JPH02220566A - Masking method for color picture data - Google Patents

Abstract

PURPOSE:To obtain a natural copied picture by setting a masking equation for each of plural areas resulting from a color space divided by the saturation or the saturation and the brightness. CONSTITUTION:Primary color components B, G, R are inputted to an adder 1, a minimum value selection circuit 2 and a maximum value selection circuit 3. The adder 1 generates sub color components b, g, r being the subtraction of a brightness V outputted from the circuit 2 from each of the components B, G, R respectively. Moreover, an adder 4 subtracts the brightness V from the maximum value of the color component of the circuit 3 to generate a saturation CR. Mashing arithmetic circuits 5-7 apply the arithmetic operation corresponding to masking matrices ANR, AP and AHC to the sub color component to generate each combination of tentative secondary color components. Furthermore, the saturation CR and the brightness V are inputted to an adaptavility output circuits 8-10, where weight coefficients WNR, WP and WHC are generated. The tentative secondary component and the weight coefficient are inputted respectively to arithmetic circuits 11-13 and a processed color components y2, m2 and c2 are generated. They are added (14) to the brightness V to generate the secondary color component.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a method of masking color image data used in a color image-scanning recording apparatus such as a color scanner for plate making.

[Conventional technology]

As is well known, primary color components R, G, B (first image data) read from a color original are converted into secondary color components Y, M, C, K (second image data) corresponding to printing ink. )
As a method of masking color image data for conversion into
There are known methods that use a masking equation that takes into consideration cross-term components RG, GB, and SR, and methods that use a masking equation that performs nonlinear calculations that take cross-term components RG, GB, and SR into account.

In the methods described above, masking coefficients are set so as to obtain good reproduction characteristics over as wide a range as possible within the color space, but it is difficult to obtain optimal characteristics for the entire color space.

In addition, as a countermeasure for this, for example, Japanese Patent Laid-Open No. 60-311
As disclosed in Japanese Patent No. 43, a method has also been proposed in which a color space is divided into several regions and an optimal masking coefficient is set for each region.

(Problem to be solved by the invention) However, with the method that takes quadratic term components and cross term components into consideration, even if good reproduction characteristics are obtained in the main color space, good reproduction characteristics are obtained in other color spaces. Reproducibility characteristics cannot be obtained.

In addition, in the method of dividing the color space into multiple regions and setting masking coefficients for each region, the reproduction characteristics may become discontinuous at the boundaries between each region, resulting in an unnatural reproduced image. There is.

(Objective of the Invention) This invention has been made in consideration of the above circumstances, and it is possible to set an optimal masking coefficient for each divided color space, and obtain good reproduction and characteristics in the golden space. The object of the present invention is to obtain a color image data masking method that makes the reproduction characteristics continuous even at mutual boundaries between regions and provides a natural duplicate image.

[Means to solve the problem]

The first color image data masking method according to the present invention scans an original image and obtains a first image data for each pixel.
The method converts first image data representing a color component into second image data representing a second color component based on a predetermined masking equation. Divide into regions.

Next, a masking equation is prepared for each of the plurality of regions.

Then, by substituting the saturation of the first image data into a predetermined membership function indicating the degree of suitability for the saturation, the suitability of the first image data to each of the plurality of regions is determined.

Further, the first image data is converted into second image data based on the masking equation and the goodness of fit to each of the plurality of regions.

Further, the color image data masking method of the second configuration according to the present invention performs the same processing as the method of the first configuration for saturation and brightness.

In general, the method for masking color image data of the third configuration according to the present invention is based on the method of the first or second configuration, in which the first image data is prepared for each of a plurality of regions in the color space. Based on the masking equation, multiple temporary second
image data, calculates a weighted average of a plurality of provisional second image data based on the degree of conformity to each of the plurality of regions, and calculates the second image data based on the weighted average. be.

However, the membership function in this invention means a function that indicates a degree of conformity between perfect conformity and non-conformity at mutual boundaries between a plurality of regions.

(Operation) Since the masking equation in the present invention is set for each of a plurality of regions obtained by dividing the color space by saturation or saturation and lightness, reproduction characteristics are individually given to each region.

Furthermore, since the degree of suitability of the first image data to multiple regions is given by a membership function defined by saturation or saturation and brightness, the reproduction characteristics at mutual boundaries between multiple regions are continuous. .

〔Example〕

A. Overall configuration and general operation FIG. 2 is a schematic block diagram of a plate-making scanner to which an embodiment of the present invention is applied. In the same figure, original picture 100
The image is read pixel by pixel by the scanning reading device 200, and the RGB image signals (density signals) thus obtained are transferred to the image processing device W1300. The image processing device 300 includes a masking circuit 400 having functions to be described later.
It performs masking processing and the like on input RGB image signals. And YMCKM after processing
The image signal (density signal) is provided to a scanning recording device 500. The scanning recording device M500 is YMCKii! ii The image signal is converted into a halftone signal, and based on the image signal, the photosensitive film 60
Both halftone dot images are exposed and recorded on 0.

B. Maskin d Next, the -next color components B, G, and R are converted into the secondary color component Y.

The masking equation for converting into M, C, and K will be explained. Equations (1) and (2) below are determinants representing standard masking equations.

-V...(1)...(2) However, V=MIN (B, G, R)...(3^
) As shown in equation (3^), the minimum Ii among the color components B, G, and R! (2) is treated as lightness for convenience, as will be described later. According to equation (2) 8, the color component and the brightness V are equal.

Equation (3B) is a defining equation for the subcolor components i, g, and r, and as shown in equation (1), the masking coefficient a 1j (i, j=
12.3) is determined for the sub-color components g, r, taking into consideration the characteristics of the color ink.

Moreover, the following equation (4) is obtained by transforming the above-mentioned equation (1).

According to equation (5), I corresponding to the secondary color components Y, M, and C
WI color components y, m, c are defined. Equation (4) calculates the first subcolor component g, r to the second subcolor component y.

This is a masking equation showing the conversion to m and c.

Furthermore, a masking equation as shown in equation (6) below, which performs correction using a quadratic term component, is also used.

However, there are also other known masking equations that perform correction using cross-term components.

C9 Processing Procedure Next, the processing procedure of a color image data masking method according to an embodiment of the present invention will be explained with reference to the flowchart of FIG. Note that the terms "saturation J" and "lightness J" used in this invention are not limited to the narrow meaning defined in academic terms,
These are broad terms indicating sharpness and brightness, respectively, and in this example, they are variables having the definitions described below.

First, in step 811, the entire color space is divided into a plurality of regions, focusing on saturation and brightness.

FIG. 3 is a diagram showing an example of a color space divided by focusing on saturation CR and brightness ■. In addition, the brightness V is given by the above-mentioned formula (3^), and the saturation CR is given by the formula (7) shown below.
given by.

CR= MAX (B, G, R) - MIN (B, G, R) (7) As shown in equation (1), the saturation CR in this example is based on the color components B, G, It is defined by the maximum value of R minus their minimum value.

In addition, the saturation CR ranges from “0” to the maximum value CR1aX.
The brightness V takes values from “0” to the maximum value vlax, and the reference value CR1°CR (cR<C
R) Set the reference values v and V (v x v2) in J5 as shown in FIG.

Using these values, the sections CR with small chroma, CR, medium CR, and CR with large chroma shown in FIG.

CR:0≦CR<CR1 CR:CR,≦CR<CR2 CR:CR2 CR≦CRIllax■ Also, similarly, the section with small brightness V■1. medium interval v
8. The large interval is set as follows.

■[2〇≦V<Vl vH: v1≦■≦V2 V■: v2kuv≦vllax Note that the reference values CR, CR, Vl, V217) Each value is based on the primary color components B, G, and According to the range of R,
For example, specified by the operator.

The above sections CR, CR, CR, and L
By H and intervals V, V, ,V, as shown in Fig. 3,
The entire color space is divided vertically and horizontally into a plurality of regions R-R9. Further, the regions R1 to R9 are classified into the following groups.

Saturation CR and brightness [V are small or medium regions R, R2, R4, R5 are the standard color group GHR
are categorized. Regions R7 and R8 where the saturation CR is low or medium and the brightness V is high are classified into a light color group G.

Region R where saturation CR is large regardless of the magnitude of brightness ■
3, R, and R9 are brightly colored group G□. are categorized.

Further, for regions near the mutual boundaries of each region R1 to R9, a fitness degree other than "O" or 1 is given by a membership function described later.Reference values CR, CR
2a and reference mv,.

A small width ΔCR1°ΔCR centered on each of v2
and width Δ■4. Each of Δv2 is set corresponding to each neighborhood area.

In step S12, masking equations defined by mutually different masking matrices are set for the groups of regions R1 to R9 classified as described above. Regions R1゜R2, R4, R5 are groups”
NR, and a masking matrix ANR that performs standard masking is set in common for these areas. Similarly, masking matrix A is applied to regions R7 and R8 belonging to group GP, and region tiI belonging to group GIIC is
! A masking matrix A is set in R3°R and R9.

11C Masking matrices A , A are light and bright
Perform appropriate masking for each HC color. By individually setting a masking matrix for each region group, for example, complex green colors of plants and trees can be reproduced three-dimensionally.

In step 813, a membership function is set that gives a degree of suitability to each section according to the saturation CR and brightness V of each point in the color space, corresponding to the division performed in step S12.

FIG. 4A is a graph showing an example of the membership function for saturation CR, and FIG. 4B is a graph showing an example of the membership function for brightness V. In FIG. 4A, sections CR and CRH.

Membership functions FC, FC, which give the goodness of fit wc, , wc, , wC, to each CRH.

H [ FC is shown. Each function FC, FC,.

HL FCH is the reference value CR1, C which is the mutual boundary of each section.
From WC=O (nonconforming) to WC within each neighboring region defined by widths ΔCR, , ΔCR2 with R2 as the center.
= 1.0 (perfect fit) or WC = 1.0 to W
Smoothly changes to C=O. Similarly, in Figure 48,
Membership functions FV, FV, FV, . . . that give degrees of fitness wv , wv, . These functions FV, FV, , FVl[ also have widths Δv1 . Within each neighborhood defined by ΔV2, from Wv−〇 to W=
1.0 or WV = Smoothly changes from 1.0 to WV-〇. Therefore, these neighboring regions (boundary regions)
Then, these membership functions indicate a degree of fitness intermediate between "perfect fit" and "no fit."

In addition, the saturation C corresponding to the points Pi and P2 shown in FIG.
R, CR and brightness V ■ are Pi P2
Pl/P2 are shown in Figures 4A and 4B, respectively. These saturations CR,1,CRp
2, brightness V, and 1°VF6 are used, for example, when calculating the degree of suitability described later.

In step 814, the saturation CR and brightness V of the image data at each point in the color space are substituted into the corresponding membership function to determine the degree of suitability for each interval and each region. Note that the method for determining the degree of fitness is based on the generally known fuzzy control theory.

Corresponding to the division of the color space in FIG. 3, inference formulas (8A) to (8G) indicating which masking matrix should be used for each point in the color space are set as follows. However, rAJ indicates a masking matrix to be obtained as a result of inference.

IF CR15CREand V is VL
then A = A NR・(8A)IF CR
iS CR[and V is VH then
A = A MR・(8B)IF CRis
CR[and V is V,.

then A=Ap'...
(8C) IF CRis CR, and V
is VLthen A = ANR・−(80)
IF CRis CR, and V is
VHthen A = A Hll
・(8E)IF CR+s CRHand
V+s Vllthen A=Ap
'...(8F)IF CRis
CRH then A = A IIc
'...(8G) First, calculate the goodness of fit for the antecedent part shown in the IF clause of equations (8^) to (8G) at point P1 shown in Figure 3.
As an example, calculate as follows (however, point P1 is region t
ii! (It is a point located within the central part of R5). Formula (8
The antecedent part of ^) indicates the condition that the saturation CR is within the interval CR, and the brightness ■ is within the interval VL.

In this way, if the antecedent part consists of two conditions 'and''', find the fitness corresponding to each of the two conditions, and select the smaller fitness of the two conditions. Let be the fitness of the entire antecedent part.

As shown in FIGS. 4A and 4B, the saturation C of point P1
The interval CR4 given by the function FC,- of R, ,
WC4 is ":0", and the interval V given by the function FV of brightness VP1.

The goodness of fit WL is also 10''. Therefore, the goodness of fit wc
The smaller value of [, wv[ is also 0′′, and the fitness degree WIA of the point P1 to the antecedent part of equation (8^) is “0”. Note that the goodness of fit wlA to the antecedent part of equation (8A) is
It corresponds to the degree of conformity to.

Similarly, point P to the antecedent part of equations (8B) to (8F)
The fitness degrees WI8 to WIF of 1 are determined as follows.

WIB=0 (WC, -0, WVM=1) Wlo=O
(WCL=O, WVH=O)Wlo-0(WCH-1,
WV, -0) Wl = 1 (WC-1, WVH-1
)E H Wl =O(WC=1.WV,,-0)F
H Furthermore, since the antecedent part of equation (8G) is composed only of conditions related to saturation CR, the fitness degree WC□ becomes the fitness degree Wt6 as it is.

Wl =WC,,=0 In addition, each of the degrees of fitness Wl -Wl, is the region R,
The suitability W16 corresponds to the suitability for each of R, R, R, and R8, and the suitability W16 corresponds to the suitability for the group G11° constituted by the regions R3°, R6, and R9.

In the manner described above, the goodness of fit Wl of the point P1 to each antecedent part is determined. Since the point P1 is located approximately at the center of the region R5, only the degree of conformity WIE is 1'', and the other degrees of conformity are 0''.

Next, as another example, it belongs to the region R6 and the regions R, R
, R5, and the degrees of conformity W1 to W+6 to the antecedent part regarding the point P2 near the boundary with the antecedent part are determined. As shown in Figure 4, the saturation C
As the goodness of fit WC for R, , each function FC, FC
H, FCII, WCl-0, WC, -o, 3. WCH=0.7
are given respectively. In addition, as shown in FIG. 4B, as the degree of suitability WV for the brightness VP2, Wv, =
0.2. WV = 0.8. WVH-0 is given. Using these numerical values, the goodness of fit WIA-WIo regarding point P2 is obtained as shown in the following equation (9).

WIA=O(WCl-0,WV,-0,2)W,=
O(WC,=O,WVH=0.8)W=0
(WC = O, WVFl = 0) CL W = 0.2 (WC = 0.3. WV, -
0,2) Mouth H W = 0.3 (WC = 0.3. WVH =
0.8>H Wl =O(WC14= 0.3.WVH=O)W
l = 0.7 (WC, = 0.7)... (
9) Since point P2 is located near the boundary with regions 1dR, R3, and R1 in region R,
Each of the HCs has a fitness that is not 0°.

In step 815, the fitness W obtained in step 814 is
l -WIGl, 1:Jl, the weighting coefficient WHR0w for each masking matrix ANR9AP'' IIc
, w, c.

The weighting coefficient WHR for □ is given to the masking matrix by equations (8^), (8B), (8D), (8E)
It is. The goodness of fit of the antecedent part of these equations Wl, WI
,Wl,.

The maximum value among A B Wl becomes the weighting coefficient WIIc. Similarly [, the maximum value of the fitness degrees Wl, WI, is the weighting coefficient W
becomes. In addition, the fitness degree WIo is the weighting coefficient W
It becomes IIc. The following formulas (10A) to (10C) are formulas showing these relationships.

WII1” MAX (WIA, WI, Wl, WIE) D ... (IOA) W -MAX (Wl, Wl,) ... (1
0B) C WII. -WIG...(IO
C) Using the relationships of formulas (10A) to (10C), point P1
When the weighting coefficient for W is calculated, the fitness W
Since only IE is “1”, W = 1. W,=W, C=O.

NR Regarding point P2, using each numerical value of the above-mentioned formula (9), WNR=0.3. wP=0. WHc=0.7
becomes.

In step 816, the primary color components B and G of each point are determined.

R or sub color components g, r are each masking matrix ANR
2A3. A and C are converted into temporary secondary color components YM and C or subcolor components y, m, and c.

For example, using Equation (4) described above, a temporary secondary color component "
NR・mNR・0NR−”P=mP・0P−”
HC・mlIC2CHe has the following formula (11A) ~(1
1C).

(12C) Note that this process can be performed independently of the process for determining the degree of suitability described above, so these processes may be performed in parallel or in the reverse order.

In step 817, the weighting coefficient WNr1. determined in step 815. Based on WP, WHc, a weighted average of the provisional secondary color components determined in step 816 is determined, and processed secondary color components y, m7. Let it be cz. Processed secondary color components y, m,
cz is given by the following formulas (12^) to Z (12C), respectively.

...(12^) ...(12B) For example, in the case of a point located in the center of one area, such as point P1, the value of any one of the weighting coefficients WNR1WP'WIIc becomes "1n". , other coefficients are 0''. Therefore, the masking matrix ANR1AP −AlI
c provides optimal masking for that point and the processed secondary color component “Z 2m2 ” Z
is obtained.

In addition, for a point, such as point P2, which is located in a region near a mutual boundary of a plurality of regions and whose boundary is a mutual boundary of each group GME GP' GIIc, the weight is Any two or more of the coefficients WNlt and WP'WHO take values other than 0°' and "1". Therefore, the masking matrix AH1<.

Masking is performed using two or more corresponding matrices of A and AH, and the results are weighted averaged.
A processed secondary color component yz 1mz 'C2 is obtained.

In this way, it is possible to make the reproduction characteristics at the boundary continuous while making the optimum masking matrix correspond to each region.

By performing the above processing, masking is performed according to subtle changes in each tone, even for patterns that are of the same family and have complex tones, such as rose red and mountain green. A three-dimensional reproduction image can be obtained. Furthermore, since smooth and continuous reproduction characteristics are provided even at boundaries where the color tone changes, a natural-looking image can be obtained as a whole.

The D0 circuit configuration in FIG. 5 shows that in order to realize the processing described above,
3 is a circuit diagram embodying the masking circuit 400 of FIG. 2. FIG.

The primary color components B, G, and R are input to an adder 1. The color components B, G, and R are also input to the minimum value selection circuit 2 and the maximum value selection circuit 3, and the minimum value (lightness V) of the color components 8°G and R and the minimum value (lightness V) of the color components 8. G.

Each of the maximum values of R is selected.

The brightness V is input to the negative input terminal of the adder 1.

The adder 1 generates sub-color components S, g, and r by subtracting the brightness V from each of the negative sub-color components B, G, and R. Also,
The brightness V is also input to the negative input terminal of the addition B4. The adder 4 subtracts the brightness V from the maximum value of the color components B, G, and R from the maximum la selection circuit 3, and generates the saturation CR based on the above-mentioned equation (7).

The sub-color components s, g, and r generated by the adder 1 are input in parallel to masking calculation circuits 5, 6, and 7, respectively.

The masking calculation circuit 5.6.7 performs calculations corresponding to the above-mentioned masking matrices ANR'' P and AlIc on the sub-color components g, r, respectively, and calculates the above-mentioned equation (
Temporary secondary color components <y, m, c> defined by 11A) to (11C).

Generate each combination of P P P MCIIc HC of NRNR811 (V , m , C), <V , m , C). In addition, the masking matrix AN
Circuits corresponding to R, A, , A□C are set in advance in the masking calculation circuit 5.6.7.

Further, the saturation CR is input to a suitability output circuit 8.9.10 composed of an LUT (look-up table), and the brightness ■ is input to a suitability output circuit 8.9. In the fitness output circuit 8.9.10, the membership functions illustrated in FIG. 4A and FIG. 4B are preset, and steps 813, 814, and 815 in FIG.
Weighting coefficients W 1 , W 2 , W, , o corresponding to the input saturation CR and brightness ■ are generated based on the rules shown in FIG.

NRP The arithmetic circuit 11 generates a temporary secondary color component "NR2"P.

y and weighting coefficients W, W, WH, are input, 11c
NRPThe above formula (12
Based on A), a processed secondary color component V7- is generated. Similarly, the arithmetic circuit 12 calculates temporary secondary color components m, rTI,
m□. and weighting coefficients are entered, the processed secondary color component m7 is generated based on the above-mentioned formula (12B), and the arithmetic circuit 13 generates a temporary secondary color component C.
, C, C and weighting coefficients are inputted, and the above-mentioned formula (12C) processed secondary color component C7 is generated.

Processed color component y generated by arithmetic circuit 11.12.13
, m, and cl are each added to the brightness V in the adder 14 to generate secondary color components Y, M, and C. Also, light mV
is output as is as a secondary color component.

1111 In the above example, processing such as dividing the color space and setting degree of suitability was performed by focusing on saturation and brightness, but similar processing may be performed by focusing only on saturation.

Further, when setting a masking equation for each region, saturation is the most important factor, but similar processing can also be performed using parameters related to the color of the square or a combination thereof.

In addition, the above-mentioned formula (3A) is used as the definition formula for saturation and brightness.
Alternatively, an equation other than equation (7) may be used. For example, the expression (3^
Instead of the brightness V shown in ), a brightness V' according to another definition may be given as shown in equation (3A').

V'""'b -B0a -G0a, -R- (38°) However, the coefficients ab, a, , a, are each color component B.

This is a predetermined coefficient set corresponding to G and R.

Further, such processing is realized, for example, by a masking circuit as shown in FIG.

Furthermore, as a more precise masking equation, a masking matrix that takes quadratic term components and cross term components into consideration may be used.

(Effects of the Invention) As described above, according to the present invention, the masking equation is set for each of a plurality of regions in which the color space is divided according to saturation or saturation and brightness. The reproduction characteristics are given individually.

Furthermore, since the degree of suitability of the first image data to multiple regions is given by a membership function defined by saturation or saturation and brightness, the reproduction characteristics at mutual boundaries between multiple regions are continuous. .

Therefore, by setting the optimal masking coefficient for each divided color space, it is possible to obtain good reproduction characteristics in the golden space, and to make the reproduction characteristics continuous even at the boundaries between regions, resulting in a natural duplicate image. A method of masking color image data can be obtained.

[Brief explanation of the drawing]

FIG. 1 is a flowchart showing the processing procedure of a color image data masking method according to an embodiment of the present invention, FIG. 2 is a schematic block diagram of a prepress scanner to which an embodiment of the present invention is applied, and FIG. 3 is a divided 4 is a graph of a membership function according to an embodiment of the present invention. FIG. 5 is a circuit diagram of a masking circuit according to an embodiment of the present invention. FIG. 7 is a circuit diagram showing a partial modification of the masking circuit shown in the figure. 400...Masking circuit, B, G, R...-Next color component (first image data), Y
, M, C, K...Secondary color component (second image data), CR...Saturation, ■...Lightness, R
1 to R9... area, ANR, AP "HC"' masking matrix, FC. WO2 WNR・yNR・FV...Membership function, WV, WI...Fitness, WP' WIIc...Weighting coefficient,

Claims (3)

[Claims] (1) Convert the first image data representing the first color component obtained for each pixel by scanning the original image into second image data representing the second color component based on a predetermined masking equation. A method for masking color image data, comprising: (a) dividing a color space into a plurality of regions according to saturation; (b) preparing the masking equation for each of the plurality of regions. (c) determining the suitability of the first image data to each of the plurality of regions by substituting the saturation of the first image data into a predetermined membership function indicating the suitability for the saturation; (d) converting the first image data into second image data based on the masking equation and the degree of compatibility with each of the plurality of regions. masking method. (2) Convert the first image data representing the first color component obtained for each pixel by scanning the original image into second image data representing the second color component based on a predetermined masking equation. A method for masking color image data, comprising: (a) dividing a color space into a plurality of regions according to saturation and lightness; (b) preparing the masking equation for each of the plurality of regions. , (c) the plurality of regions of the first image data by substituting the saturation and brightness of the first image data into a predetermined membership function indicating the degree of suitability for each of the saturation and brightness. (d) converting the first image data into second image data based on the masking equation and the degree of fit to each of the plurality of regions; A method of masking color image data including . (3) The step (d) according to claim 1 or 2 comprises (d-1
) converting the first image data into a plurality of provisional second image data based on masking equations prepared for each of the plurality of regions in the color space; A method for masking color image data, comprising the steps of determining a weighted average of the plurality of provisional second image data based on the degree of conformity to each, and determining second image data based on the weighted average.