JPH0440158A - Method for calculating color correction - Google Patents

Abstract

PURPOSE:To always execute almost optimum color correction by calculating a color correction coefficient from a picture element data from a digital data in respect to a color original in plural total spectroscopic sensitivity characteristics concerning one separated color. CONSTITUTION:First of all, a color original to an (n) number of various colors is read in the initial state of a color picture reader so as to obtain digital color picture element data (R1, G1a, B1)-(Rn, Gna, Bn) in respect to the (n) kinds of colors. Next, when the total spectroscopic sensitivity characteristic of the color picture reader is changed, the color original is read in respect to the same (n) number of various colors so as to obtain digital color picture element data (R1, G1b, B1)-(Rn, GnbBn) in respect to the (n) colors. For example, concerning R, the two pairs of (R1, G1a, B1) and (R1, G1b, B1) are made correspondent to R01 in respect to the 2n pairs of data as mentioned above, and even concerning the other pairs, the data are similarly made correspondent. Then, the condition of miniaturizing the residual sum of squares is calculated.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a method for calculating color correction coefficients, and in particular, a method for reading a color original to generate digital color pixel data, and performing color correction on the digital color pixel data. The present invention relates to a method for calculating color correction coefficients used in a color image input device having a function to perform color correction.

[Conventional technology]

2. Description of the Related Art Conventionally, color image reading devices are known that read color originals and generate digital color pixel data.

FIG. 2 is a block diagram showing an example of the structure of a conventional color image reading device. 7 is a color original, 1 is a color image reading means, 2 is a digital processing means, 3 is a color correction means, 4 is a color correction coefficient storage means, and 5 is a color image display device.

FIG. 3 is a side view showing an example of the configuration of the reading means 1 of the color image reading device. A color document 7 on a document table 6 made of glass or the like is illuminated by an illumination light source 8 , and the reflected light passes through a lens 9 and forms an image on a color image sensor 10 .

A color filter is provided on each reading pixel of the color image sensor 10, and a color-separated image is input to each reading pixel, and a signal corresponding to the intensity of light is output. .

The color image sensor with the above configuration normally performs color separation for three color separations.
Green) and B (blue) are often used, and in that case, the color separation system is often composed of color filters of the above-mentioned separated colors, but it can also be composed of color filters of complementary colors. Based on the spectral characteristics of the color filter, illumination light source, image sensor, and lens, the color image reading means as a whole has spectral sensitivity characteristics, for example, as shown by the solid line in FIG. 1 with respect to wavelengths in the visible region. The spectral sensitivity characteristics of the color image reading means as a whole will be referred to herein as the overall spectral sensitivity characteristics. When a color original is read by such a system, a color image signal is output based on the relationship between the spectral reflectance characteristics of the color original and the overall spectral sensitivity characteristics of the color image reading means.

The color image sensor is provided to have a reading width for a color document, and is scanned in the direction of the arrow by a scanning mechanism (not shown) in FIG.
The entire color document or a portion of it is scanned.

In FIG. 2, the color image signal generated by reading the color original 7 by the color original reading means 1 is as follows.
The digital processing means 2, which includes an A/D converter, shading correction means, etc., converts the data into digital color pixel data.

If you try to display the above color pixel data as is on a CRT or output it to a color printer, you will not be able to obtain good color reproduction for the original due to distortions in the spectral sensitivity characteristics of the color image reading means 1, differences in color system, etc. , Therefore, color correction processing is usually performed on digital color pixel data. In order to simplify the explanation below, the color image is output using an additive color color image display bag such as a CRT!
The following explanation assumes that .

As a means of color correction, there is a color correction matrix calculation method in which a color correction matrix is prepared and color correction is performed by multiplying the digital color pixel data by the color correction matrix, and a color correction matrix calculation method that performs color correction by multiplying the digital color pixel data by the color correction matrix. A look-up table reference method is known in which relationships are prepared in a look-up table and color correction is performed by referring to the look-up table for digital color pixel data. Here, the former color correction matrix calculation method is used as the color correction method.

In FIG. 2, the color correction calculation means 3 performs a color correction calculation on digital color pixel data using the color correction coefficients stored in the color correction coefficient storage means 4. The color correction calculation is mainly for correcting distortion in the spectral sensitivity characteristics of the color image reading means 1. Assuming that the digital color pixel data is three-color data of R (red), G (green), and B (blue), for example, when using the simplest 3×3 matrix, color correction processing is performed using the following equation.

Here, R, G, and B are digital color pixel data obtained by the digital processing means 2, R'G'B' are digital color pixel data after color correction obtained by the color correction calculation means 3, and A-( atj) is a color correction matrix used by the color correction calculation means 3. Note that although the simplest 3×3 matrix has been described here, a higher-order matrix may also be used.

The digital color pixel data subjected to color correction processing by the color correction calculation means 3 is sent to a color image display device 5 such as a CRT.
will be displayed. Note that the digital color image data subjected to color correction processing by the color correction calculating means 3 may be subjected to T correction processing for correcting non-linearity between input and output of a normal color image output device.

Next, we will explain the color correction matrix of formula 0 (1) to explain an example of the method of awarding the color correction matrix.
For example, a color image reading device reads a color document for many colors, and a digital processing device obtains digital color pixel data. From the set of digital data (R, G, B) for each color of the obtained color document and the set of color reproduction target values (R,,G,,B,) for each color chart, (R.

Minimum 2 so that G, B) approaches (R,,G,,B,'')
Calculate using multiplication etc.

Here, an example of a method for calculating a color correction coefficient using the minimum 2 multiplication method for color correction in equation (1) will be briefly described.For example, a case where R data is obtained by color correction will be described. First, the correction target value R0 and the data R, G, and B obtained by reading a color original are related by the following equation.

Ro-α1R1αAG+α,B...(2)
Next, for a color original for n different colors, a correction target value R0 and data R, G, and H obtained by reading the color original are obtained. Here, the data on R0 is usually obtained by spectroscopically measuring a color original. R, G, and B data can be obtained by directly reading a color original with a color image input device without performing color correction processing, or from the overall spectral sensitivity characteristics of the color image reading means and the spectral reflectance of the color original. It can also be obtained by calculation. Now, let the estimated value of α be ai
Then, the following formula is obtained.

Here, "l+" for the condition that minimizes the sum of squares of the residuals Rt==Σ(R-tRi)! ＋” 3
seek.

Similarly, perform the same total of 1 for G and B, and finally 9
The color correction coefficients aijN,j''1 to 3) are determined.The n color originals to be used at this time include, for example, many color originals in which the colors of the original are distributed as evenly as possible within the uniform color space. is desirable.

A color correction coefficient is calculated by the color correction coefficient calculation method as explained above, and the color correction coefficient is used to perform color correction processing on the digital color pixel data obtained by the color image reading means. was.

In the color correction method described above, color correction coefficients optimized for the overall spectral sensitivity characteristics of the color image reading means are calculated in advance, and color correction is performed using the color correction coefficients of the color correction matrix. I was walking.

[Problem to be solved by the invention]

However, the overall spectral sensitivity characteristics of a color image reading means may change over time over a long period of time. For example, in a color image reading means having a color separation system using color filters, Depending on the usage environment, the spectral transmittance characteristics of the color filter change after a long period of time, and a portion of the overall spectral sensitivity characteristics of the color image reading means changes, for example, from Ga shown by the solid line in FIG. 1 to Gb shown by the broken line in FIG. There was a tendency to shift to

The color correction coefficients calculated for the initial overall spectral sensitivity characteristics of the color image reading means are initially optimal and provide sufficient color correction, but due to changes in the spectral transmittance characteristics of the color filters as described above. After the overall spectral sensitivity characteristic of the color image reading means is shifted, the initial color correction coefficient is no longer optimal. If color correction processing is performed in such a state using color correction coefficients for the initial state, there is a problem in that good color reproduction cannot be obtained because the color correction coefficients are not optimal.

[Means to solve the problem]

In order to solve these problems, the present invention provides a color correction method used in a color image input device that has a function of reading a color original, generating digital color pixel data, and performing color correction processing on the digital color pixel data. In the method of calculating the coefficient, if the total spectral sensitivity characteristics of the plurality of color separations possessed by the color image input device are small (both for one color image input device), The color correction coefficient is calculated based on the data.

[Effect]

If performed as described above, even if the color separation characteristics of the color separation system change over time, close to optimal color correction can always be performed on a color image.

〔Example〕

An embodiment of the present invention will be described below based on the drawings.

Here, the overall decomposition sensitivity characteristics of R and B remain unchanged;
A case in which only the overall spectral sensitivity of G is shifted will be explained as an example. In FIG. 1, the curve vAGa is the overall spectral sensitivity characteristic curve of G at the initial stage of the color image reading means, and the curve cb is the overall spectral sensitivity characteristic curve of G after shifting. First, a color original for n different colors is read in the initial state of the color image reading device, and digital color pixel data (Rr, Or-, B
+) ~ (R,,G,,,B,) The overall spectral sensitivity characteristic G of the zero-order color image reading device is the curve G in Figure 1.
In state b, a color original for the same n different colors is read, and the digital color pixel data for the n colors (
RI+ G+b, B 1)～(R,,G,,,B
, ), for example, for R, as shown in equation (4), (Rl+ G ja+
The two sets B+) and (R,, G□b, B+) are made to correspond to R□, and the other sets are made to correspond in the same way.

And the sum of squares of the residuals Rt-Σ(R,i-R5)! ``l+'' for the conditions that minimize ).

...(4) In the above explanation, the explanation has been made assuming that the overall spectral sensitivity characteristics for one color separation shift, but the same procedure can be applied to the case where the overall spectral sensitivity characteristics for two or more separation colors shift. It is possible. In addition, each digital color pixel data may be weighted if necessary. Furthermore, the case where the color correction coefficient is calculated using two types of comprehensive spectral sensitivity characteristics for one color separation has been explained. ,3
A total spectral sensitivity characteristic of more than one species may be used.

Note that in the above embodiment, when using digital color pixel data read by an actual color image reading device, it is necessary to take actions such as artificially changing the overall resolution sensitivity characteristics of the color image reading means. It is also possible to carry out experiments such as these to understand the trends in changes in the overall spectral sensitivity characteristics, and then use R, G, and B data obtained by measurement from the spectral reflectance characteristics of color originals for n colors. It is.

〔Effect of the invention〕

As described above, according to the present invention, in color correction processing for color pixel data generated by reading a color original with a color image reading device, changes over time occur in the overall spectral sensitivity characteristics of the color image reading means. Now, it is now possible to always perform close to optimal color correction on the read color single pixel data.

[Brief explanation of drawings]

Fig. 1 is a schematic diagram showing an example of the general attitude characteristics of an image reading means, Fig. 2 is a pronk configuration diagram showing an example of a conventional image reading device, and Fig. 3 is a side view showing the configuration of a direct image reading means. be. Applicant: Seiko Electronic Industries Co., Ltd. Representative: Patent Attorney: Keinosuke Hayashi

Claims (1)

[Scope of Claim] A method for calculating a color correction coefficient used in a color image input device having a function of reading a color original, generating digital color pixel data, and performing color correction processing on the digital color pixel data, comprising: Calculate a color correction coefficient based on digital color image data for a color original in a plurality of comprehensive spectral sensitivity characteristics for one color separation, for at least one of the comprehensive spectral sensitivity characteristics of a plurality of color separations possessed by a color image input device. A color correction coefficient calculation method characterized by: