JPH0722340B2 - Normalized curve generation method - Google Patents

Description

TECHNICAL FIELD The present invention relates to a normalization curve generation method, and more particularly to a method suitable for automatic generation of a normalization curve in a platemaking scanner or the like. Further, as one application thereof, a method of systematically correcting the color cast of only the shadow portion of the color original image in a color image processing apparatus such as a plate-making color scanner is also targeted.

[Conventional technology]

In a plate-making color scanner or the like, an image of a color original image is color-separated, and then each color component is converted into a dot signal to record a duplicate image. When a color cast or the like exists in the original image, the color separation condition is corrected by the operator's judgment so that the gray balance is ensured in the recorded image.

In the first conventional method for correcting the color separation condition, a process for converting the density signal for each color component obtained by reading the original image so as to match the density range of the color separation device, that is, a normal process. In the conversion, the highlight density value and the shadow density value are set to different values for each color component.

Further, in the second conventional method, the highlight density value and the shadow density value are common values for the respective color components, and yellow (Y),
At the stage of obtaining halftone dot signals for each color plate of magenta (M), cyan (C) and black (K), the halftone dot area ratio for each color plate is corrected.

[Problems to be Solved by the Invention]

There are two major problems with such conventional methods.

One of them is that the correction of the color separation condition is basically performed based on the manual operation of the operator, so that a skilled operator is required to make the appropriate correction. Then, this kind of problem exists not only in the correction of the color separation condition but also in the normalization conversion in the monochrome image.

The other one is a problem peculiar to the above gray balance. That is, the above-mentioned two conventional methods are used when a color cast is generated in the low and middle density areas due to the deviation of the light emission characteristics of the light source when the subject is photographed for creating the original image, or when the original image or the entire image is faded. , The effect of improving gray balance is recognized.

However, if there is color cast only in the shadow area due to the maximum expressible density of the original film itself being different for each color component, it is difficult to make appropriate corrections using these conventional methods. Is. That is, in each of the above-mentioned conventional methods, since the correction is made uniformly over a wide range of the density, when the correction for removing the color cast of the shadow portion is attempted, the effect of the correction also affects the low and medium density areas, and This results in the loss of gray balance in the density area.

In order to remove the color cast only on the high density side while preventing such a situation, the UCA / UCR (under color addition / removal) function in the color separation device may be used. By using this function, the halftone dot area ratio can be adjusted for each color plate for a gray component having a predetermined density or higher.

However, when using the UCA / UCR function, the lower limit threshold of the density range to be corrected must be manually set for each color plate. Also. Since the correction amount is also based on the darkest part in the original image, the determination of the correction amount largely depends on the intuition of the operator. That is, the work of removing the color cast by UCA / UCR is a one-off
It is not systematic. For this reason, it is impossible to systematically and quantitatively manage the amount of color cast correction by UCA / UCR, which is a cause of impeding automation of color separation conditions.

Further, since the UCA / UCR circuit is provided on the rear side of the black plate signal generation circuit of the color separation device, each circuit existing on the front side of the UCA / UCR circuit outputs a signal before gray balance correction. It must be processed, and the setting of its processing characteristics becomes complicated.

Furthermore, there is no technology developed to handle gray balance correction in cases where the entire original image has faded, and color cast correction only in the shadow area, and these must be considered individually. There are also problems.

[Object of the Invention]

The present invention is intended to overcome the above-mentioned problems in the prior art, and is to provide a method for generating a normalized curve that can be systematically carried out without the need for a skilled operator and is particularly suitable for automation. The purpose of.

A second object is to provide a normalized curve generation method which is applied to the correction of the color separation condition for the color original image and which can systematically and quantitatively manage the color cast correction amount of only the shadow portion.

A third object is to facilitate setting of processing characteristics in these circuits by giving a signal after gray balance correction to each circuit in the color separation device.

Furthermore, it is possible to handle gray balance correction and color cast correction only in the shadow area in a unified manner when the entire original image has faded, thereby providing a technology particularly suitable for automating color separation conditions. The fourth purpose is to do so.

A fifth object is to achieve the first object, and in particular, to generate a normalization curve that can prevent the density expression of the converted image from being largely influenced by the density range of the original image. is there.

[Means for Solving the Problems]

A first configuration of the present invention provides a normalization curve for converting a density value of each pixel of an image of an original image to be processed into a value within a normalized density range suitable for image processing in a predetermined image processing apparatus. It is generally applied when generating. In this method, (a) with respect to the original image to be processed, a highlight density value and a shadow density value are determined from the range of possible density values before conversion, and (b) according to a predetermined determination rule, A dummy shadow density value for the density value before conversion is determined.

(C) Then, as the normalization curve, (c-1)
On the coordinate plane on which the normalization curve should be defined, a first point that is determined according to the highlight density value and a second point that is determined according to the shadow density value are substantially passed. Condition 1 and (c-2) in the density range near the highlight density value, both the first point and the third point on the coordinate plane determined according to the dummy shadow density are set. A curve satisfying both the second condition of changing along a passing straight line and the second condition is generated as a normalized curve.

In the second configuration of the present invention, when the color original image is color-separated for each pixel using a predetermined color separation device,
In the process of generating a normalized curve for each color component for converting a density value for each color component of each pixel into a value within a normalized density range suitable for color separation in the color separation device, the first configuration described above. Method is applied. Since the original image is a color original image, each of the highlight density value, the shadow density value, the dummy shadow density value, and the normalization curve is determined for each color component.

Then, in the step (b), (b-1) when color cast correction of only the shadow portion of the color original image is required, a value deviated from the shadow density value by a value corresponding to the color cast amount is used. Obtained for each component and set them as the dummy shadow density value for each color component, and when color cast correction of only the shadow portion is not required, a value substantially the same as the shadow density value is set as the dummy shadow. It is executed so as to include the step of setting the density value. That is, in this method, when the density value for each color component of the original image is normalized and converted,
It is also configured to perform color cast correction only on the shadow portion.

In the third configuration, in the above second configuration, (d) in order to generate a normalized curve, (d-1) having a highlight density value, a shadow density value and a dummy shadow density value as parameters, and converting Having the previous density as a variable, (d-2) when the shadow density value and the dummy shadow density value are different values, the variable is non-linear, and (d-3)
When the shadow density value and the dummy shadow density value are the same value, a linear function for the variable is determined in advance.

Then, (e) the normalized curve is (e-1) after determining whether or not the original color image requires color cast correction of only the shadow portion, and (e-2) color cast correction of only the shadow portion. Is required, the shadow density value and the dummy shadow density value are set to different values to specify the parameter of the function, and (e-3) it is necessary to correct the color cast of only the shadow portion. If not, it is determined by setting the shadow density value and the dummy shadow value to the same value and specifying the parameter of the function.

On the other hand, the fourth configuration is applicable regardless of whether the original image to be processed is a monochrome original image or a color original image. In the method of the first configuration, the step (b) includes (b-1). The method includes a step of setting a value obtained by adding a value proportional to the density range width of the standard original image and a value proportional to the highlight density value as a dummy shadow density value. Unlike the second and third configurations, the fourth configuration does not target color cast correction of only the shadow portion.

In the fifth configuration, in the fourth configuration, the weighted average value of the shadow density value of the standard original image and the density range width is set to a value proportional to the highlight density value of the original image to be processed determined in step (a). The dummy shadow density value is determined as the added value.

It should be noted that the term "density" in the present invention is a general term that represents not only the optical density but also the signal level obtained by photoelectrically reading the density, the Munsell value, and other quantities that express the density of image gradation. Is a term.

[Action]

In the first configuration of the present invention, the first to third points are specified on the normalized conversion coordinate plane in accordance with the highlight density value, the shadow density value and the dummy shadow density value. In the highlighted portion on this coordinate plane, the normalization curve has a shape along a straight line connecting the first and third points. Then, in the shadow portion, the normalization curve deviates from this straight line and passes through the second point.

In the second configuration, the shadow density value and the dummy shadow density value are determined such that the difference therebetween corresponds to the amount of color cast of only the shadow portion of the original image. Therefore, the deviation of the normalization curve from the straight line has a function of removing the color cast of only the shadow portion. Since this deviation substantially occurs only on the shadow portion or in the vicinity thereof on the coordinate plane, the gray balance in the low and middle density areas is not disturbed.

In the third configuration of the present invention, the normalization curve is generated by specifying the parameter value of the function that satisfies the conditions (d-1) to (d-3). This function can be commonly used when the color cast is generated only in the shadow portion and when the color cast is not generated.

In the fourth configuration, by using the value of the density range width of the standard original image in the process of determining the dummy shadow density value, even if the density range of the original image to be processed changes, the highlight density in the highlight portion or the intermediate density portion is increased. Density values having the same density difference from the values are converted to density values that are substantially the same or relatively close by normalization.

Further, in the fifth configuration, the dependency of the dummy shadow density value on the value of the density range width of the standard original image can be changed by changing the weighting coefficient in the weighted average.

〔Example〕

A. Overall structure and schematic operation of the first embodiment FIG. 1 is an external view of a plate-making color scanner to which the embodiment of the present invention is applied, and FIG. 2 is a block diagram showing its internal structure.

In FIG. 1, the plate-making scanner 100 is configured by electrically connecting a scanner main body 200 and a color separation auxiliary device 300 by a transmission cable 101. Scanner body 10
Reference numeral 0 has an external configuration roughly divided into an input unit 201 and an output unit 202, and the input unit 201 includes an original image drum 2 and a pickup head 3. Also, the output unit 202 is the recording drum 1
The operation panel 203 of the scanner body 200 is provided on the input section 202 side in the illustrated example.

The scanner body 200 itself does not have an automatic setting function of color separation conditions, and the function is achieved by the color separation auxiliary device 300. The color separation auxiliary device 300 has an external appearance in which the CRT 18 and the console 17 are provided on the device main body 301.

Next, referring to FIG. Of the respective elements shown in FIG. 2, elements other than the elements inside the broken line frame showing the color separation assisting device 300 are constituent elements of the scanner body 200 of FIG. A positive color original image film 1 is wound around an original image drum 2, and a pickup head 3 is provided so as to face the drum 2. The main scan and the sub-scan of the original image 1 are achieved by the rotation of the drum 2 in the α direction and the translation of the pickup head 3 in the β direction, and the image of the original image 1 is photoelectrically photoelectrically pixel by pixel in the scanning line by the pickup head 3. Read.

The image information of the original image is given to the input circuit 4 for each color component of blue (B), green (G) and red (R), and in this input circuit 4, digital color density for each of B, G and R is given. Signals D _B , D _G , D _R.

These signals D _B , D _G , and
D _R is stored in the frame memories 13B, 13G, and 13R for each pixel. The information processing device 14 is, for example, a microcomputer, and receives these signals D from the frame memories 13B, 13G, 13R.
_B , D _G , D _R are taken in and a normalization curve for each color component is generated. The details of this process will be described later, but CPU 15 and memory 1
6 is used in this process. Also console 17
Is for manually inputting various commands and data to the information processing apparatus 14. Then, the normalized curve created by the information processing device 14 is used as numerical data for each color component as a look-up table memory (LUT) 5B, 5B.
It is loaded into G, 5R.

Next, a main scan of the original image 1 is performed for image recording. The signals D _B , D _G , and D _R obtained at this time are subjected to normalization conversion by the LUTs 5B, 5G, and 5R to become signals D _NB , D _NG , and D _NR , which are given to the color operation circuit 6 in the next stage. In the color calculation circuit 6, the signals D _NB , D _NG , D
Color calculation based on _NR is performed, and each color plate signal for Y, M, C, and K is generated. The respective road version signals are converted into Y, M, C and K halftone signals in the halftone converting circuit 7, combined in a time division manner by the selector 8 and given to the output circuit 9.

The recording head 10 has a built-in laser light source,
The laser beam is ON / OFF-modulated by the modulation signal given from the output circuit 9. The photosensitive film 12 is wound around the recording drum 11, and due to the combination of the α rotation of the drum 11 and the translation of the recording head 10 in the β direction, the exposure of the photosensitive film 12 by the laser beam is performed in scanning line sequential and It is performed for each pixel. In the following, it is assumed that the Y, M, C and K color plate images are positively recorded on the photosensitive film 12.

B. Normalized Curve Generation Operation in First Embodiment FIGS. 3A and 3B are flowcharts showing the operation of the color scanner 100 described above, and this operation is suitable for color cast correction of only the shadow portion. ing. Also,
FIG. 4 is a diagram conceptually showing the process of generating the normalized curve line, and S1, S101, ... In FIG.
Corresponds to the step numbers in the figure and Figure 3B. In addition, white arrows indicate data usage relationships.

In the following, the process of normalizing curve generation will be explained step by step.

(B-1) Characteristics of Average Color Density Value First, in step S1 in FIG. 3A, the entire copy target portion of the original image 1 is prescanned. Then, the color density signals D _B , D _G , D _R (FIG. 2) for each pixel are stored in the frame memories 13B, 13B.
Store in G, 13R. Hereinafter, the color density values designated by these signals D _B , D _G , and D _R will be represented by the same symbols D _B , D _G , and D _R. When the size of the original image 1 is large, the color density values D _B , D _G ,
This storing operation of D _R is executed while thinning out pixels.

Next, in step S100, from these color density values D _B , D _G , D _R , highlight side average color density values D _HAB , D _HAG , D _HAR and shadow side average color density values D _SAB , D _SAG , D _SAR. This is a subroutine for obtaining and. This subroutine is described in detail in Japanese Patent Application No. 63-312684 previously filed by the applicant of the present invention, but only an outline thereof will be described here.

In FIG. 3B showing the details of step S100, in the first step S101, the density values D _B , D _G ,
The average density value for each pixel: D _M = (D _B , D _G , D _R ) / 3 is calculated from D _R. Further, this calculation is performed for all the pre-scanned pixels, and the abscissa is the class indicating the range of the average density value D _M , and the ordinate is the number of pixels N _P , and the average density value frequency histogram is as shown in FIG. To create. In FIG. 5, the median of the classes is indicated by D _M i (i = 1 to n).

In step S102, the density values D _B , D _G , and D _R for each color component B, G, and R of each pixel included therein are cumulatively added for each class of the average density value frequency histogram. This process is performed independently for each class. After performing this calculation, each class value D _M i (i = 1 to n) is plotted on the horizontal axis, and the cumulative density values ΣD _B , ΣD _G , ΣD _R corresponding to the pixels included in each class are plotted on the vertical axis. , A cumulative density value histogram is created for each color component. FIG. 6 shows an example of this cumulative density value histogram, and such a cumulative density value histogram is obtained for each of X = B, G, and R.

As an example, a class with a class value D _M i = 1.0 and a class width of 0.1 (0.95 ≦
D _M <1.05 will be described. For the sake of convenience, the number of pixels included in this class is set to three, and the density data of each pixel is pixel 1: D _B = 1.10, D _G = 0.90 D _R = 0.95 (D _M ≈0.98) Pixel 2: D _B = 1.00 , D _G = 1.10 D _R = 0.90 (D _M ≈ 1.00) Pixel 3: D _B = 1.00, D _G = 0.95 D _R = 0.95 (D _M ≈ 0.97)

When the cumulative density value histogram shown in FIG. 6 is X = B, the cumulative density value ΣD _B in that class (0.95 ≦ D _M <1.05) is ΣD _B = 1.10 + 1.00 + 1.00 = 3.10. Similarly, the other cumulative density values ΣD _G and ΣD _R are also ΣD _G = 0.90 + 1.10 + 0.95 = 2.95 ΣD _R = 0.95 + 0.90 + 0.95 = 2.80.

This kind of processing is performed for each class and each color component B, G, R
Each cumulative density value histogram is completed for each. (See also the block of step S102 in FIG. 4. Incidentally, in FIG. 4, each histogram is approximately drawn by a curve. Further, steps S101 and S102 are actually processed in parallel.) In step S103, from the average density value frequency histogram shown in FIG. 5, the relative frequency RN (%) of the pixels cumulatively added from the lower density is plotted on the horizontal axis of the class value D _M i (i = 1 to n). The cumulative relative frequency histogram as shown in FIG. 7 is created on the vertical axis. The histogram has a shape that changes from 0% to 100% within the range of the minimum and maximum generated density values D _M min and D _M max. However, in FIG. 7, this cumulative relative frequency histogram is approximated by a curve on the assumption that the class width is sufficiently small.

In the next step S104, the predetermined cumulative concentration appearance rate RN
_H and RN _S are applied to the cumulative relative frequency histogram shown in FIG. 7 to obtain the temporary highlight average density value D _MH and the shadow average density value D _MS , respectively. The values of the cumulative density appearance rates RN _H and RN _S are values obtained in advance by analysis of a large number of sample original images so as to give statistically optimum highlight points and shadow points, and for example, 1
%, About 98%.

In step S105, among the cumulative density value histograms for each color represented in FIG. 6, as shown by hatching in FIG. 8, the highlight side is less than the temporary highlight average density value D _MH. region _{(D M min ≦ D M ≦} D MH), the shadow side, the temporary shadow average density value D _MS or more regions (D
Focusing on each of _MS ≤ D _M ≤ D _M max), pixel averages of the cumulative density values ΣD _B , ΣD _G , and ΣD _R within that range are performed for each color component.

That is, if the pixel average on the highlight side is written as <...> _H and the pixel average on the shadow side is written as <...> _S , the highlight side average color density value D obtained by this pixel average is written.
_HAB , D _HAG , D _HAR and shadow side average color density value D _SAB , D _SAG , D
_SAR is D _HAX = <ΣD _X > _H (X = B, G, R) (1) D _SAX = <ΣD _X > _S (X = B, G, R) (2)

For example, X in FIG. 8 is B, and the temporary highlight density D
_{When MH} is D _M5 , D _HAB = (D _M1 + D _M2 + D _M3 + D _M4 + D _M5 ) / (N _P1 + N _P2 + N _P3 + N _P4 + N _P5 ). However, N _P i (i = 1 to 5) is the average density value D _M
Is the number of pixels belonging to the class D _M i, and is obtained from the histogram in FIG.

Thus, the highlight side and shadow side average color density values D _HAX , D for each color component shown in the center of FIG.
_SAX (X = B, G, R) is obtained. With the above, the subroutine corresponding to step S100 is completed, and the process returns to the main routine of FIG. 3A.

(B-2) Calculation of Reference Value and Correction Amount In the next step S2 in the main routine, the highlight reference value D _H0 and the shadow reference value D _S0 common to the B, G, and R color components are calculated by the following (4 ), Equation (5).

D _H0 = FH (D _HAB , D _HAG , D _HAR ) (4) D _S0 = FS (D _SAB , D _SAG , D _SAR ) (5) However, the functions FH, FS are FH (D _HAB , D _HAG , D _HAR ) = r _H・ MIN (D _HAB , D _HAG , D _HAR ) ＋ (1-r _H ) D _HF (6a) FS (D _SAB , D _SAG , D _SAR ) = r _S・ MAX ( D _SAB , D _SAG , D _SAR ) ＋ (1-r _S ) D _SF … (6b), MIN (…): Minimum value selection operation, MAX (…): Maximum value selection operation, D _HF , D _SF : Standard standard highlight density value and shadow density value, r _H , r _S : 0 <r _H <1,0 <r _S <1, a preselected constant. That is, the highlight reference value D _H0 is D _HAX (X =
B, G, R) is the weighted average of the minimum value and the standard value D _HF , and the shadow reference value D _S0 is the maximum value of D _SAX (X = B, G, R) and the standard value D
_It is a weighted average with _SF . The functions FH and FS are D _HAX (X
= B, G, R) and D _SAX (X = B, G, R).

In the next step S3, the gray balance correction amount ΔD _HX (X = B, G, R) in the highlight part and the gray balance correction amount ΔD _SX (X = B, G, R) in the shadow part are It is calculated using equations (7a) and (7b).

ΔD _HX = K _H · GH (D _HAX −D _H0 ) (X = B, G, R) (7a) ΔD _SX = K _S · GS (D _SAX − D _S0 ) (X = B, G, R) (7b) However, K _H and K _S are positive constants that have been experimentally determined in advance, and the function GH and GS is, for example, GH (D _HAX −D _H0 ) = (D _HAX −D _H0 ) / [1+ (D _M max −D _M min) / A _H } m] (8) GS (D _SAX −D _S0 ) = D _SAX −D _S0 (9) However, D _M max = MAX (D _HAB , D _HAG , D _HAR ) ... (10a) D _M min = MIN (D _HAB , D _HAG , D _HAR ) ... (10b) A _H : Preselected positive constant , M: a preselected positive constant, eg "3" ,.

As can be seen from the equations (6a) to (9), the correction amount ΔD _HX of the highlight portion is (D _HAX −D _H0 ), and the correction amount ΔD _SX of the shadow portion is (D _SAX −D _S0 ). Each is proportional. (D _HAX
−D _H0 ) reflects the amount of color cast in the highlight part, and (D _SAX −D _S0 ) reflects the amount of color cast in the shadow part.

By the way, focusing on the gray balance correction amount ΔD _HX in the highlight part, (7a), (8), (10a), (10
by b) formula, B in the highlight portion, G, variations in the average density value of R each color component _{Width: ΔD H m≡D H max-D} H min ... (11) the larger the gray balance correction amount [Delta] D _HX Becomes smaller. Therefore, when the highlight point does not exist in the original image 1, it is possible to prevent the gray balance correction amount in the highlight portion from becoming more than necessary. The constant A _H is a threshold value corresponding to the presence / absence of the highlight point, and when ΔD _H m> A _H (12), the above correction suppression effect becomes large.

(B-3) Calculation of highlight density value, shadow density value and dummy shadow density value In the next step S4, the reference values D _H0 , D _S0 and the correction amounts ΔD _HX , ΔD are calculated.
_SX (X = B, G, R) is used to calculate the highlight density value D _HX and the shadow density value D _SX (X = B, G, R) for each color component,
Calculate using the following equations (13) and (14).

D _HX = D _H0 + ΔD _HX (X = B, G, R) (13) D _SX = D _S0 + ΔD _SX (X = B, G, R) (14) In the next step S5, the original image 1 The operator determines whether or not the color cast correction of only the shadow portion is necessary. Generally, it is difficult to visually determine the color cast of only the shadow portion, so it is determined whether or not the original image 1 has faded,
When the color is not faded, it is considered as the color cast of only the shadow portion, and in the next step S6, the dummy shadow density value D
_DSX (X = B, G, R) is set to a common value using the shadow reference value D _S0 . That is, D _DSB = D _DSG = D _DSR = _SS0 (15).

On the other hand, if the original picture 1 is entirely faded,
The dummy shadow density value D _DSX (X = B, G, R) is set to be the same as the shadow density value D _SX (X = B, G, R) of the corresponding color component. That is, D _DSX = D _SX (X = B, G, R) (16) (step S7). In step S7,
The dummy shadow density value D _DSX is not _{exactly the same} as the shadow density value D _SX, and may be set to a value near D _SX .

(B-4) Generation of Normalized Curve In the next step S8, the highlight density value D _HX , the shadow density value D _SX and the dummy shadow density value D _DSX (X = B, G, R) is used to characterize the normalized curve for each color component. For this purpose, the following (17) ~
The non-linear conversion function f defined by the equation (19) is determined in advance, and information expressing this function is stored in the memory 16.

D _N = f (D: u, v, w) ≡p + (p−q) log ₁₀ [g ₁ + g ₂ ] (17) g ₁ ≡10 ^{− (D−u) / (v−u)} (( 18) g ₂ ≡0.1−10 ^{− (w−u) / (v−u)} (19) where p and q are constants set near the lower and upper limits of the normalized concentration range before masking, D : Density value before conversion, D _N : Density value after conversion, u, v, w: Parameters, and constants p and q are predetermined for each color component. Then, as the values of the parameters u, v, w, u = D _HX , v = D _DSX , w = D _SX (X = B, G, R) (20) for each color component (17) to (19) ), The conversion characteristic of the function f is specified for each color component.

FIG. 9 is a diagram exemplifying the shape of the function f specified in this way on the density normalized conversion coordinate plane. However,
p _X and q _X are values of constants p and q for the color component X. Further, the density variables D _X and D _NX are provided with a subscript “X” to indicate that they correspond to the color component X. The characteristics of the function f will be described later, but an example of the function f embodied in this way is shown as Ca and Cb in FIG. 9 (the difference between Ca and Cb will also be described later).

When the function f is specified for each color component, the CPU 15 generates numerical data expressing the conversion characteristic of the function f, that is, data numerically expressing the curve Ca in FIG. Then, the data is loaded into the LUTs 5B, 5G, 5R of FIG. 2 for each color component. This completes the preparation process for normalization conversion.

In step S9 of FIG. 3A, the main scan of the original image 1 is performed for the purpose of actual image duplication. The operation during this main scan is described above because it has already been described, but the normalized curves in the LUTs 5B, 5G, and 5R generated as described above are the signals D _B ,
Used for conversion of D _G and D _R.

(B-5) Characteristics of Normalized Curve The equations (17) to (19) described above and FIG. 9 are referred to again.
Then, according to the equations (17) to (19), the function f (and thus the normalization curve generated thereby) is
It can be seen that it has the property of (iv).

(I) Since the highlight density value D _HX is considerably smaller than the shadow density value D _SX and the dummy shadow density D _DSX ,
In equation (20), u << v, w (21) holds. Further, since the difference between the shadow density value D _SX and the dummy shadow density D _DSX is relatively small, v and w are about the same value. Therefore, the amount appearing on the right-hand side of (19): the value of (w-u) / (v -u) ... (22) is substantially "1", the definition formula amount g ₂ ((19) The second term on the right side in) is almost "0.1". Therefore, the absolute value of the quantity g ₂ is considerably 0 by subtracting the first term and the second term on the right side.
It is a value close to.

On the other hand, when the density value D (= D _X ) is the same as the value of the parameter u (that is, the highlight density value D _HX ), the exponent on the right side of the equation (18) is “0” and g ₁ is “1”. Become.

For these reasons, the sum (g ₁ +
The main term of g ₂ ) is g ₁ , and g ₁ + g ₂ 1 (23). As a result, D _N p (= p _x ) ... (24) is obtained.

That is, D = D _HX is substantially converted into D _N = P _X by the function f, and the normalized curve in FIG.
Ca and Cb substantially pass through the first point Q _H defined by the two-dimensional coordinate value (D _HX , p _X ).

(Ii) Next, in equations (17) to (19), D = D _SX (shadow density value) (25). Since the value of the parameter w is given by the shadow density value D _SX by the equation (20), the quantity g _{1 of the} equation (18) and (1
It is completely the same as the second term on the right side of equation (9). as a result,
The value in [...] of the logarithmic function of equation (17) becomes "0.1", and D _N = q (= q _X ) ... (26). That is, D = D _SX is converted to D _N = q _X by the function f, and the normalized curve C in FIG.
a, Cb is the second point Q _S defined by the two-dimensional coordinates (D _SX , q _X ).
Pass through. However, the two points Q _S = Q _S shown in FIG.
Of the a, Q _S = Q _S b, the former is the second point Q _S when the gray balance correction amount ΔD _{SX a} in the shadow portion is a positive value ΔD _SX a. In the latter, the correction amount ΔD _SX is a negative value ΔD _SX b
Is the second point Q _S when Density value before conversion D _SX = D
_S a and D _S b indicate respective shadow density values in these two cases. Then, the normalized curves in these cases are shown as curves Ca and Cb, respectively.

(Iii) When the density value D is not the same as the value of the parameter u (= highlight density value D _HX ), but relatively close to it, that is, in the highlight part, the absolute value of the index on the right side of the equation (18) is It is close to “0”. Therefore, the value of g ₁ is “1”
Which is much larger than the value of g ₂ and is log ₁₀ [g ₁ + g ₂ ] log ₁₀ [g ₁ ] = − (D−u) / (v−u) (27). Therefore, the function f in the highlight part is approximated as D _N p− (p−q) (D−u) / (v−u) (28).

By the way, the equation: D _N = p- (p-q) (D-u) / (v-u) (29) is the first point Q _H determined according to the highlight density D _HX (= u). And a third point determined according to the dummy shadow density D _DSX (= v): Q _D = (D _DSX , q _X ) ... (30), the equation of a straight line L (see FIG. 9). Therefore, in the highlight portion HLP of FIG. 9, the normalization curves Ca and Cb change along the straight line L.

On the other hand, as explained in (ii), the normalized curves Ca, C
b is through the second point Q _S, in the example of the first illustrated position of the second point Q _S is different from the third point Q _D. Therefore, in the shadow portion SDP, the normalization curves Ca and Cb deviate from the straight line L, and the deviation amount becomes larger as the density value D becomes larger. Further, the shift amount in the medium density region MDP is larger than that in the highlight portion HLP, but not so large as the shift amount in the shadow portion SDP. And. Since the shift amount in the shadow portion SDP is set to a value corresponding to the color cast amount only in the shadow portion of the original image 1, the normalized curve of this embodiment exemplified by the curves Ca and Cb in FIG. It is a curve that can remove the color cast only in the shadow portion without breaking the gray balance in the low and middle density regions.

(Iv) As described above, the shadow density value D _SX is the same as the dummy shadow density value D _DSX when the entire original image 1 is faded, not only the color cast of the shadow portion. It The normalization curve at this time is
By substituting v = w into equations (17) to (19), (2
It is expressed by the same formula as 9). That is, the normalization curve at this time is the first point Q _H and the second point Q _S (= Qsa or Q
It becomes a straight line that passes through _S b) and. This straight line when the second point Q _S is Q sa is shown as the straight line L _{0 in} FIG. The normalization conversion characteristic indicated by the straight line L ₀ suppresses the density of a specific color component in any of the high-ride portion HLP, the medium-density area MDP, and the shadow portion SDP, and secures the gray balance between the color components. To function. In other words, original picture 1
The overall fading is corrected.

Therefore, the non-linear function f used in this embodiment is an effective function for both the color cast correction of only the shadow portion and the fading of the whole of the original image 1, and is unified for these various corrections. Can be used for Even if the original image 1 has no color cast or fading, a straight line expression can be obtained from the function f by setting the shadow density value D _SX and the dummy shadow density value D _DSX to be the same. In this case, the function f is effective.

C. Second Embodiment The second embodiment of the present invention can be carried out independently of color fog correction, and the basic construction of the plate-making color scanner to which the second embodiment is applied is shown in FIGS. The scanner 100 is the same as the scanner 100 shown in FIG. 2, and only the normalization curve generated by the information processing device 14 is different from that of the first embodiment. In the second embodiment, after the highlight density value D _HX and the shadow density value D _SX (X = B, G, R) for the original image 1 to be reproduced are obtained by steps S1 to S4 in FIG. 3A, steps S5 to S7 are performed. Instead of this, step S50 in FIG. 10 is executed.

In order to execute this step S50, the highlight density value D _HFX and the shadow density value D for the standard original image are set in advance.
_{Specify SFX} . However, the standard original image refers to an original image that has empirically standardized color tones and gradations, and an original image that is in an appropriate exposure state when using a photograph or the like as the original image. pointing. Then, the highlight density value and the shadow density value for each color component, which are empirically found for such a standard original _image , are set as D _HFX and D _SFX , respectively.

In step S50, using the standard highlight density value D _HFX and the shadow density value D _SFX thus determined, the dummy shadow density value D _DSX (X = B, G, R) is calculated by the following equation (31). decide.

D _DSX = K ・ D _HX ＋ K ・ ΔD _FX ＋ (1-K) ・ D _SFX (X = B, G, R) (31) However, ΔD _FX is the following formula (32): ΔD _FX ≡D _SFX −D _HFX (X = B, G, R) (32), which corresponds to the density range width for each color component in the standard original image (hereinafter referred to as “standard density range width”). Further, K is a constant that satisfies 0 ≦ K ≦ 1 (33) and is determined in advance.

The expression (31) is proportional to the standard density range width ΔD _FX : K · ΔD _HX + (1-K) · D _SFX (34) and the highlight density value D _HX of the original image 1 to be reproduced. Value: K · D _HX (35) is added to determine the dummy shadow density value D _DSX . According to another point of view, the standard density range width ΔD _FX and the standard shadow density value D _SFX are weighted averaged using the weighting coefficient K, and the value of the equation (34) obtained thereby is given by the equation (35). _Is added to the value of to obtain the dummy shadow density value D _DSX . The advantages of determining the dummy shadow density value D _DSX by such a method will be described later.

Upon completion of step S50 in FIG. 10, step S8 in FIG. 3A.
A normalization curve is generated by, and in step S9, the main scan of the original image 1 to be reproduced is performed while performing the normalization density conversion using the normalization curve. The principle of halftone image recording is the same as that of the first embodiment.

FIG. 11 shows the two normalized curves C ₁ and C ₂ determined according to this second embodiment. Of these, the first normalized curve C ₁ is the highlight density value D _HX1 and the shadow density value D
_It shows the case where the value of _SX1 is small, and the second normalization curve C
₂ indicates a case where the highlight density value D _HX2 and the shadow density value D _SX2 are large. Ie these two normalization curves
C ₁ and C ₂ are curves for two types of original images to be reproduced having different density ranges.

Similar to the first embodiment, by using the non-linear conversion function f defined by the equations (17) to (19), the first normalized curve C ₁ can be obtained in the highlight portion HLP and the medium density range MDP. , The straight line L ₁ connecting the highlight point Q _H1 and the dummy shadow point Q _D1
It has a shape that follows. Then, the curve C ₁ passes through the shadow point Q _S1 . Similarly, the second normalization curve C ₂ shows the highlight point Q in the highlight part HLP and the medium density region MDP.
It changes along a straight line L ₂ connecting _H2 and the dummy shadow point Q _D2, and passes through the straight line L ₂ or the shadow point Q _S2 in the shadow portion SDP.

Here, [Delta] H the difference between two highlight density value D _HX1, D _{HX2 X}
Then, the following equation (36) is established.

_{ΔH X = D HX2 -D HX1 ...} (36) On the other hand, by applying respectively (31) below to the first and second normalizing curve C _1, C _2, the following equation (37), is (38) To establish.

D _DSX1 = K ・ D _HX1 ＋ K ・ ΔD _FX ＋ (1-K) ・ D _SFX … (3
_{7) D DSX2 = K · D} HX2 + K · ΔD FX + (1-K) · D SFX ... (3
8) Therefore, the two dummy shadow density values D _DSX1 , D _DSX2
When the difference of ΔD _DSX is written, the following equation (39) is obtained from the equations (37) and (38).

_{ΔD DSX ≡D DSX2 -D DSX1 = K} · (D HX2 -D HX1) ... (39) Therefore, by combining with each other and (36) and (39) below, the following equation (40) A proportional relational expression is obtained.

ΔD _DSX = K · ΔH _X (45) The following can be understood from this equation (40). That is,
The mutual distance ΔD _DSX between the dummy shadow points Q _H1 and Q _H2 in Fig. 11 is 2
It is K times the mutual distance ΔH _X between the two highlight points Q _H1 and Q _H2 , and especially when K = 1, these mutual distances ΔD _DSX and ΔD _HX
Are equal to each other. Therefore, when K = 1, the straight lines L ₁ , L
₂ are parallel to each other. This is two shadow points D _SX1 ,
This is a property that is established regardless of the mutual distance ΔS _X value of D _SX2 .

Now consider an arbitrary positive value ΔD _T that is not very large.
This value ΔD _{T is} more than the first highlight density value D _HX1.
The density value D ₁ (see Fig. 11) that is larger than the first normalization curve
The density value obtained by conversion with C ₁ is D _NT . At this time,
Since the straight lines L ₁ and L ₂ are parallel to each other as described above, the density value D ₂ which is larger than the second highlight density value D _HX2 by the same value ΔD _{T is} generated by the second normalization curve C ₂ . The density value obtained by conversion is also D _NT .

Therefore, if the normalization conversion according to this embodiment is performed on each of a plurality of original images to be reproduced having different density ranges, an appropriate reproduced image can be always obtained regardless of such a difference in the density range. In the shadow part SDP, the normalized curves C ₁ and C ₂ have their own shadow points Q _S1 and Q
_Since the condition of passing _S2 is also satisfied, the normalization conversion having a preferable property is executed over the entire density value.

By the way, when the constant K is smaller than "1", the straight lines L ₁ and L _{2 in} FIG. 11 are not completely parallel but approximately parallel.
Therefore, at this time, the density values D ₁ and D _{2 in} FIG. 11 are not completely the same but are converted to relatively close values by the normalization conversion.

FIG. 12 is a graph showing by a solid line how the dummy shadow density value D _DSX changes when the value of the constant K is changed, and is drawn using the equation (31). When K = 0, the dummy shadow density value D _DSX matches the standard shadow density value D _SFX . Since this standard shadow density value D _SFX is a constant value that is unrelated to the image of the original image 1 to be copied, a dummy shadow density value D _DSX common to all original images 1 to be copied is set. As the constant K becomes larger, the dummy shadow density value D _DSX also increases linearly, and at K = 1, as described above, it becomes the sum of the highlight density value D _{HX of the} original image 1 to be copied and the standard density range width ΔD _FX. is there. Therefore, when the coefficient K is set to "1" or a value close thereto (for example, 0.5 or more), the dummy shadow density value D _DSX changes relatively greatly according to the highlight density value D _HX of the original image 1, and the coefficient K is set to, for example, When the value is 0.5 or less, the dependency of the dummy shadow density value D _DSX on the highlight density value D _HX of the original image 1 is suppressed, and the value is close to the standard shadow density value D _SFX . Therefore, the characteristics of the normalized curves C ₁ and C ₂ can be changed by arbitrarily designating the value of the coefficient K.

By doing so, it is possible to automatically generate a normalization curve that is desirable as an empirical rule using a small number of parameters.

D. Modified Example [1] FIG. 13 shows a plate-making color scanner 400 in which a scanner body and a color separation auxiliary device are integrated. This scanner 400 is also roughly divided into an input section 401 and an output section 402, and an operation panel 403 is provided above the input section 401. Further, each circuit in the color separation auxiliary device 300 of FIG. 1 is housed in the main body of the input unit 401 in FIG. The CRT 418 and the operation panel 417 are the CRT 18 and the console 17 of FIG. 1, respectively.
Equivalent to. The function and operation of this scanner 400 are substantially the same as those of the scanner 100 of FIG. 1, but the scanner 400 is relatively compact because it is not necessary to separately provide a color separation auxiliary device.

[2] In generating the normalized curve, a non-linear function other than the above-described function f can be used. Generally, the straight line L
Let f ₁ (D) be the function that expresses f ₂ (D _HX ) = 0… (41) f ₁ (D _SX ) ＋ f ₂ (D _SX ) ＝ q _X … (42) If a function f ₁ (D) + f ₂ (D) (43) is created by using a non-linear function f ₂ (D) that is monotonically increasing or monotonically decreasing, a function according to the requirements of the present invention is obtained. .

Note that (17) of the function f, (43) f ₁ (D) In the equation = p- (p-q) ( D-u) / (v-u) ...
(44) f ₂ (D) = (p−q) log ₁₀ [1 + g ₂ / g ₁ ] (45) However, g ₁ and g ₂ are (18) and (1
It is a function defined by the equation 9).

[3] In the above embodiment, the normalization conversion is performed digitally using the LUTs 5B, 5G, 5R, but it may be performed analogically using a function generator.

[4] Highlight density value D _HX and shadow density value D _SX
Is preferably determined through a statistical method using a histogram or the like as in the above embodiment, but when a highlight point and a shadow point are found in the original image 1, the operator designates them to specify the color density. _May be measured and the measured values may be adopted as D _HX and D _SX .

[5] When generating the normalized curve according to the present invention, correction may be performed according to the subject of the original image to be processed. Such amendment is described, for example, in Japanese Patent Application No. 1-308485 filed by the applicant of the present invention. For example, highlight density correction amount Δ
d _X (X = B, G, R) is determined in advance, and the highlight density value D _HX is shifted by this correction amount Δd _X as a new highlight density value D _HCX as shown in FIG. Good. In this way, the normalization curve C without correction is
₁₀ becomes the curve C ₁₁ after the correction.

[6] Although the first implementation order described above is applied only to color original images, the second embodiment can be applied to both color original images and monochrome original images. In the case of a monochrome original image, each of the highlight density value, the shadow density value, and the dummy shadow density value is determined, and only one normalization curve is determined.

[8] The present invention is not limited to the plate-making color scanner, but can be applied to all image processing apparatuses that perform color separation and other processing on an original image.

〔The invention's effect〕

As described above, according to the first configuration of the present invention,
In addition to the highlight density value and the shadow density value, the dummy shadow density value is determined, and the shape of the normalization curve in the highlight area and the middle density area is determined by the highlight density value and the dummy shadow density value. The shadow density value is reflected only in the shape at. Therefore, the shape of the normalization curve in each concentration range can be managed in a unified manner with a relatively small number of parameters, and a skilled operator is not required. Therefore, the method is suitable for automation (corresponding to the first purpose).

In the second configuration of the present invention, the highlight density value and the shadow density value for each color component are given, and the dummy shadow density value according to the color cast amount of only the shadow portion is given, so that only the shadow portion is colored. It is possible to systematically generate a normalization curve suitable for fogging removal. Since the normalization curve is generated based on the values that can be objectively grasped as described above, the correction amount of color cast can be quantitatively managed (corresponding to the second purpose).

Also, since color cast correction is performed during normalization conversion,
The setting of the processing characteristics in other circuits in the color separation device is performed on the signal after the color cast correction. Therefore, regardless of whether or not the original image has a color cast, these circuits can determine the processing characteristics as if the original image has no color cast (corresponding to the third purpose).

Further, according to the third configuration of the present invention, only by changing the parameter of the predetermined function, it is possible to handle the color cast removal only in the shadow portion and the correction of the overall fading of the original image in a unified manner. Therefore, the present invention is a method particularly suitable for automating the color separation condition (corresponding to the fourth object).

Further, according to the fourth and fifth configurations of the present invention, since the dummy shadow value is determined while reflecting the value of the density range width of the standard original image, even if the density range of the original image to be processed changes, the highlight part In the medium density range, density values having the same difference from the highlight density value can be converted into density values that are substantially the same or relatively close by normalization. The methods according to the fourth and fifth configurations can be applied to both a color original image and a monochrome original image regardless of whether or not they are for color cast correction, and can be applied to the density range of the original image to be processed. Regardless, it is possible to always ensure proper density expression in the image after the normalization conversion (5th example).
Corresponding to the purpose of).

In particular, in the fifth configuration, since the dependency of the dummy shadow density value on the standard density range width can be changed arbitrarily by changing the weighting coefficient in the weighted average, its versatility is also high.

[Brief description of drawings]

FIG. 1 is an external view of a plate-making color scanner to which an embodiment of the present invention can be applied, FIG. 2 is a block diagram of the plate-making color scanner of the embodiment, and FIGS. 3A and 3B are first diagrams. FIG. 4 is a flow chart showing the operation of the embodiment, FIG. 4 is a view conceptually showing the data processing means in the first embodiment, FIG. 5 is a view showing an average density value frequency histogram, and FIG. 6 is a color component. FIG. 7 is a diagram showing a cumulative density histogram for each image, FIG. 7 is a diagram showing a cumulative relative frequency histogram, and FIG. 8 is an explanatory diagram of a process of calculating an average color density value from a part of the histogram in FIG. Figure is a graph showing an example of a normalization curve generated by the first embodiment, Figure 10 is a flow chart showing part of the process of the second embodiment of the present invention, Figure 11 is a second An example of a normalization curve generated by the example of FIG. 12 is a graph showing the dependency of the dummy shadow density value on the weighting coefficient of the weighted average, and FIG. 13 is an external view showing another example of a plate-making color scanner to which the present invention can be applied. The figure is a diagram exemplifying correction of highlight density values according to the subject of the original image to be processed. D _X ... Density value for each color component, D _NX ... Normalized density value for each color component, D _HX ... Highlight density value for each color component, D _SX ... Shadow density value for each color component, D _DSX …… Dummy shadow density value for each color component, Q _H …… First point, Q _S …… Second point, Q _D …… Third point, Ca, Cb, C ₁ , C ₂ … Normalized curve D _HFX …… Standard highlight density value, D _SFX …… Standard shadow density value, ΔD _FX …… Standard density range width, K …… Coefficient in weighted average

Claims (5)

[Claims] 1. A method for generating a normalized curve for converting a density value of each pixel of an image of an original image to be processed into a value within a normalized density range suitable for image processing in a predetermined image processing apparatus. (A) For the original image to be processed, a highlight density value and a shadow density value are determined from the range of the density values before conversion, and (b) the conversion is performed according to a predetermined determination rule. A dummy shadow density value for the previous density value is determined, (c) as the normalization curve, (c-1) on a coordinate plane on which the normalization curve should be defined,
A first condition of substantially passing through both a first point determined according to the highlight density value and a second point determined according to the shadow density value; (c-2) the highlight In the density range near the density value, the second condition that it changes along a straight line passing through both the first point and the third point on the coordinate plane determined according to the dummy shadow density value A method for generating a normalized curve, characterized in that a curve that satisfies both of (1) and (2) is generated. 2. When a color original image is color-separated for each pixel using a predetermined color separation device, a density value for each color component of each pixel is normalized to a normalized density range suitable for color separation in the color separation device. The method according to claim 1 is applied to a process for generating a normalization curve for each color component for conversion into a value within, each of a highlight density value, a shadow density value, a dummy shadow density value and a normalization curve. Is determined for each color component, and (b-1) requires (b-1) color cast correction of only the shadow portion of the color original image, the shadow density is increased by a value corresponding to the color cast amount. A value deviating from each value is obtained for each color component, and these are set as the dummy shadow density values for each color component, and when the color cast correction of only the shadow portion is not required, the shadow density value Containing substantially step of the same value as the dummy shadow density value, normalizing curve generation method. 3. The method according to claim 2, wherein (d) in order to generate a normalized curve, (d-1) has a highlight density value, a shadow density value and a dummy shadow density value as parameters, and before conversion. As a variable, (d-2) when the shadow density value and the dummy shadow density value are different values, the variable is non-linear, and (d-3) the shadow density value and the shadow density value When the dummy shadow density value is the same value, a linear function with respect to the above variables is determined in advance. (E) The normalized curve is (e-1) The color original image is corrected for color cast only in the shadow portion. (E-2) When it is necessary to correct the color cast of only the shadow portion, after determining whether or not the shadow density value and the dummy shadow are required. (E-3) When the color cast correction of only the shadow portion is not required, the shadow density value and the dummy shadow value are set to different values. A method for generating a normalized curve, which is determined by setting the same values to each other and specifying the parameters of the function. 4. The method according to claim 1, wherein the step (b) comprises: (b-1) a value proportional to the density range width of the standard original image;
A method for generating a normalized curve, comprising the step of adding a value proportional to the highlight density value to a value obtained as a dummy shadow density value. 5. The method according to claim 4, wherein the dummy shadow density value is the weighted average value of the shadow density value of the standard original image and the density range width determined in step (a). A method for generating a normalized curve, which is determined as a value added to a value proportional to.