JP2984276B2 - Color image processing method - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a color image processing method for creating a conversion table used in color processing.

[Conventional technology]

In color image processing, the circuit configuration becomes complicated and large-scale in order to perform a number of corrections, and labor is required to determine or adjust a correction coefficient in each correction circuit. Therefore, the device is expensive and large, and it is difficult to integrate the device for personal use.

To solve this, a method of performing color conversion using a table is conceivable. For example, three signals of R, G, and B are collectively used as input addresses, and outputs of Y, M, and C are obtained. Calculates the theoretical formula including various characteristics such as the spectral characteristics of illumination light, the spectral transmittance of filters, the spectral sensitivity characteristics of CCDs, and the characteristics of ink, and processes the data using a computer.
Table contents can be determined.

[Problems to be solved by the invention]

However, the above-mentioned theoretical formula does not always agree with the actual value, and it is necessary to adjust the color conversion coefficient several times while changing it slightly, and there is a problem that it is difficult to obtain correct color reproducibility.

Also, as a method of realizing correct color reproducibility, all possible dimensional functions are arranged, their coefficients are registered as variables, and the conversion results Y, M, C obtained according to the input of the read signals R, G, B In practice, there is a method of determining a coefficient by repeating the operation of actually printing according to the above, reading this again, and monitoring the R, G, B signals. However, this method requires a complicated program for performing feedback, and has a problem that it takes a long time since this operation must be performed until all the colors are satisfied.

SUMMARY OF THE INVENTION It is an object of the present invention to solve the above problems and to provide a color image processing method for easily creating a conversion table having good color reproducibility.

[Means and actions for solving the problem]

In order to achieve the above object, the present invention is characterized by having the following configuration.

The present invention is a color image processing method for creating a multidimensional table, wherein two color charts having different color component X values after color conversion are input by image input means,
Based on the data of the color chart, A component data a and B component data b are obtained from the color components before the color conversion.
From the distribution of (a, b) in the space, an approximate expression f (a) and an approximate expression g (a) in which the color component X takes two values are obtained, and the approximate expressions f (a) and g (a) are obtained. A table for the color component X is created based on the color component X.

〔Example〕

An embodiment of a color conversion table creation method according to the present invention will be described below with reference to the drawings.

FIG. 1 is an electric block diagram of a color scanner as an application example of the present invention. In FIG. 1, reference numeral 1 denotes a CCD with R, G, and B filters. The details of the CCD are shown in FIG. 2,3,4 are alternately attached. In FIG. 1, a serial analog signal transmitted from a CCD 1 is amplified by an amplifier 5 and input to an A / D converter.
Reference numeral 7 denotes a reference voltage control circuit, which inputs the reference voltage set for each of R, G, and B to the A / D converter 6 in synchronization with the transmission pulse of the pulse generator 8. In general, the spectral characteristics of illumination light and the spectral characteristics of R, G, B filters and C
Due to the spectral sensitivity characteristics of the CD, when reading a white background
Since there is a difference between the outputs of R, G, and B, R,
A reference voltage is set for each of G and B.

The A / D converter 6 has a 5-bit output, that is, 3 to 0 to 31.
It is a 2-tone output, and the data when reading a white background is
The gain of the amplifier 5, the reference voltage setting value, and the like are determined so that R, G, and B are all output at 1. 9,10,11 is A
R, G, B 5bit each transmitted serially from / D converter 6
Each of the three primary colors of R, G, and B signals are sampled by the latch 12 and input to the ROM 13. The ROM 13 is a color conversion table in which data is written so as to output a YMC 4-bit signal in accordance with the input R, G, B 5-bit signals.

The YMC signal output from the ROM 13 is
And printed by the printer 15 on the paper. The printer 15 is a thermal transfer printer and has an expression capability of 16 gradations by area gradation.

Next, a method of creating a color conversion table of the ROM 13 which is an image processing unit will be described in the order of cyan, magenta, and yellow.

First, a standard color chart for determining cyan data is created by the printer 15. The standard color charts are two types of color charts having different cyan component values as shown in FIGS.

The color chart shown in FIG. 3 does not contain cyan, that is, C = 0, and shows the density of yellow in the horizontal axis direction and the density of magenta in the vertical axis direction. In the present embodiment, the output is 4 bits, so that for both yellow and magenta, 16 gradations from 0 to 15 are realized from a density of 0 to a solid color.

The color chart shown in FIG. 4 is a case where cyan is solidly applied, that is, C = 15, and the way of obtaining the density of yellow and the density of magenta is the same as that of FIG.

Here, two types of standard color charts having different cyan component values are required. In this embodiment, in order to easily obtain two straight lines representing the maximum and minimum values of cyan, C = 0 and C are used.
= 15, but is not necessarily limited to these. For example, C = 1 and C = 14.

Further, since it is sufficient to obtain a straight line indicating the maximum value and the minimum value for the cyan data, charts for all cyan gradations are not required, and at least two charts are sufficient.

In this embodiment, the printer 15, that is, the color chart is created by the output means of the apparatus in which the color conversion table which is the object of the present invention is actually used, but the color chart is created by using the same ink. A commercially available color chart can also be used.

Next, a measurement system in which the output of the A / D converter 6 of the scanner is sampled by a logic analyzer or the like is constructed, and the color chart is read by the scanner, and the R, G, and R at respective yellow and magenta densities at that time are read.
B Collect data of 5bit, 32 gradations each. Then, among the obtained signals, the graph is plotted with R on the horizontal axis and G on the vertical axis according to the R and G signals. Similarly, plotting is performed in accordance with the signal obtained by reading FIG. 4 on the same graph. The graph thus obtained is shown in FIG.

Next, a straight line r ₁ = f ₁ (G) approximating C data = 0,
A straight line r ₂ = f ₂ (G) approximating C data = 15 is obtained from the above plot. Then, the equations of the approximate straight lines r ₁ and r ₂ are obtained. In this embodiment, the _{r 1 = 0.48G + 3 r 2} = 0.12G + 24.6. This is substituted into the following equation.

C = (R−r ₁ ) / (r ₂ −r ₁ ) × 15 Here, the formula divides C = 0 represented by r ₁ and C = 15 represented by r ₂ into 15 equal parts. , R and G to determine which of the 16 gray levels 0 to 15 the cyan output C corresponds to. However, C is an integer, and data after the decimal point is unnecessary. Further, when C is 0 or less, all are 0, and when C is 15 or more, all are 15.

In this way, read the color chart with the scanner,
By arranging the data on a graph and obtaining an equation of an approximate straight line, a calculation formula for determining a conversion table can be obtained.

The above formula includes all of masking, logarithmic conversion, γ correction, ink color correction, and the like necessary for color image processing. In determining C data, in addition to R data, which is a complementary color thereof, and G data The correlation is taken. Especially B
It has been experimentally confirmed that there is no need to correlate with the data, and this is supported by the fact that yellow ink contains little cyan in view of the ink components.

The method of determining cyan data has been described above.
In the case of magenta, a standard color chart is created and read as in the case of cyan, and the equation for determining the magenta component is: M = (G-g ₁ ) / (g ₂ -g ₁ ) × 15 g ₁ = f ₃ (R) g ₂ = f ₄ (R)

It is basically the same for yellow, but it has been empirically found that in the case of yellow, a correct color tone cannot be obtained unless it is represented by the three functions of R, G, and B. Perform various data processing.

First, data obtained when a color chart containing no yellow is read is grouped by R size (FIG. 6).
For example, when R = 4 to 8 with R = 6 as the center,
= 28, R = 26-30
Each is plotted on a graph with B on the horizontal axis and G on the vertical axis. That is, FIG. 6 shows a case where the white circle plot is R = 6 and a case where the half white circle plot is R = 28.

Then, the yellow data Y = 0 approximation of two straight lines that represent the formula _{b 1 (R = 6) =} f 5 (G) = 0.74G + 3.8 ... b 1 (R = 28) = f 6 (G) = 1.08G- 6.7…

Next, when the value of R changes from 6 to 28, a straight line b ₁ (R
= 5) moves to a straight line b ₁ (R = 25), and an equation is established. That is, b ₁ (R) = aG + d, and the parameters a and d are considered to be functions of R. This is replaced by b ₁ (R, G) = (eR + h) G + (jR + k)...
To obtain b ₁ (R, G) = (0.015R + 0.65) G−0.48R + 6.7. Then, similarly to the case of cyan, the output gradation Y of yellow is as follows: Y = (B−b ₁ ) / (b ₂ −b ₁ ) × 15 b ₁ = (0.015R + 0.65) G−0.48R + 6.7 b ₂ = 0.2G + 23 If the conversion formula from R, G, B data to Y, M, C data is determined as described above, it is easy to create ROM data by a personal computer or the like based on this. .

As described above, according to the present embodiment, the color chart is read, and the points at which the output gradation becomes 0 and the points at which the gradation becomes 15 are plotted in accordance with the R, G, B signals at that time. ,
An equation for linearly approximating this is obtained, and a conversion table can be created by an arithmetic equation that divides the two straight lines into 15 equal parts. The good color reproducibility of the conversion ROM created by this method has been confirmed on actual equipment.

In this embodiment, the three color input signals of R, G and B are used.
An example of a table for performing color conversion into output signals of three colors of Y, M, and C has been described as an example. The method can be applied.

Further, in the present embodiment, the present invention is used for creating a color conversion ROM, but the present invention is not limited to ROM and can be used for creating a RAM or a digital arithmetic circuit.

Further, the present invention can be applied to both a manual scanning type and an automatic scanning type scanner.

Further, the input gradation and the output gradation are not limited to the case of the present embodiment, and the number of gradations can be freely set.

Further, the printer as the output means is not limited to the thermal transfer printer, but can be used for all output devices such as an ink jet printer and a laser beam printer.

Further, when an approximate expression is obtained from a plot in a coordinate space having two color components as variables before color conversion, it is not always necessary to perform a linear approximation as in the present embodiment, and an approximate expression such as a parabola or a hyperbola is used. You may.

In this embodiment, a color chart having the minimum color component after color conversion, for example, C = 0 (FIG. 3) and a maximum color chart, for example, C = 15 (FIG. 4), are input, and the approximate values are obtained from the read data. Although the equation has been obtained, the value is not necessarily the minimum or maximum, but may be any value close to this.

Also, the two straight lines for approximation are always equally divided,
It is not necessary to assign to the output gradation. For example, it is possible to appropriately determine, for example, that the low gradation corresponds to a wide range and the high gradation corresponds to a narrow range. In other words, this means that the space between the two straight lines is divided at a predetermined ratio to correspond to the output gradation of the color component for which the table is created.

〔The invention's effect〕

According to the present invention, a multidimensional table with good color reproducibility can be easily created.

[Brief description of the drawings]

FIG. 1 is an electric block diagram of a scanner device to which the present invention is applied, FIG. 2 is a diagram for explaining the configuration of the CCD 1, FIG. 3 is a color chart for cyan 0, and FIG. FIG. 5 is a graph of cyan data for creating a conversion table, and FIG. 6 is a graph of yellow data for creating a conversion table. 1 ... CCD 5 ... AMP 6 ... A / D converter 7 ... Reference voltage control circuit 8 ... Pulse generator 13 ... ROM 14 ... Host 15 ... Printer

Claims (1)

(57) [Claims] 1. A color image processing method for creating a multidimensional table, comprising: inputting two color charts having different color component X values after color conversion by image input means; A component data a and B component data b among the color components before color conversion are obtained based on the distribution of (a, b) in the AB space.
Takes an approximate expression f (a) and an approximate expression g (a), which take two values. Based on the approximate expressions f (a) and g (a), the color component X
A color image processing method, wherein a table relating to the color image is created.