 26 is a digital still camera 1
4, scanner 16, image processing device 18, color conversion device 10,
And a command to transfer image signals from the digital still camera 14 or the scanner 16 to the image processing unit 18, and a CD-ROM or DVD loaded in the recording media drive 30. And instructs the image processing unit 18 to transfer image signals by calling image data recorded on various recording media such as a smart media and the like, or instructs the control device 26 using a keyboard, a mouse or the like (not shown). In accordance with the input content, the image processing unit 18 instructs the image processing content, and further, information of the input device that is the source of the image signal (input image signal) input to the color conversion device 10, for example, information of the input device. Identification information such as a name and a model, spectral characteristic information such as a spectral sensitivity distribution of an input device and a spectral intensity distribution of an illumination light source, an NTSC display monitor 20 and an HD Color conversion of information on output devices such as the V display monitor 22 and the CTP output device 24, for example, identification information such as the name and type of the output device, and spectral characteristic information such as the spectral sensitivity distribution of the output device and the spectral intensity distribution of the illumination light source. Apparatus 10
Transfer to

The color conversion device 10 assumes that at least the spectral sensitivity distribution of the input device and the spectral sensitivity distribution of the output device are known.
From these spectral sensitivity distributions, a conversion coefficient of a color conversion matrix for converting an image signal obtained by the input device into an output image signal suitable for the output device is derived, and the obtained conversion coefficient is used to convert the input image signal into an input image signal. This is a device that performs color conversion to output an image signal, and as shown in FIG. 2, includes a spectral characteristic information obtaining unit 10A, a color conversion coefficient deriving unit 10B, a color conversion processing unit 10C, and a memory unit 10D.

The spectral characteristic information acquiring section 10A is provided with the control device 2
6 to obtain the identification information of the input device and the output device, and the spectral characteristic information of the input device and the output device. The spectral sensitivity distribution in the input device and the spectral sensitivity distribution in the output device are obtained based on the identification information and the spectral characteristic information. This is a part for acquiring a spectral sensitivity distribution and, in some cases, a spectral intensity distribution of illumination light when obtaining an input image signal by an input device and a spectral intensity distribution of illumination light of an output device. The spectral characteristic information of the input device or the output device is numerical data of the spectral sensitivity distribution or the spectral intensity distribution of the input device or the output device, and the identification information of the input device or the output device is the name of the input device or the output device. It refers to information that can specify a spectral sensitivity distribution or a spectral intensity distribution by a model or the like. When the information about the input device and the output device transferred from the control device 26 is the identification information of the input device and the output device, the spectral sensitivity distribution and the illumination corresponding to the identification information such as the name and model of the input device and the output device are provided. Based on the obtained identification information, the spectral sensitivity distribution of the input device or the output device, or, in some cases, the spectral intensity distribution of the illumination light is called from the memory unit 10D in which the spectral intensity distribution of light is recorded. Is obtained.

The color conversion coefficient deriving section 10B is a section for deriving conversion coefficients using the obtained spectral sensitivity distributions of the input device and the output device. For example, the spectral sensitivity distribution of the input device is the spectral sensitivity distribution D _j (λ) (j = 1 to n), and the spectral sensitivity distribution of the output device is the color matching function x of the XYZ color system.
(Λ), y (λ), and z (λ), the conversion coefficient k _ij (i = 1 to 3,
j = 1 to n), the spectral sensitivity distribution D _j (λ) (j = 1
.Function. ^* (. Lambda.), Y ^* (. Lambda.) And z
^* (Λ), which is used to calculate the following formulas (8) to (1)
0), d ₁ (D, x), d ₁ (D, y) and d ₁
A transform coefficient k _ij that minimizes (D, z) is obtained.

[0032]

(Equation 13)

[Equation 14]

(Equation 15)

(Equation 16) However, the inner product operator (,) and the L ₂ norm refers to the following equation (11) and (12).

[Equation 17]

(Equation 18) Here, a conversion coefficient k _1j (j = 1 to n) is obtained from Expression (8), a conversion coefficient k _2j (j = 1 to n) from Expression (9), and a conversion coefficient k _3j ( j = 1 to n) are obtained.

FIG. 3A shows the flow of processing for deriving a conversion coefficient k _ij in the color conversion coefficient deriving section 10B using the equations (13) to (15). Each of the spectral sensitivity distributions is subjected to an integration operation with the visible region as an integration range, and (1, x ^* (λ)) or (1, x
(Λ)) and the like, and then d ₁ (D, x), d ₁ (D, y), and d ₁ (D,
The transform coefficient k _ij that minimizes z) is obtained. Conversion coefficient k _ij
Is nonlinear, but is calculated using a known method.

Further, when an input image signal is obtained by the input device,
Spectral intensity distribution P (λ) of illumination light and output device
The spectral intensity distribution Q (λ) is obtained by the spectral characteristic information acquisition unit 10A.
If acquired, as shown in equation (7), the conversion coefficient k_ij
(I = 1 to 3, j = 1 to n) and the spectral sensitivity distribution D_j
A function x represented by a linear sum of (λ) (j = 1 to n)
^*(Λ), y^*(Λ) and z^*Using (λ),
D shown in equations (13) to (15)_Two(D, x), d
_Two(D, y) and d_TwoTransformer that minimizes (D, z)
Number k _ijAsk for.

[Equation 19]

(Equation 20)

(Equation 21) Here, the inner product operator (,) and the L ₂ norm mean equations (11) and (12).

Here, the equations (8) to (10) are obtained by the equation (1)
3) to (15) when the values of P (λ) and Q (λ) are set to 1, that is, when the illumination light of the input device and the output device is completely white light. . Further, a conversion coefficient k _1j (j = 1 to n) is obtained from Expression (13),
A conversion coefficient k _2j (j = 1 to n) is obtained from Expression (14),
The conversion coefficient k _3j (j = 1 to n) is obtained from Expression (15).

FIG. 3B shows a flow of a process for deriving the conversion coefficient k _ij using the equations (13) to (15). The spectral sensitivity distribution of the input device and the output device and the distribution of the illumination light source are shown. From the spectral intensity distribution, an integral operation is performed on the product of the spectral intensity distribution and the spectral sensitivity distribution in the input device and the product of the spectral intensity distribution and the spectral sensitivity distribution in the output device, respectively, using the visible region as an integration range, and (1, P (λ ) · X ^* (λ)), (1, Q (λ) · x (λ)), etc.
A conversion coefficient k _ij of a color conversion matrix that minimizes d ₂ (D, x), d ₂ (D, y), and d ₂ (D, z) shown in (15) is obtained. Although the equation for obtaining the conversion coefficient k _ij is non-linear, it is calculated using a known method. The color conversion coefficient deriving unit 10B calculates the obtained conversion coefficient k _ij (i = 1
, 3, j = 1 to n) to the color conversion processing unit 10C.

The color conversion processing section 10C converts the conversion coefficients k _ij (i = 1 to 3, j = 1) sent from the color conversion coefficient deriving section 10B.
To n) as the conversion coefficients of the color conversion matrix K, and performs color conversion using the color conversion matrix K from an image signal obtained by the input device and subjected to image processing by the image processing unit 18 to produce an output suitable for the output device. This is a part for outputting an image signal.
That is, the spectral sensitivity distribution D _j (λ) (j = 1) of the input device
To the input image signal composed of the image signals S _j (j = 1 to n) corresponding to the color matching functions x (λ) and y, which are the spectral sensitivity distributions of the output device, according to the following equation (16). (Λ)
And X (Y) and Z (λ), which are also tristimulus values.

(Equation 22)

The conversion coefficient k _ij of the color conversion matrix K obtained by the color conversion coefficient deriving unit 10B and used by the color conversion processing unit 10C is, for example, the digital still camera 14 and the scanner 16 as input devices, and the HDTV monitor 22,
The conversion coefficients output from the NTSC display monitor 20 and the CTP output device 24 take values as shown in Table 1 below. The reason why the color conversion matrix K is a 3 × 3 matrix is that the spectral sensitivity distributions of the digital still camera 14 and the scanner 16 are the three spectral sensitivity distributions of R, G and B, the color matching function x (λ), The input image signal is composed of three spectral sensitivity distributions of y (λ) and z (λ), and the input image signal is also composed of image signals of R, G and B, and signals of three stimulus values of X, Y and Z. In the output device, the three spectral sensitivity distributions of R, G, and B, and the color matching functions x (λ), y
(Λ) and z (λ), and the output image signal is also composed of R, G, and B image signals, and X, Y, and Z tristimulus value signals. Because.

[Table 1]

Since such a color conversion matrix K is uniquely determined when an input device and an output device are set,
The color conversion matrix K is recorded in advance in the memory unit 10D for each combination of the input device and the output device, and based on the identification information of the input device and the output device transferred from the control device 26 in the spectral characteristic information acquisition unit 10A. Memory unit 10
A conversion matrix K is called from D, and a color conversion processing unit 10 is called.
It is also possible to configure to transfer to C. Thereby, the calculation process performed by the color conversion coefficient deriving unit 10B can be omitted, and the color conversion process can be performed quickly.

The color conversion coefficient deriving unit 10B in this embodiment.
Calculates the conversion coefficient k _ij from Expressions (8) to (10) or Expressions (13) to (15) using Expression (7). At the time of calculation, x ^* ( It is preferable to apply a constraint to the transform coefficients such that λ), y ^* (λ) and z ^* (λ) are non-negative. Thereby, as described later, stable color conversion independent of the illumination light source and the subject can be achieved. The color conversion device 10 is configured as described above.

Next, a method of deriving a color conversion coefficient of the present invention will be described based on an image output system 12 using a color conversion device 10 which is an example of the color conversion device of the present invention.

First, in the control device 26, identification information such as the name and model of an input device which is an input source of an image signal and an output device which is an output destination of the image signal is input from a mouse or a keyboard (not shown). Characteristic information such as numerical data of the spectral intensity distribution and spectral sensitivity distribution of the output device and the spectral sensitivity distribution are manually input from a mouse or a keyboard, or the spectral characteristic information recorded on the recording medium is called out via the recording medium drive device 30. , To the color conversion device 10. On the other hand, in the spectral characteristic information acquisition unit 10A of the color conversion device 10, the identification information of the input device and the output device and the spectral characteristic information of the input device and the output device transferred from the control device 26 and the recording media device 30 are acquired. ,
Based on the identification information and the spectral characteristic information, the spectral sensitivity distribution of the input device, the spectral sensitivity distribution of the output device, and, in some cases, the spectral intensity distribution of the illumination light when the input image signal is obtained by the input device. And the spectral intensity distribution of the illumination light of the output device (acquisition of spectral characteristic information).

When the information transferred from the control device 26 is the identification information of the input device or the output device, it corresponds to the identification information such as the name or model of the input device or the output device based on the obtained identification information. From the memory unit 10D in which the spectral sensitivity distribution and the spectral intensity distribution of the illumination light are recorded, the spectral sensitivity distribution of the input device and the output device, or in some cases, the spectral intensity distribution of the illumination light is called and acquired.

The spectral sensitivity distributions of the input device and the output device obtained in this way are transferred to the color conversion coefficient deriving section 10B, and in some cases, together with these spectral sensitivity distributions, when the input device acquires an image signal. Is transmitted to the color conversion coefficient deriving unit 10B.

In the color conversion coefficient deriving section 10B, the spectral sensitivity distribution D _j (λ) (j = 1 to n) of the input device and the spectral sensitivity distribution of the output device, for example, the color matching functions x (λ) and y (λ) ) And z (λ) described above, d ₁ (D, x), d (d) are calculated using equations (8) to (10).
A transform coefficient k _ij (i = 1 to 3, j = 1 to n) that minimizes ₁ (D, y) and d ₁ (D, z) is derived (derivation of a transform coefficient). Spectral intensity distribution P of illumination light in input device
When (λ) and the spectral intensity distribution Q (λ) in the output device are obtained, d is calculated using Expressions (13) to (15).
₂ (D, x), d ₂ (D, y) and d ₂ (D, z) are derived to obtain conversion coefficients k _ij (i = 1 to 3, j = 1 to n). Derived).

The conversion coefficients thus obtained are transferred to the color conversion processing unit 10C, and the color conversion processing unit 10C creates a color conversion matrix K using the conversion coefficients as matrix components.

On the other hand, according to the instruction from the control device 26,
An image signal of an image captured by the digital still camera 14 or an image signal of an image read by the scanner 16 is called, or an image signal recorded on a CD-ROM, a DVD, a smart media, or the like is transmitted to a recording media drive device. 30 and called out through the image processing apparatus 1
After the image data is subjected to desired image processing, it is input to the color conversion processing unit 10C of the color conversion device 10 as an input image signal.

The color conversion processing section 10C performs a matrix calculation on the transferred input image signal using the color conversion matrix K to obtain an output image signal suitable for the output device (color conversion). The obtained output image signal is output to the NTSC display monitor 20, HDTV display monitor 22, or CTP output device 24 according to the instruction of the control device 26.

According to the color conversion method of the present invention, the color conversion coefficients k _ij (i = 1 to m, j = 1 to
n), the color conversion processing unit 10C performs color conversion on the input image signal. The color conversion matrix K derived in advance is previously recorded in the memory unit 10D for each combination of the input device and the output device. In advance, the spectral characteristic information acquisition unit 10
In step A, a conversion matrix K determined by a combination of the input device and the output device is called based on the identification information of the input device and the output device transferred from the control device 26, and the color conversion processing unit 10C uses the conversion matrix K to A conversion may be performed. In this way, an output image signal suitable for the output device is output, and an image signal that performs the same color reproduction as the image acquired by the input device can be provided to the output device.

The color conversion coefficient deriving method of the present invention is obtained by the following equation (8).
-D ₁ (D, E _i ) in equation (1) that generalizes equation (10)
(I = 1 to m) or d ₂ (D, E _i ) (i = 1 to m) in equation (4), which is a generalization of equations (13) to (15).
Is minimized by appropriately setting the conversion coefficient, which is derived based on the following idea.

Originally, humans do not perceive the color of light in a form proportional to the spectral waveform, but perceive the color by a relative comparison based on a reference color, for example, perfect white. Moreover, the degree of color perception is not linear but non-linear. This means that the values of L ^* , a ^* and b ^* in a uniform color space that optimally quantify the color perceived by humans are X
Relative ratios of the tristimulus values X, Y, and Z in the YZ color system with respect to the tristimulus values X ₀ , Y _0, and Z ₀ on the perfect diffuse reflection surface, ie, X / X ₀ , Y / Y ₀ , Z / Z
_It can also be seen from the fact that it is represented by a linear combination of values obtained by raising ₀ to the third power.

The method for deriving the color conversion coefficient according to the present invention is described above.
It is derived taking into account the perceptual characteristics of human color.
For example, a uniform color sky taking into account the perceived characteristics of human color
L ^*a^*b^*In consideration of the input image signal of the input device,
L obtained^*, A^*And b^*To the output device.
L of the color to display^*, A^*And b^*Best value for
This is to obtain a conversion coefficient that approximates well.

[0053] L in the input device ^*, a L ₁ ^{* *} and b ^* values, respectively, a ₁ ^* and b ₁ ^*, 3 stimulus values as X _1, Y ₁ and Z _1, L in the output device ^*, a
The values of ^* and b ^* are L ₂ ^* , a ₂ ^* and b _{2 respectively.}
^*, If the tristimulus value was X _2, Y ₂ and Z _2, the input device in L ₁ ^*, a ₁ ^* and b ₁ ^* of the values, L ₂ ^* in the output device, a ₂ ^* and b ₂ ^* Difference from the value of Δ
Considering L ^* , Δa ^*, and Δb ^* , the expected value is evaluated as in the following equations (17) to (19). Where E
[] Is an operator representing an expected value when the spectral reflectance distribution ρ _l of the subject is statistically dispersed.

E [ΔL ^* ] = 116 · E [Δ (Y / Y ₀ ) ^1/3 ] −16 (17) E [Δa ^* ] ≦ 500 · E [Δ (X / X ₀ ) ^1/3 ] + 500 · E [Δ (Y / Y ₀ ) ^1/3 ] (18) E [Δb ^* ] ≦ 500 · E [Δ (Y / Y ₀ ) ^1/3 ] + 500 · E [Δ (Z / Z ₀ ) ^1/3 ] (19) Here, the expected values E [ΔL ^* ], E [Δa ^* ] and E
To minimize [Δb ^* ], E [Δ (X /
X ₀ ) ^1/3 ], E [Δ (Y / Y ₀ ) ^1/3 ] E [Δ (Z /
Z ₀ ) ^1/3 ] should be minimized. That is, to make the values of L ^* , a ^* and b ^* in the output device most approximate to the values of L ^* , a ^* and b ^* in the input device, E [Δ (X / X ₀ ) ^1/3 ] , E [Δ (Y / Y ₀ )
^1/3 ] and E [Δ (Z / Z ₀ ) ^1/3 ].

Therefore, as one example, the spectral sensitivity distribution D _j (λ) (j = 1 to n) of the input device, the spectral intensity distribution P (λ) of the illumination light in the input device, and the spectral sensitivity distribution x of the output device (Λ), y (λ) and z (λ) (x (λ), y
(Λ) and z (λ) are color-matching functions of the XYZ color system) and the above-described example in which the spectral intensity distribution Q (λ) in the output device is obtained, and the expected value E [Δ (X / X ₀ )
^1/3 ] is required. Δ (X
The expected value of / X ₀ ) ^1/3 is expressed by the following equation (20), where X ^* ₁ is expressed by the equation (7) among the three stimulus values XYZ for the subject having the spectral reflectance ρ _l. Spectral sensitivity distribution x formed by linear sum of spectral sensitivity distribution D _j (λ) of input device
^* X stimulus value characterized by (λ), where X _l is the color matching function of the output device x (λ), y (λ) and z (λ)
Characterized by a X stimulus value among the tristimulus values of a subject having a spectral reflectance distribution [rho _l.

(Equation 23)

By rearranging the above equation (20), the following equation (2)
1) is derived.

(Equation 24) Here, _Denom is represented by the following equation (22).

(Equation 25)

As can be seen from equation (21), equation (21)
By making the numerator on the right side close to 0, that is, the equation (1)
By making the right side of 3) close to 0, the expected value E [Δ (X /
X₀) ^1/3] Can be minimized. Equation (2)
As shown in 1), in equation (21),
Spectral reflectance distribution ρ_l(Λ) corresponds to the right side of equation (13)
Term, the spectral reflectance distribution ρ of the subject
_l(Λ), the expected value E [Δ (X / X₀)
^1/3] Can be minimized. In a similar way,
Expected value E [Δ (Y / Y₀)^1/3] And the expected value E [Δ
(Z / Z₀)^1/3] Can be minimized. This
Such minimization is performed, as described above, by the transformation expressed by equation (7).
Conversion factor k_ijIs appropriately determined, and equations (14) and (15) are
Achieved by minimizing.

By the way, according to the conventional method, the spectral sensitivity distribution of the output device is set so that the output image signal has the same value as the tristimulus value or the RGB image signal obtained from the input image signal of the color sample object such as the Macbeth chart. The conversion coefficient of the conversion matrix is determined according to the spectral intensity distribution of the illumination light, that is, the spectral waveform of the photographing object that is most approximated to the input image signal, the spectral sensitivity distribution of the input device, and the illumination intensity distribution of the illumination light. The conversion coefficient of the conversion matrix is determined so as to have a spectral waveform to be obtained. Therefore, in the conventional method,
The conversion coefficient must always be calculated using a subject such as a Macbeth chart, and the conversion coefficient greatly depends on the spectral reflectance distribution of the selected subject. Therefore, the range in which the conversion coefficient color conversion can be appropriately performed using the calculated conversion coefficient as a coefficient of the color conversion matrix is narrow. In order to perform appropriate color conversion over a wide range, it is necessary to calculate conversion coefficients by using a large number of color samples having various spectral reflectance distributions as subjects.

Further, in the conventional method, the color conversion of the output image signal is performed so as to have a spectral waveform most similar to the spectral waveform represented by the input image signal of the subject such as a Macbeth chart. Do not always match. Furthermore, since there are input devices and output devices whose spectral sensitivity distributions show negative values, the value of the matrix conversion coefficient to be obtained becomes unstable depending on the spectral intensity distribution and spectral sensitivity distribution of the illumination light, and the stable and favorable In some cases, it may not be possible to perform color conversion.

However, according to the color conversion coefficient deriving method of the present invention, the conversion coefficient is obtained in a manner not depending on the spectral reflectance distribution ρ _l of the subject, as shown in equation (21).
It is not necessary to calculate the conversion coefficient using a color sample such as a Macbeth chart, and as a result, the conversion coefficient does not depend on the spectral reflectance distribution of the chart. Equations (8) to (1)
When the minimization is performed in 0), the conversion coefficient is derived without including the spectral intensity distribution of the light source, so that the conversion coefficient independent of the light source can be obtained. Therefore, the conversion coefficient does not depend on the spectral intensity distribution of the light source. Further, in the uniform color space L ^* a ^* b ^* , the output image signal is color-converted so that the L ^* value, the a ^* value, and the b ^* value are best approximated. Color conversion can be realized. Further, in Expressions (13) to (15),
Since Q (λ) · x (λ) is standardized as shown as Q (λ) · x (λ) / (1, Q (λ) · x (λ)), for example, Even when the sensitivity distribution has a negative region, and a value such as (1, Q (λ) · x (λ)) shows a value close to 0, the expression (13) to the expression (15) Since the obtained value has a finite value, a stable conversion coefficient k _ij can be obtained.

The method for deriving the color conversion coefficients in this embodiment is as follows.
Using the equation (7), the equations (8) to (10) or the equation (1)
From 3) to (15), the conversion coefficient k_ijIs to derive
Is P (λ) · x when deriving the conversion coefficient.^*(Λ), P
(Λ) · y^*(Λ) and P (λ) · z^*(Λ) is non-negative
It is desirable to set constraints on the conversion coefficients so that
No. For example, equation (13) is derived from equation (21)
Is the denominator D in equation (21)_{en om}Is changed so that
Conversion factor k_ij, Where P (λ),
Consider that Q (λ) and x (λ) are greater than or equal to 0
And P (λ) · x^*(Λ) is preferably non-negative
It is because that is led. Similarly, P (λ) · y
^*(Λ) and P (λ) · z^*(Λ) is also non-negative
Is preferred. Where x^*(Λ), y
^*(Λ) and z^*(Λ) is expressed by equation (7)
From P (λ) x^*(Λ), P (λ) · y^*(Λ)
And P (λ) · z^*Constraint such that (λ) is non-negative
Conversion factor k below_ijMay be derived. in this way,
P (λ) x^*(Λ), P (λ) · y^*(Λ) and
P (λ) · z^*A transform coefficient k for which (λ) is non-negative_ijDerived
By doing, the denominator D_enomIs never zero,
Achieves more stable color conversion independent of illumination light source and subject
it can.

When the spectral intensity distribution P (λ) of the illumination light source in the input device and the spectral intensity distribution Q (λ) of the illumination light source in the output device are not obtained, the expressions (13) to (1) are obtained.
In 5), the illumination light source is a white light source, that is, P
(Λ) = Q (λ) ≡1. Alternatively, equation (13)
To (15), C.I. F. Borges approximation formula (J. Opt. Soc. Am, 8 (8), p1319−
1323 (1991)), Equations (8) to (10) can be obtained using Equations (23) and (24) below.

(Equation 26)

[Equation 27] As a result, d represented by Expressions (8) to (10) is obtained.
By calculating a conversion coefficient k _ij that minimizes ₁ (D, x), d ₁ (D, y) and d ₁ (D, y), the expected value E [Δ (X / X ₀ ) ^1/3 ], E [Δ (Y /
Y ₀ ) ^1/3 ] and E [Δ (Z / Z ₀ ) ^1/3 ] can be minimized, thus minimizing E [ΔL ^* ], E [Δa ^* ] and E [Δb ^* ]. can do.

Here, it does not depend on the illumination light source or the subject.
To achieve stable color conversion, x ^*(Λ), y
^*(Λ) and z^*(Λ) is preferably non-negative
Is derived from equation (21). Where x
^*(Λ), y^*(Λ) and z^*(Λ) is the equation
From the expression (7), x^*(Λ), y^*(Λ),
And z^*Under the constraint that (λ) is non-negative
Conversion factor k_ijMay be derived. Such a conversion coefficient k
_ijBy deriving the
Stable color conversion can be achieved.

The conversion coefficients for the color conversion are derived from the device image signal (input image signal) as shown in FIG. Signal)
It can be carried out. For example, [4] → [5] →
This will be described using the route [7]. The device image signal is characterized by a spectral sensitivity distribution D _j (λ) (j = 1 to n), the intermediate signal is characterized by a spectral sensitivity distribution F _l (λ) (l = 1 to k), and the monitor image signal Is the spectral sensitivity distribution E _i (λ) (i =
1 to m), a conversion coefficient a _lj (l = 1 to k, j =
1 to n) are obtained by deriving a coefficient that minimizes the left-hand side of equation (1), or E.sub.1 defined by equation (2).
^* _i (λ) is obtained by deriving a coefficient that minimizes the left side of equation (1) under the constraint that it is non-negative, and a conversion coefficient c for performing color conversion from the intermediate signal to the monitor image signal.
_il (i = 1 to m, l = 1 to k) is obtained by deriving a coefficient that minimizes the left side of equation (1), or E ^* (λ) defined by equation (2) is It is obtained by deriving a coefficient that minimizes the left side of Equation (1) under a non-negative constraint.

Further, white balance adjustment in the intermediate signal,
That is, in the white standard, a white balance adjustment coefficient b _ll (l = 1 to 1) for adjusting the value of the image signal to a predetermined value.
k). Such a white balance adjustment coefficient b _ll is
The device image signal of the white standard is taken in advance, and it is determined based on this. Here, the white balance adjustment coefficient b _ll
Simply changes the contribution ratio of each spectral waveform of the spectral sensitivity distribution F _l (λ) of the intermediate signal to make the white standard.
The coefficient that each spectral waveform exerts on each other is 0.

Using the transform coefficients obtained in this way, the transforms [4], [5] and [7] are synthesized according to equation (3), and [4] → [5] shown in FIG. → A conversion coefficient f _ij for color conversion along the path [7] is obtained. The device image signal is subjected to color conversion using the conversion coefficient f _ij to obtain a white balance adjusted monitor image signal.

In the above example, [4] and [7] in FIG.
Are derived using equation (1). In the present invention, both of the conversion coefficients [4] and [7] are calculated using equation (1).
It is not necessary to derive from Equation (1), and either one may be derived from Equation (1), and the remaining transform coefficients may be obtained using a known conventional method. In the above example, the case where the spectral intensity distribution of the illumination light source is not taken into account is described. However, when the spectral intensity distribution of the illumination light source is obtained, as in the above method, [4], [5], and [7] ] Are obtained, and these are combined, and [4] → [5] → [7]
May be obtained, and color conversion may be performed using this.

Table 2 shows specific examples of the conversion coefficient a _lj and the conversion coefficient c _il obtained under the constraint that E ^* _i (λ) is non-negative.
Listed in Table 2 shows the device image signals in FIG.
A device image signal photographed by a digital still camera characterized by the spectral sensitivity distribution shown in FIG. 4A is used as a monitor image signal for an HDTV display monitor signal (FIG. 4C). BT709 signal) and the conversion coefficient a _lj and the conversion coefficient c _il when the LMS signal characterized by the spectral sensitivity distribution of the physiological primary color (LMS) shown in FIG. When an XYZ signal using X, Y and Z signals, which are tristimulus values of an XYZ color system characterized by color matching functions x (λ), y (λ) and z (λ), as image signals is used as a signal. shows the coefficient of conversion a _lj and transform coefficients c _il. Note that the intermediate signal is a spectral signal like the color matching functions x (λ), y (λ), and z (λ) which are the spectral sensitivity distribution of the LMS signal and the spectral sensitivity distribution of the XYZ signal shown in FIG. It is preferable to use a signal characterized by a spectral sensitivity distribution that is non-negative in a predetermined range of the sensitivity distribution, for example, a wavelength range of 400 to 800 nm.

[0069]

[Table 2]

Although the derivation of the conversion coefficients for color conversion has been described above, the derivation of the conversion coefficients may be performed by a color conversion device using a dedicated circuit, or the computer exhibits the above-described effects. It may be processed by a program. When processing is performed by a program, the program including the above-described steps (procedures) is recorded on a recording medium such as a CD-ROM, and is executed as needed by being read by a computer. The present invention provides a recording medium on which such a program is recorded. Further, such a program may be processed by adding various image processing functions in the image processing device 18 or may be processed together with white balance adjustment.

Although the color conversion coefficient derivation method, color conversion method, recording medium and color conversion device of the present invention have been described in detail above, the present invention is not limited to the above embodiments, and does not depart from the gist of the present invention. Of course, various improvements and modifications may be made within the scope.

[0072]

As described above in detail, when a color image signal is output as an output image signal, the spectral sensitivity distribution of an input device such as a digital still camera or a scanner for obtaining the image signal and the spectral information of an illumination light source are obtained. In consideration of the spectral sensitivity distribution of the output device and the spectral information of the spectral intensity distribution of the illumination light source, stable color conversion can be achieved for any illumination light source or subject without deteriorating color reproducibility. Transform coefficients can be derived. Further, E _i ^* (λ) in equation (1) and P (λ) · E in equation (4)
By deriving the conversion coefficient under the constraint that the value of _i ^* (λ) is non-negative, it is possible to derive a conversion coefficient that achieves more stable color conversion regardless of the illumination light source or the subject.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of an example of an image output system using a color conversion device according to the present invention.

FIG. 2 is a block diagram showing a configuration of a color conversion device used in the image output system shown in FIG.

FIGS. 3A and 3B are diagrams illustrating the flow of a color conversion method according to the present invention.

FIGS. 4A to 4E are diagrams illustrating an example of a spectral sensitivity distribution of an input device or an output device.

FIG. 5 is a diagram illustrating a flow of an image signal until a monitor image is obtained from a subject.

FIG. 6 is a diagram illustrating a conventional flow for obtaining a conversion coefficient.

[Explanation of symbols]

 Reference Signs List 10 color conversion device 10A spectral characteristic information acquisition unit 10B conversion coefficient derivation unit 10C color conversion processing unit 10D memory unit 12 image output system 14 digital still camera 16 scanner 18 image processing device 20 NTSC display monitor 22 HDTV display monitor 24 CTP output device 26 Control device

 ──────────────────────────────────────────────────続 き Continued on the front page F-term (reference) 5B057 AA11 BA02 BA19 CE17 DA17 DB02 DB06 5C066 AA01 AA11 CA08 DA05 EA13 EB02 EC01 EE04 FA02 GA01 GB01 JA01 KA12 KD06 KE02 KE03 KE07 KF05 KM01 KM11 5C037 MM27PP08 SS01 SS06 TT02 TT09 5C079 HB01 HB05 HB08 JA12 JA23 LA02 LA23 LB02 MA11 NA03 NA15

Claims (12)

[Claims] A first characteristic characterized by a first spectral sensitivity distribution.
Is a method for deriving a conversion coefficient for converting the image signal of (a) into a second image signal characterized by a second spectral sensitivity distribution, wherein the first spectral sensitivity distribution and the second spectral sensitivity distribution Deriving the transform coefficient, or a first spectral sensitivity distribution, a first image signal further characterizing the first image signal.
Deriving a conversion coefficient from the spectral intensity distribution of the illumination light, the second spectral sensitivity distribution, and the spectral intensity distribution of the second illumination light further characterizing the second image signal. Method. 2. The method according to claim 1, wherein the first spectral sensitivity distribution is D _j (λ) (j
= 1 to n), and the second spectral sensitivity distribution is E _i (λ)
When (i = 1 to m), the conversion coefficient is d ₁ (D, E _i ) shown in the following equation (1).
2. The color conversion coefficient deriving method according to claim 1, wherein the coefficient k _ij minimizes (i = 1 to m). (Equation 1) However, E _i ^* in equation (1) is represented by equation (2) below. (Equation 2) 3. The coefficient k _ij is derived under a constraint that the value of E _i ^* (λ) in the equation (1) is non-negative.
The color conversion coefficient derivation method described in 1. 4. A color conversion coefficient deriving method for deriving a conversion coefficient for performing color conversion from an input image signal characterized by an input spectral sensitivity distribution to an output image signal characterized by an output spectral sensitivity distribution, A conversion coefficient for color conversion from an input image signal to an intermediate signal characterized by a predetermined spectral sensitivity distribution is a _lj (l = 1 to k, j
= 1 to n), a white balance adjustment coefficient for performing white balance adjustment on the intermediate signal is b _ll (l = 1 to k), and a conversion coefficient for performing color conversion from the intermediate signal to the output image signal is c _il. (I = 1 to m, l = 1 to k), the conversion coefficient f of the color conversion from the input image signal to the output image signal
_ij (i = 1 to m, j = 1 to n) derives at least one of the conversion coefficient a _lj and the conversion coefficient c _il using the color conversion coefficient deriving method according to claim 2 or 3. ,
A color conversion coefficient deriving method characterized by being obtained by combining by the following equation (3). (Equation 3) 5. The color conversion coefficient deriving method according to claim 4, wherein the spectral sensitivity distribution characterizing the intermediate signal is non-negative. 6. The spectral intensity distribution of the first illumination light is represented by P
(Λ), the first spectral sensitivity distribution is represented by D _j (λ) (j = 1
To n), the spectral intensity distribution of the second illumination light is Q (λ),
When the second spectral sensitivity distribution is E _i (λ) (i = 1 to m), the conversion coefficient is d ₂ (D, E _i ) shown in the following equation (4).
2. The color conversion coefficient deriving method according to claim 1, wherein the coefficient k _ij minimizes (i = 1 to m). (Equation 4) However, E _i ^* in Expression (4) is represented by Expression (5) below. (Equation 5) 7. P (λ) · E in equation (4)
7. The color conversion coefficient deriving method according to claim 6, wherein the coefficient k _ij is derived under a constraint that the value of _i ^* (λ) is non-negative. 8. The spectral intensity distribution of input illumination light and input spectral sensitivity.
From the input image signal characterized by the distribution, the output illumination light
Output characterized by spectral intensity distribution and output spectral sensitivity distribution
Color transformation to derive transformation coefficients for performing color transformation on image signals
A method for deriving a transposition coefficient, comprising:
Color conversion to an intermediate signal characterized by a given spectral sensitivity distribution
Conversion factor_lj(L = 1 to k, j = 1 to n),
White balance adjustment coefficient for performing white balance adjustment on the intermediate signal
To b_ll(L = 1 to k), and from the intermediate signal
The conversion coefficient for performing color conversion to the output image signal is c _il(I = 1 ~
m, l = 1 to k), the conversion coefficient f of the color conversion from the input image signal to the output image signal
_ij(I = 1 to m, j = 1 to n) are the conversion coefficients a_ljYou
And the conversion coefficient c_ilClaim 6 at least one of
Or using the color conversion coefficient derivation method described in 7,
It is determined by combining with the following equation (6).
A color conversion coefficient deriving method characterized in that: (Equation 6) 9. The color conversion coefficient deriving method according to claim 8, wherein the spectral sensitivity distribution characterizing the intermediate signal is non-negative. 10. A color conversion method for a color image signal, wherein color conversion is performed using a conversion coefficient derived by the color conversion coefficient derivation method according to claim 1. Conversion method. 11. When the first image signal is color-converted into a second image signal, a first spectral sensitivity distribution characterizing the first image signal and a second spectral sensitivity characterizing the second image signal. A first spectral sensitivity distribution that characterizes the first image signal and a spectral intensity distribution of the first illumination light, and a second spectral sensitivity distribution and a second spectral sensitivity distribution that characterize the second image signal. Obtaining a spectral characteristic information obtaining procedure for obtaining a spectral intensity distribution of the illumination light; and deriving a conversion coefficient for performing color conversion of the first image signal from the first spectral sensitivity distribution and the second spectral sensitivity distribution. Alternatively, a deriving step of deriving a conversion coefficient from the first spectral sensitivity distribution, the spectral intensity distribution of the first illumination light, the second spectral sensitivity distribution, and the spectral intensity distribution of the second illumination light; Using the conversion coefficient Causing the computer to execute a color conversion procedure of color-converting the first image signal to obtain a second image signal, thereby color-converting the first image signal into the second image signal. A computer-readable recording medium that records a characteristic program. 12. A color which converts a first image signal characterized by a first spectral sensitivity distribution into a second image signal characterized by a second spectral sensitivity distribution by using a conversion coefficient, and outputs the second image signal. An image signal color conversion device, wherein the first spectral sensitivity distribution and the second spectral sensitivity distribution are obtained, or in addition to the spectral sensitivity distributions, the first spectral sensitivity distribution and the second spectral sensitivity distribution are acquired. Spectral characteristic information for acquiring a spectral intensity distribution of first illumination light further characterizing one image signal and a spectral intensity distribution of second illumination light further characterizing the second image signal together with the second spectral sensitivity distribution. An acquisition unit, which derives a conversion coefficient for converting the first image signal into the second image signal from the first spectral sensitivity distribution and the second spectral sensitivity distribution, or Sensitivity distribution, said Spectral intensity distribution of one of the illumination light, the second
A conversion coefficient deriving unit for deriving a conversion coefficient from the spectral sensitivity distribution of the second illumination light and the spectral intensity distribution of the second illumination light;
A color conversion processing unit for performing color conversion from the image signal to the second image signal.