JPH0197652A - Method of color printing - Google Patents

Abstract

PURPOSE:To realize high-fidelity reproduction of original color images by a method wherein optimum area factors of the paper color and the synthesized colors derived from ink primaries are determined for three stimulus values to assign serial numbers to the regions corresponding to the optimum area factors, and the boundary values of these regions are compared with the corresponding values in a judgement matrix to position the colors in an output matrix. CONSTITUTION:The area factors of the colors based on an RGB signal are determined by a process using a table 10 for generating conditions. The values corresponding to these colors are produced in accordance with a coding signal for print primaries. A matrix circuit 11 (e.g., 4X4) reads matrix values in a predetermined sequence and delivers them to a comparator 12. The comparator 12 compares the value from the table 10 with the value from matrix circuit 11. When the comparation agrees with given conditions, the color is printed.

Description

DETAILED DESCRIPTION OF THE INVENTION "Field of Industrial Application" This invention relates to a printing method for a color printer capable of reproducing colors with high fidelity.

``Prior Art'' Screens displayed on color displays are often printed out using a printer. In this case, the printer converts the RGB (red, green, blue) signals for display supplied from the computer etc. into YMC (yellow, magenta, cyan) signals for printing, and then prints each color based on the converted signals. Print and overlap.

FIG. 3 is a block diagram showing a schematic configuration of an in-film color printer.
This is a signal conversion unit that converts the signal into a signal.

The processing in this signal converter l can be easily explained, for example. The conversion process is performed using the following calculation.

Reference numeral 2 denotes a pseudo gradation processing section, and when processing in this section is performed using, for example, a 4.times.4 threshold matrix x (BAYER type) using the density pattern method, the result will be as shown in FIG. In other words, if the original pixel data (any one color of YMC) is "7" as shown in Figure 4 (A), the sample value is set in the 4x4 matrix as shown in Figure 4 (B). Create sample values in which all values are "7".

And a pre-stored threshold matrix (
The matrix shown in (c) of the same figure is compared with the sample value, and the matrix shown in (d) of the same figure is obtained as a comparison result. The hatched areas in this matrix are areas where the sample value is larger, and are printed island areas. That is, 16 dots (4x4) are printed for each dot of the original pixel, and (4x4)
), the number of dots printed corresponds to the gradation of the original pixel.

Then, based on the above processing results, the YMC printing section 3 (third
Printing is performed in (Fig.). This printing is performed for each color of YMC, and overlapping printing is performed.

"Problems to be Solved by the Invention" By the way, when printing is performed using the method described above, as a result of pseudo gradation processing, the primary colors of YMC are overlapped and subtractively mixed, and the YMC A portion where each of the primary colors is juxtaposed and mixed occurs. That is, subtractive color mixture and juxtaposed color mixture coexist within the matrix that reproduces the original pixel. For this reason, the color of the original image and the printed color may differ, resulting in a problem that the fidelity of color reproduction is reduced.

The present invention has been made in view of the above-mentioned circumstances, and an object of the present invention is to provide a printing method for a color printer that can reproduce the colors of an original image with high fidelity.

``Means for Solving the Problems'' In order to solve the above problems, the present invention provides a color printer that prints one dot of an input image in correspondence with a plurality of dots arranged in a matrix. The composite color obtained by superimposing the ink primary colors and the optimal area ratio of the paper color are determined corresponding to the given tristimulus values, and the consecutive numbers are set within a range corresponding to each of the optimal area ratios. Then, the position of each color in the output matrix is determined by dividing each color according to the preset printing order and by comparing the six values in the matrix provided for discrimination with the boundary value of the range. It is characterized by deciding.

"Operation" The optimal area ratio of the color to be printed is determined according to the three stimuli vi (RGB signals, etc.) supplied as input data,
Consecutive numbers are assigned according to a predetermined printing order within a range corresponding to this area ratio. Further, the six values of the discrimination matrix are compared with the boundary value of the range, and the arrangement of each color in the output matrix is determined according to the comparison result.

"Embodiments" Hereinafter, embodiments of the present invention will be described with reference to the drawings.

First, assuming that printing is performed using the three primary colors of YMC (black ink will not be used for simplicity), printing in the same color as the original image means that the data supplied as input data will be printed in the same color as the original image. For each color value of RGB, it boils down to finding each area ratio of YMC so as to correspond to the color of the original image.

Furthermore, Y and M inks can be combined with each YMC ink.

If the area ratios of C, R, G, B, K (black) and the paper W are determined individually, the reproducibility of the color produced by a set of these can be made higher.

Hereinafter, based on the above principle, Y, M, C, R, G.

In B, the process of determining the area ratio of W will be explained.

First, tristimulus values corresponding to each color of Y, M, C, R, G, B, and W, that is, six values of R, G, and B signals are measured. Then, R and G were measured.

The six values of B are expressed as a vector. Therefore, for example, the tristimulus values corresponding to R or Y are each expressed as follows.

Also, if the area ratio of each color of W is expressed as r, g, b, y, m, c, and w in l, G, B, Y, M, -C, respectively, then
The following formula holds.

However, r+ g+ b+ y+ m+ c+ k+
w= 1...(2) If the area ratio is determined from the above formula, it can be seen that each of RGB is uniquely determined, but conversely, when calculating the area ratio when RGB is given, cannot uniquely find a solution because there are three equations for eight unknowns. Therefore, as an example, the area ratio is determined by the following method. Note that in this example, an approximate solution is first obtained, and then an optimal solution is obtained using this as an initial value.

Now, the maximum values of each rLGB signal are RO and GO.

In the case of BO, white is where RGB overlaps, so its area/ratio W is w=min'' [R/RO1G/60, B/BO]...
...(3) becomes. Note that min means the minimum value among the numerical values in [].

Also, since black is the area where there is no RGH signal, the area ratio is k=m1n[1-It/RO11-G/GO11-B/
BO]...(4) On the other hand, the following equation approximately holds true for each of RGB.

r+m+y+w=R/RO (color including red component)g+y+
c+w=G/'GO (color containing green component) b+m+c+w
=B/BO (color containing blue component)... (5) c+g+b+kl-R/RO (color containing cyan component) m
+r+b+=l-G/Go (color containing ?zenta component) y+r+g+kl-B/BO (color containing yellow component) ......(6) Then, from equations (3) to (6) above, r , g, y.

It can be seen that four variables among m and c are always 0.

For example, in equations (3) and (4), w = B /
Assuming that B 01k=l-R/RO, by substituting these relationships into the third equation in equation (5) and the first equation in equation (6), b = m = c = g = It turns out that it becomes 0. By using this relationship, (5
) and (6), y=G/GO-B/B0r=R/RO
It becomes G/GO. In this way, by using equations (3) to (6), the area ratios r, g, b, y, and m of each color can be determined.

c, we can find w.

By the way, the possible value of each area ratio is determined by the number of gradations that can be realized, and when the number of gradations is N, the minimum value of the area ratio d
is d=1/(N-1) (7).

Then, each area ratio is rounded by an integral multiple of the minimum value d shown by formula (7), and this is substituted into formula (1) as an initial value (approximate solution), and the resulting R, G, B Let each position be the initial value R", G-, B- of each color signal. Here, ΔR=11-R- ΔG=G-G" ΔB=I3-B-, that is, the difference between the initial value and the true value. Define a value indicating the error with rtGB. Ideally, the errors ΔR9ΔG and ΔB should be 0, but there is a limit because the minimum value d of the area ratio is finite. Therefore, by correcting the area ratio obtained at the beginning, R-, G-.

B' is corrected to find the area ratio that gives the minimum error.

For this process, calculate the following values.

ΔRi j=d (Ri−Rj) ΔGi j=d (Gi−Gj) ΔBi j=d (Bi l3j) However, j=R, G, B, Y, M, C, ni', W, and jf -i. Further, i is as described above.

Here, is defined as a correction vector, and is defined as an error vector. Then, a correction vector P that makes the error vector ■ smaller is searched for, and the area ratio hli is corrected so that it matches the searched correction vector P.

That is, first, on the condition that the area ratio has a positive value of 6, a combination of i and j is found such that the error vector II after correction is smaller than before correction, and this discovered correction Vector P is added to (R', G'', B') instead of one error vector H as shown in the following equation, and the value of (1''t-, G-, 13') is updated by this calculation process.

R-=R'+ΔRij C;'=G-+ΔGij B'=B-+ΔBij This updating process is repeated and ends when there is no correction vector P that makes the error vector ■ any smaller. Simultaneously with the above update process, the following calculations are performed to correct the area ratio. In addition, in this formula, the area ratio is S k (k = R, G, B, Y, M.

In C, W, -9R=r, Sσ=g, 5B=b.

S Y = y, S M = m, S C
= c, S K = k, and S W = W).

5k=Sk+dxlF However, (IF=l; to=+11F=-1; to=j.

IF=0;kf-i,j).

The meaning of the above formula is as follows. For example, i=R,j
If a combination of =G is found, the correction vector P is obtained, but the area ratio when k is i, that is, the area ratio 5R (=r
) is updated by performing the calculation S n = S R + d and adding the minimum value d. Further, the area ratio when k is j, that is, the value of the area ratio S6 (=g), is updated by performing the calculation 5c=Scd and subtracting the minimum value d. Area ratios other than these (SY, SC, etc.), i.e.
For those (kf-i, j) in which k does not match either i or j of the correction vector P, the previous value is maintained as 5k=Sk.

As a result of the above processing, when the error vector ■ becomes the minimum, the corresponding area ratio is simultaneously determined for each color.

Next, a method for quickly converging the above correction process will be described.

Assume now that the optimal correction vector is PP, and that the discovered correction vector is P. Then, set a λ value that satisfies the following equation.

PP-λP In this case, λ satisfies the following conditions.

(2) The area ratio determined by the correction vector P does not become negative after correction.

■λ satisfies the following formula.

! ≦λ<2(IT・P)/IP+” However, (■・P) is the internal debt, and IPI is the length of the vector.

■λ is a positive integer close to the following value.

λ=(n・P)/! PM The conditions for each of the above equations are derived from the following.

First, if the angle between the vector 11 and P is θ, then the following equation holds true. Furthermore, as shown in Figure 2, if we draw a sphere whose radius is the length of the error vector H with the tip of the desired color vector as the center, the corrected approximate vector must fall inside that sphere. . Therefore, the length of the correction vector H must satisfy the following equation.

1λP l <2 I n I CO3θ From this equation, the following equation is derived: λlpM<2(TI ・P). Here, if the length of the error vector H after correction is L, then L'=(If-λP)! So, if we differentiate this and set it as O, then λ=(n ・ P)/lp1! becomes.

The above is the method for determining the area ratio in one embodiment of the present invention.

Next, how to arrange colors will be explained. First, the area ratios r and g obtained by the process described above.

Rewrite w as b, y, m, c, using the minimum area ratio d as a unit, and rewrite this as r, g, b, y.

Rewrite it as m, c, and w. In this case, the following formula holds from the definition of area ratio.

r+g+b+y+m+c++w=N-1 Next, arrange each color in the order of, for example, KMRYGBW and number the area. As a result, the correspondence between colors and numbers will be as shown in the table below.

(Hereinafter, blank space) According to the table above, the numbers to be printed for each color on the YMC ribbon are: ■ For Y (yellow ribbon), ■ to 1 and to + m + 1 to k 10 m + r + y + g For ■M (magenta ribbon), the range is 10 m from ■.
+r and k+m+r+y+g+c+1 to +m
For C (cyan ribbon), the range is from 1 to k and from +m+r+y+1 to +m+r+y''+g+c+b.When k (black ribbon) is used, the range from indigo to k is overlapping. Just print it out.

From the above, k% +m+l, k+m+r.

k+m+r+y+1. +m+r+y+g%+m+
r ten y + g + c ten plates, k + m + r + y +
By knowing the seven values g + c and + b, the printing position of each color on the ink ribbon can be determined.

Next, corresponding to each of the above, K y 1 = k s
Ky2=+m%Ky 3=+m+ r+y+g, l
Kml=+m+r, Km2=+m+r+y+g+c
, Km3=Kc3, Kc1=k, K
c 2 = k + m + r + y, Kc3 = + m +
Define r+y+g+c+b. And N-1=
Consider an M*M matrix and decide how to arrange the colors within it. As already mentioned, each color is numbered in the order KMRYGCBW, so filling in the numbers in the matrix corresponds to arranging the colors. In this case, various matrices can be used, but in principle they cannot be applied to the dither method. As a matrix, a centralized type is suitable, so
- As an example, an 8*8 fattening type is applied, which is not shown below.

In the above matrix, k (black) is placed in the center, surrounded in the order of MRYGCB, and finally surrounded by W. If the 6 values of this matrix are set as Tij, each YMC ink can be printed based on the following judgment.

Y--...-T i j≦Ky 15Ky 2<T i
j<Ky 3M...T i j≦Kml, Km2
<'r i j≦K m 3C...Tij≦K
c1. Kc 2<T i j≦K m 3 Also, K(
If you want to print overlapping colors (black), you can use the following formula to judge.

K...Tij≦Kkl However, Kkl is k or αk (α≦1).

To summarize the above, RG supplied from a computer etc.
The area ratio of each color is calculated from the B signal, and from this result KylS
The arrangement of colors to be printed is determined by calculating values such as Ky2 and comparing these values with numerical values in a predetermined matrix.

In this case, 3/l corresponds to various values of RGB signals.
, KV2... etc., if you store them in ROM,
Valley values can be generated quickly, and the arrangement of colors to be printed can be easily determined using a comparison circuit or the like.

Next, FIG. 1 is a diagram showing a schematic configuration of a printer to which the present invention is applied. In this figure, IO is a condition occurrence table, in which the area ratio of each color is determined based on the r(GB times signal) using the processing method described above, and KyI corresponding to the color is determined based on the printing primary color code designation signal. ni3/
Outputs numerical values such as 2... 11 is a matrix circuit (4×4 in this example), and the six values in the matrix are read out in a predetermined order and supplied to the comparison circuit 12. In the comparison circuit 12, the numerical values such as Kyl, Ky2, etc. supplied from the condition generation table IO,
Compare the numerical value read from matrix circuit II,
When this satisfies a predetermined condition, printing is performed in that color.

"Effects of the Invention" As explained above, according to the present invention, one of the input images
In a color printer that prints dots in a matrix of multiple dots, primary colors of ink,
The composite color obtained by overlapping these primary ink colors,
and the optimal area ratio of each color of the paper are determined corresponding to the given tristimulus values, and consecutive numbers are assigned within the range corresponding to each optimal area ratio and according to the printing order set in advance for each color. In addition, the position of each color in the output matrix is determined by comparing the six values in the matrix provided for discrimination with the boundary value of the range, so the colors of the original image can be reproduced with high fidelity. This provides the advantage of being able to be reproduced at a high degree.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing a schematic configuration of a printer to which an embodiment of the present invention is applied, FIG. 2 is a diagram for explaining the invention in detail, and FIG. 3 is a block diagram showing the configuration of a conventional printer. , FIG. 4 is a diagram showing the printing principle in a conventional printer. IO... Condition occurrence table, 11...
Matrix circuit, 12... Comparison circuit.

Claims (1)

[Claims] In a color printer that prints one dot of an input image in correspondence with multiple dots in a matrix,
The optimum area ratios of the primary ink colors, the composite color obtained by superimposing the primary ink colors, and the color of the paper are determined corresponding to the given tristimulus values, and the consecutive numbers are determined according to each of the optimal area ratios. The position of each color in the output matrix is calculated by comparing each value in the matrix provided for discrimination with the boundary value of the range. A printing method of a color printer characterized by determining the following.