JPH0240778A - Color converter - Google Patents

Abstract

PURPOSE:To stably specify a color even against a halftone document by smoothing a color signal which is brought to sampling and bringing an area of microscopically discontinuous colors to dummy control to an area of continuous colors. CONSTITUTION:A scanner 1 reads a document by bringing it to color separation by R, G and B, and a minute area for specifying a color (extraction color) which goes to symmetry of a color conversion and a color (substitution color) reproduced instead thereof is designated by using a digitizer 4 or ten keys of an operating part 5, etc. R, G and B data which are inputted to an image processing part 2 are brought to smoothing processing in a smoothing circuit 21 in order to eliminate a halftone frequency component, and thereafter, converted to an HSL (hue, saturation, lightness) color specification system in a first color specification system converting circuit 22. An extraction color designating circuit 23 and a substitution color designating circuit 24 generate data H1, S1 and L1, and H2, S2 and L2 of the extraction color and the substitution color which are specified, respectively, and a color converting circuit 26 converts them to substitution color signals corresponding to extraction color signals. In such a way, the color can be specified stably even against a halftone document.

Description

DETAILED DESCRIPTION OF THE INVENTION ■Field of Industrial Application The present invention relates to a color conversion device that replaces a specific reproduced color with another reproduced color or shifts that color. The present invention relates to a technique for reproducing only an image expressed in a specific color on a document by replacing it with a different color, or reproducing the image by shifting its color.

■Prior art If it is possible to replace only an image expressed in a specific color on a manuscript with a different color, or reproduce it by shifting the color, for example, it is possible to change the ``red flower'' displayed on the manuscript to `` It can be converted into ``blue flowers'' or ``yellowish red flowers'', which is often convenient.

In this case, the RGB data [Re
d, Green, and Blue].

Conversely, it is difficult for humans to imagine colors expressed as RGB data.

Therefore, the applicant specified an area within the manuscript by inputting coordinates, and based on the colors within that area, the color to be converted (hereinafter referred to as "extracted color") and the color to be replaced (hereinafter referred to as "replacement color").
We proposed a technology to identify the technology and filed an application first (Patent application 1986-2).
No. 01196), in this case, for each specified color, the variation within the specified area is set as an allowable range.

According to this, since the extracted color and the replacement color can be specified by looking at the colors appearing on the document, the colors recognized by humans and the colors recognized within the device match.

■HM that the invention seeks to solve It is true that this method has the desired effect on density gradation images such as silver halide photographs, but when the sampling density of the input system increases, it was found to be unable to identify colors properly.

The reason for this will be explained with reference to FIG.

Currently, the full-color printed matter commonly used is 1
It is composed of halftone dots of about 50 to 175 lines (150 to 175 dots/1ch). On the other hand, the input system of a digital copying machine employs a sampling density of about 300 to 400 dots/1 nch. Figure 11 shows the relationship between these halftone dots and the sampling grid.
・ is the halftone dot of M agent (M), 0 is Yel
The low (Y) cutting points and the grids indicate the pixels to be sampled, respectively. For convenience of explanation, each pixel is numbered 1 to 14. Number a-Ω, (1=aL (82g
), etc., specify the coordinates of each pixel.

This document is drawn with 150 to 175 lines of halftone dots, and when viewed with the ordinary eye, the halftone dots are barely noticeable and are recognized as a continuous tone, pale red image. However, 300-40
If this is observed in sampling at a density of 0 dots/1 nch, it cannot be said that the ink is uniformly distributed.

For example, pixels designated by coordinates (6, b) and (2, f) both have approximately equal amounts of Y and M, making them red, but their brightness is very different. Also the coordinates (
The pixels specified by 4, e) have 7 colors.

The pixel specified by the coordinates (4, f) is in the M color category.

In this image, there are pixels with a wide range of hues MRY.

In other words, in a halftone image, even if it is a macroscopically continuous area, it is a microscopically discontinuous area, so there is a drawback that the color specified by the coordinate specified pixel becomes unstable. .

The present invention aims to solve the above-mentioned problems and provide a color conversion device that can stably and appropriately specify colors even for halftone originals, and improves the applicability of the color conversion device. The purpose is to

■Means for Solving the Problems The present invention focuses on the fact that the above-mentioned problems occur because the sampling aperture is too small relative to the dot density. Therefore, in the present invention, in order to achieve the above object, a color conversion device that performs the above-mentioned color replacement or shift is configured to generate color signals of multiple components in correspondence with each micro-area into which the original image is micro-divided. Color signal generation means; minute area specifying means for specifying arbitrary minute areas from the original image divided into minute areas; color signals associated with each of the minute areas of interest and minute areas in a predetermined vicinity thereof, for each component; a smoothed signal generating means for generating a smoothed signal in association with the minute region of interest;
The apparatus includes a signal generating means for generating a signal specifying at least one component from the smoothed signal.

In this case, smoothing means that each color signal of a minute area in a predetermined vicinity of the minute area of interest (for example, a minute area included in an area expressed by 3X3, 5X5 with the minute area of interest as the center) is Processing that simply averages each component, or weighting and smoothing processing (this involves converting the color signal of the minute area of interest into the color signal of the minute area value in a predetermined neighborhood, smoothed for each component. (including replacement processing), etc.

■ Effect According to this method, by smoothing the sampled color signal, a microscopically discontinuous color area is transformed into a continuous color area, so it is stable even for halftone originals. As a result, colors can be appropriately specified, and the applicability of the color conversion device is improved.

Other objects and features of the present invention will become apparent from the following description of embodiments with reference to the drawings.

(1) Embodiment (1) First Embodiment FIG. 1a is a block diagram of a color image recording apparatus embodying the present invention. This device has scanner 1
1 image processing section 2. Printer 3. Digitizer for area specification 4. It is comprised of an operation section 5 for setting processing modes and parameters for processing, displaying set values, etc., and a system controller 6 for controlling each of these elements.

The image processing unit 2 also includes a smoothing circuit 21 . First color system conversion circuit 22. Extracted color designation circuit 23. Replacement color designation circuit 24
9 color recognition circuits 25 2 color conversion circuits 26° 2nd color system conversion circuit 27. UCR color correction circuit 28 and gradation processing circuit 29
It is equipped with

First, an overview will be explained.

When the operator selects a color conversion mode from the operation panel, area data is used to specify the color to be converted (hereinafter referred to as extracted color) and the color to be reproduced in its place (hereinafter referred to as replacement color). It becomes an input holding state. During this time, the operator inputs area data (coordinate data) using the digitizer 4 or the numeric keys of the operation unit 5 according to the message displayed on the operation unit 5, and also determines the range to be included in each color (hereinafter referred to as the allowable range). input.

When the copy start button is pressed after inputting area data,
First, the printer 3 enters the pre-scan mode and does not operate. In this mode, scanner 1 scans the original in R (red),
The color is separated into G (green) and B (blue) and read, converted into 6-bit digital data for each color, and sequentially sent to the image processing section 2.

The KGB data input to the image processing section 2 is processed by the smoothing circuit 2.
1, the first color system conversion circuit 22 converts the HSL color system [H: Hue (hue).

S: 5ajration (saturation L L: Li
lightness (brightness)]. I will add an explanation to this.

First, it should be understood that this color system conversion is a process to facilitate color identification. Namely.

Conventionally, when performing color conversion, the extracted color (Rt#G1eB1) specified in the RGB space is converted into the replacement color (R29G2tB2) specified in the same RGB space.
) was common. However, in practical terms, for example, if you want to change a "red flower" to a "blue flower," in the case of a full-color image, the abbreviation for "red" that should be specified as the extracted color is "bright red" or ""darkred" is also included, and similarly, "blue", which should be specified as a replacement color, also includes "bright blue" and "dark blue". In other words,
In many cases, the operator specifies a color without considering the brightness of the three color elements, so one color is recognized as being specified by the operator and the other is recognized as being specified by the equipment. There will be a discrepancy between the colors. Therefore, in the apparatus of this embodiment, a color system conversion process is performed to separate the color into three elements, each of which can be selected independently. So even if the operator specifies the color without having lightness in mind.

If color designation using hue and saturation components, which are the three elements excluding the lightness component, is selected, the above discrepancy will almost disappear.

Additionally, apart from this, a color such as "red" may be specified. In this case, after specifying one color, you can specify the range (tolerable range) to be included in that system, but in KGB space, not only the allowable range of R2O, B, but also the range between R, G, and 8. It is difficult to specify this tolerance range because balance must also be taken into consideration. However, if the color system conversion process is performed and the color system is separated into three elements as in this example, this kind of problem can be achieved by selecting the color specification based on only one color composition and specifying one color and the allowable range. Color can be specified.

In this way, color specification is facilitated by converting the color system from the RGB system to the H8L system and handling color images in the H8L system. Here, H9S,L is L''=116(Y/Yo)v3-16
...(2.1)a" = 500((X/Xo
)' to (Y/Yo)/3) "(24)b"
= 200・((X/Xo)1'3 (Z/
Zo ``) ・---・When (2.3), H=jan'a' / b'
...'(34)S=(a"+b"")l/2・---
(3-2) L=L''...
... shall be defined as (3.3).

However, R, G, and B refer to the CIE RGB color system.

, and xo , Y, , Z are the values of X, Y, and Z on the perfect diffuse reflection surface.

The H3L data converted and generated from the KGB data in the first color system conversion circuit 22 is converted to the extracted color designation circuit 23. Replacement color designation circuit 242 color recognition circuit 25 and color conversion circuit 26
given to.

In the extraction color designation circuit 23, when the coordinate data of the pixel of interest matches the coordinate data of the extraction color specific area input by the operator, the H8L data of the pixel of interest is stored in the circuit as extraction color data (Hly St e Lt). When the coordinate data of the pixel of interest matches the coordinate data of the extraction color specific area input by the operator, the replacement color designation circuit 24 stores the H8L of the pixel of interest in the replacement color designation circuit 24.
The data is stored in a storage device within the circuit as replacement color data (H2t S2 e L2).

This concludes the processing in pre-scan mode.

Color recognition circuit 251 Color conversion circuit 26. Second color system conversion circuit 27. The UCR color correction circuit 28 and the gradation processing circuit 29 do not operate.

When the pre-scan is completed, the main scan mode is entered, and the scanner 1 starts reading again.

In this case, as in the case of pre-scanning, the image data separated into KGB is smoothed in the smoothing circuit 21, converted to HSL color in the first color system conversion circuit 22, and the converted data ()ISL data) is extracted by the color specifying circuit 23. [Color change designation circuit 249 is applied to the color recognition circuit 25 and color conversion circuit 26. However, in this case, the extraction color designation circuit 23 and the replacement color designation circuit 24 do not carry out the operation of specifying the color described above, but instead use the extraction color data (Hl) stored in the prescan mode.

St*Lt) is output to the color recognition circuit 25, or replacement color data (Hl w S2 t L2) is output to the color conversion circuit 26.

In the color recognition circuit 25, the H5L data of the pixel of interest inputted in synchronization with the pixel clock and the extracted color designation circuit 2
Extracted color data (Illy St e
Lt), and when they match within the tolerance specified by the operator, the extracted signal is '1''.
and if they do not match, output "jJo".

In the color conversion circuit 26, when the extraction signal is "1", the replacement color data (H
l, S2 w L2), and when the extraction signal is “0”, H8L given from the first color system conversion circuit 22
Output data. That is, in this color conversion circuit 26, extracted color data (Hl) corresponding to the extracted color specified by the operator.

5xyLt) is the replacement color data (Hl
t S2 e L2 ).

Data conversion by this color conversion circuit 26 is independent for H9S and L, and the selection depends on the operator's designation. That is, for the reasons mentioned above, the operator performs "color conversion by hue Z1g color conversion by saturation 1', L
I Color conversion by brightness n, u Color conversion by hue and saturation", ``Color conversion by saturation and brightness'', ``Color conversion by brightness and hue'', or 6 Color conversion by hue, saturation and brightness ” can be specified arbitrarily. Therefore, for example, when specifying "color conversion using one hue," only the H component is converted, and H1-ΔH≦H≦H1+ΔH(ΔH
A pixel having H data (hue data) (within a permissible range) is extracted, and its H data is converted to Hl.

The HSL data that has undergone color conversion processing is inversely converted from the H8L color system to the RGB color system in the second color system conversion circuit 27, and then subjected to UCR processing in the UCR color correction circuit 28. Ru. This UCR processing is performed by printer 3.
This process corrects the deviation from the ideal spectral characteristics of the coloring material used in the process, calculates the amounts of Y, M, and C, and generates the BK acid component.

The data (Y,
M, C, BK) are binarized by the systematic dither method in the gradation processing circuit 29 and sent to the printer 3. The printer 3 prints Y, M, C and BK using these data.
Color superimposition recording is performed to reproduce the color-converted image.

Next, each processing block will be explained in detail.

Smoothing processing in the smoothing processing circuit 21 is performed for each KGB data, and is performed by selecting one of the digital filters shown in FIGS. 3a to 3d according to the operator's designation. Here, an example of a circuit of a digital filter consisting of a 3×3 pixel matrix will be described (a portion related to one piece of data).

The 3 x 3 pixel matrix is used to simultaneously extract nine pixel data arranged two-dimensionally in the main scanning direction and sub-scanning direction as shown in Figure 2a, and each pixel in the matrix is
When referred to by the numbers shown in Figure a, ■ is the pixel of interest.

This pixel matrix is as shown in Figure 2b.

It is composed of line buffers LBI and LB2 for two lines and data latches L1 to L9 for nine pixels. When the data of the pixel of interest and its 8 neighboring pixels are extracted here, a smoothing operation is performed in the circuit shown in FIG. 2c.

The circuit that performs the smoothing operation includes eight adders Add 1 to
Add8 and four multipliers Mull to Mu14.

Pixel ■ in adders Addl, Add2 and Add5.

The data of ■, ■, and ■ are added, and the 1 multiplier Mull adds the coefficient ■ (set to 0=1 for the filter in Figure 3a, and ■=2 for the filter in Figure 3b) to the addition result.
). Also, the adder Add3. Ad
In d4 and Add6, pixels ■, ■.

The data of ■, and ■ are added, and the multiplier Mu12 adds the coefficient ■ (Fig.
For the filter in the figure, set ■=1). The outputs of multipliers Hull and Mu12 are added in adder Add7.

In this embodiment, by configuring a set of data to be multiplied in this way, a filter that weights data of pixels having the same distance from the pixel of interest as shown in FIGS. 3a and 3b is created. The number of multiplications when using one is reduced.

On the other hand, the data of the pixel of interest ■ is sent to the multiplier Mu13 with a coefficient O (in the filter of FIG. 3a, ■=1 is set, and the third
In the filter in figure b, the output data of the adder Add7 and the output data of the multiplier Mu13 are added together in the adder Add8, and the result of the addition is multiplied by the coefficient ■( For the filter of FIG. 3a, set ■=179; for the filter of FIG. 3b, set ■=1/16). This output data becomes the data of the pixel of interest that has been subjected to the smoothing process.

The above circuit expanded to 5.times.5 pixels becomes the circuit for performing the smoothing process using the filter shown in FIG. 3C or FIG. 3D, but its explanation will be omitted here.

Further, this embodiment includes a circuit that performs smoothing processing using a 3x3 pixel matrix and a circuit that performs smoothing processing using a 5x5 pixel matrix for each RGB data.

Next, the first color system conversion circuit 22 will be explained with reference to FIG. 4a.

This circuit includes nine multipliers Mull to Mu133 and three adders A for performing the calculation of equation (1) described above.
dd9 to AddlO, the above-mentioned (2.1) to (2.
3) Three memories ROM 1 to 1 for calculating formulas
ROM3, and three memories ROM4 to ROM6 for performing calculations of equations (3.1) to (3.3).

Multiplier Mu111. Mu112. Mu113. Mu12
1. Mu122. Mu123°Mu131. Mu132
and Mu133, adder Add9. AddlO and Addll convert data R, G, B to data X, Y.

Converting to Z. In this case, in the above-mentioned equation (1), the coefficient matrix A is set assuming the RGB color system defined by CIE, so the spectral characteristics of the color separation filter of the scanner l, illumination system, sensor, etc. Coefficient matrix C (matrix element activation cij
) is used. This coefficient matrix C is set as follows.

First, the data R, G, and B obtained from the scanner are
Data R', G', and B' are obtained by correcting to the CIE RGB color system using the following correction formula.

However, . Data R', G'B obtained by correction at this time
' is the CIE RGB color system, so the above (1)
It can be converted into data X, Y, and Z using the formula.

Here, by setting AB'-C, a coefficient matrix C can be obtained.

As described above, the data R9G, B obtained from the scanner is corrected to the CIE RGB color system, and the data x, y,
The calculation to convert to z is realized in one step.

Memories ROMI to ROM3 refer to pre-stored arithmetic tables to calculate data L" a from data X, Y and Z.
``and b'' are generated, and the memories ROM4 to ROM6 refer to pre-stored calculation tables to generate data L'' a
data L, S, and I] are generated from "and b'. In this embodiment, the data L, S, and I are generated from
Since the conversions using equations 2.3) and equations (3.1) to (3.3) are nonlinear operations, the ROM table was used in this way, but these conversions are performed using only one or Since the converted element is determined from two elements, a relatively small capacity ROM is sufficient.

Here, each element has an 8-bit precision.

For example, the memory ROM2 that generates data a'' from data X and Y has 64KwordX 8bit=512K
A bit ROM is used.

FIG. 5 shows the configuration of the extracted color designation circuit 23. As shown in FIG.

This circuit consists of storage devices Lll, LL2, 13, comparator Cmp1, and the like.

The storage bags [Ll and L2 are stored by the operator on the digitizer 4 or the operation unit 5 under the control of the system controller 6.
The coordinate data (pixel address) x 1-addr and y 1-addr of the extracted color designated area input from are stored. The comparator Cmpl compares this address given to the comparison input port P with the addresses x-addr and y-addr of the pixel of interest based on the pixel clock and line clock counts given to the comparison input port Q, and writes 1 if both match. Outputs a control signal.

In pre-scan mode, system controller 6
Since the pre-scan mode signal is given as ``1'', this write control signal is applied to the AND gate ANDI.
is applied to the memory bag ML3 via the memory bag ML3.

At this time, the H8L data of the pixel of interest is input to the storage device L3, so is the data stored in the storage device? The extracted color data (Ht t S 1 t Ll) is stored in ZL3. In addition, when scanning one line, the system controller 6
Since the pre-scan mode signal is given as "0", the data in the storage device L3 is not updated.

The replacement color designation circuit 24 has the same configuration as the extraction color designation circuit 23 described above, so a description thereof will be omitted here.

Please refer to FIG. 6a. As shown in this figure, the color recognition circuit 25 includes a subtracter 5ubll and an adder Add1.
2. It consists of three blocks including comparators CmpH, Cmp12 and an AND gate ANDII, and an AND gate AND12. Here, a block that processes the H component will be described.

The subtractor 5ubll and the adder Add 12 are each supplied with extracted color data H1 from an extracted color designation circuit 23 and tolerance range data ΔH designated by the operator from the system controller 6. The subtracter 5ub11 performs subtraction using these data to obtain data (Hl-ΔH
) is given to comparator C+spH.

The adder Add12 performs addition using these data, and the data (H1+ΔH) is sent to the comparator Cmp12.
give to

Comparator CmpH compares data H and data (T11-ΔH) of the pixel of interest, and if the former is larger, It1
1#, and if the latter is larger, it outputs "0".Similarly, the comparator Cmp12 compares the data H of the pixel of interest with the data (H1+ΔH), and if the former is smaller, it outputs "1".
``'', and if the latter is smaller, outputs "O".The outputs of these comparators CmpH and Cmp12 are combined in an AND gate ANDII.In other words, the subtracter 5ubll, the adder Add12, and the comparator Cm
pH, Cmp12, and the AND gate ANDII constitute a window comparator whose center value and width are variable.

Similar processing is performed on the S component and the L component, and each output is applied to the AND gate AND12. AND GATE AN
D12 outputs the extraction signal as '1' if all of the data H, S and L match the extraction color data H1tS1 and Ll within the allowable range, and outputs the extraction signal as 'O' if any one of them differs. .

In this case, the H component is used when specifying the extracted color.

When excluding the S component or the L component, the maximum value is set to the allowable data.

As shown in FIG. 7, the color conversion circuit 26 includes three AND gates ANI to AN3 and three selectors 5ell to
Consists of 5e13.

AND gate ANI is H from system controller 6.
When the conversion signal is “1”, the extraction signal from the color recognition circuit 25 is “1”.
Outputs '1' only when the signal is 1'', and the AND gate AN2
outputs “1” only when the S conversion signal from the system controller 6 is “1” and the extraction signal from the color recognition circuit 25 is “I11”, and the AND gate AN3 outputs “1” only when the L conversion signal from the system controller 6 is “1”. "1" is output only when the extraction signal from the color recognition circuit 25 is "1".These outputs are input to the selection control terminals of the respective selectors.

When the input to the selection control terminal is 0'', the selector 5ell selects and outputs the [■] data from the first color system conversion circuit 22, and when the input to the selection control terminal is ## 1 pg, it selects and outputs the data from the replacement color designation circuit 24. Similarly, the selector 5e12 selects the S data from the first color system conversion circuit 22 and the S2 data from the replacement color designation circuit 24 in accordance with the input from the selection control terminal. and output
The selector 5e13 converts the L data from the first color system conversion circuit 22 and the replacement color designation circuit 24 according to the input from the selection control terminal.
The selected L2 data is selected and output.

Therefore, for example, the L-converted signal and the S-converted signal are
, H conversion signal is set to 1, only the hue of the pixel to be color-converted is converted to the hue (H2) of the replacement color, and the L and S components are preserved as original signals.

The second color system conversion circuit 27 converts the H8L system data H'S', L' selectively converted in the single color conversion circuit 26.
As mentioned above, this is a circuit for inversely converting the data into RGB data, but since this configuration is almost the same as the first color system conversion circuit 22, the explanation here will be omitted.

Further, as for the UCR color correction circuit 28 and the gradation processing circuit 29, well-known ones can be used, so a detailed description thereof will not be given here.

(2) Modified Embodiments of Each Processing Block The ones shown in FIGS. 3e to 3j show other examples of digital filters. Since these digital filters differ only in coefficients, they can be used as they are in the circuit described above.

FIG. 4b shows a modified example in which the first color system conversion circuit 22 is configured by memories ROM7 to ROM8 in which calculation tables are stored. Here, 6-bit RGB data is used. 256 K words x 8 bits
The RGB data read by the scanner 1 is converted into H8L in one stage using three 2 Mbit memories.

Of course, the same functionality can be obtained even with a smaller capacity memory by using more than this. Furthermore, in general, when specifying a color, if consideration is given to setting a tolerance range, there will be no practical problem even if the accuracy of each component before and after color system conversion is reduced. However, as mentioned above, it is considered that the brightness component is often saved before and after color conversion processing, so it is desirable to maintain high accuracy.

FIG. 4c shows an example of the configuration of a table-reference type color system conversion circuit that takes the above into account. Here, the color system conversion process is performed using the upper 5 bits of the 6-bit data for each of R, G, and B. However, regarding the L component, data R and G for it.

Considering the contribution rate of B, select data G with the highest contribution rate by 6
The configuration is such that the L component can be obtained with 6-bit precision using a total of 15 bits of data, with data R and B being 5 bits and 4 bits, respectively, in order of contribution rate. In this way, by considering the contribution rate, it becomes possible to perform conversion using less input data without reducing accuracy. In addition, in this example, data H, S, and L are each obtained from 15-bit RGB data, so 32K
vordX 8bi7 configuration memory ROM10. R.O.
MII and ROM12 are used.

Note that even when the color system conversion circuit 22 is configured using an arithmetic circuit as in the embodiment described above, by considering the contribution rate, it is possible to simplify the hardware while maintaining the accuracy after conversion. is possible.

FIG. 6b shows the addition, subtraction and comparison described above in memory Ro.
T'! 21. A modification of the color recognition circuit 25 using the ROM 22 and the ROM 23 is shown. H2 of these memories
t11 and Δ■] or the area addressed by ts*s1 and ΔS or L.

In the area addressed by Ll and ΔL, 1 (match) when the above conditions are satisfied, 0 (mismatch) otherwise.
is memorized. In this case, each memory has multiple ΔH
The table that stores the match/mismatch determination results when 9ΔS or ΔL is set, and refers to the table corresponding to the specified ΔH.

FIG. 6c shows one color recognition circuit 25 in one word 8 bi
This is a modification example configured using a memory with a t configuration. The H component will be explained.

Each bit of the area (8 bits) addressed by H and Hl of the memory ROM 31 is associated with ΔHo to ΔH7, and the determination result for each ΔH is stored.

Outputs 60 to 67 of this memory output judgment results for eight types of ΔH at the same time.
One of them is selected by a 4-bit select signal 5elH corresponding to the operator's designation.

In this case, if the H component is to be removed when specifying the extraction color, the non-reference signal, which is the least significant bit of the select signal 5elH, is set to 1411H and the OR gate OR1 to 1'1 is set to 1411H, regardless of the values of H and Hl. (match)" is output, so the H component is not referenced. This function can be performed by the memory POM31, but there are seven types of determination data.

The block for determining the S component and the L component has a similar configuration, and II I H is outputted as an extraction signal from the AND gate AND31 only when the determination data of all components match.

(3) Second Embodiment FIG. 1b shows an example of the configuration of the image processing section 2 of the second embodiment.

In this device, a "character area" and a "photograph/halftone image area" are distinguished by smoothing processing and gradation processing. This was added in view of the problem that the sharpness of character images deteriorates due to the smoothing process being performed in the first embodiment apparatus described above regardless of the type of input image (character image, photo image, halftone image). This is a function that eliminates this type of problem by not applying smoothing processing to the data in the character area, but applying gradation processing to the data with an emphasis on resolution performance.

The details of the area determination unit 2B are detailed in Japanese Patent Application Laid-Open No. 62-252357, which was filed by the present applicant, and will not be described here, but the character area and the halftone dot area are separated. The separated signal is given to selectors 2A and 2C.

Selector 2A selects smoothed data and non-smoothed data using this separation signal, and selector 2C
The first gradation processing circuit 291 and the second gradation processing circuit 292
The output of is selected by the separated signal. The first gradation processing circuit 291 is a processing circuit for picture areas with emphasis on gradation characteristics, and the second gradation processing circuit 292 is a processing circuit for character areas with emphasis on resolution performance. Therefore, the character image is not subjected to smoothing processing, but is subjected to processing that emphasizes resolution, so that a sharp output image can be obtained. Further, in the picture area (photo image and voiced-tone area), smoothing processing is performed and processing that emphasizes gradation is performed, so that an output image with smooth gradation can be obtained.

(4) Third Embodiment FIG. 1c shows an example of the configuration of the image processing section 2 of the third embodiment.

This device is characterized in that when the data of the pixel of interest is not subjected to color conversion, KGB data without color system conversion processing is used. According to this, when the calculation accuracy of the first color system conversion circuit 22 and the second color system conversion circuit 27 is low, data of non-converted pixels is disturbed by the color system conversion process and its inverse conversion process. This will be prevented.

In this embodiment, since the extraction signal is not input to the color conversion circuit 26, color conversion processing is performed on the data of all pixels, but there is an intermediary between the second color system conversion circuit 27 and the UCR color correction circuit 28. In the inserted selector 2D, data R, G, B from the selector 2A and the second
Data R', G', and B' from the color system conversion circuit 27 are selected.

(5) Fourth Embodiment FIG. 1d shows an example of the configuration of the image processing section 2 of the fourth embodiment.

This device is configured so that three extraction designated colors and three replacement designated colors can be designated. That is, the extraction color designation circuit 23 of this embodiment device includes three extraction color designation circuits 23 having the configuration shown in the first embodiment, and the replacement color designation circuit 23.
4 includes an extraction color designation circuit 24 having the configuration shown in the first embodiment.
Contains three sets.

In this case, the operator associates the first extracted color with the first replacement color, the second extracted color with the second replacement color, and the third extracted color with the third replacement color, and uses the digitizer 4 or the numeric keypad of the operation unit 5 to Specify from.

As shown in FIG. 8, the color recognition circuits 25 each have a sixth
It has three sets of color recognition circuits 251 to 253 configured as shown in FIG.
5ISO". 11 S, So+# outputs "oo" when it does not match any of the three colors, "01'" when it matches the first extracted color, and "10" when it matches the second extracted color.
'', and when it matches the third extracted color, it becomes "11'". However, if it matches multiple extracted colors, the extracted color specified later is given priority. Therefore, for example, if it matches the first extracted color, When the third extracted color matches, "11" is output giving priority to the matching with the third extracted color.

The color conversion circuit 26 of this fourth embodiment receives the color recognition signal It S
, the first ff color change (Hals4
#L4 L first replacement color (Hs, Ss, Ls
), the third substitution (Hs t SS p Ls ), or the non-substitution (H, S, L,) is selected and output. FIG. 9 shows a color conversion circuit 26 constructed using a multiplexer MPX that selects one type from four types of input. This multiplex MPX uses data H when the color recognition signal is 'oo'.
,S.

When L is atOLpp, data H4t S4 t L
When m is "10", data H5*SS w L5 is selected, and when ##11jl, data H6v ss l L6 is selected and output.

In this case, the configuration may be such that conversion of the H component, S component, or L component is selected in the same manner as described above.

(6) Fifth Embodiment FIG. 10 shows a color correction circuit 2E that replaces the color conversion circuit 26 in the apparatuses of the first to fifth embodiments described above.

This color correction circuit 2E is a circuit that shifts the hue, saturation, and/or brightness of a specific color, and by replacing the color conversion circuit 26 of each embodiment device described above with this circuit, for example, [Red Flower It is possible to make a relative color shift, such as making the red of "" a little yellower. In this case, the above-mentioned extraction color designation circuit 23 and color recognition circuit 24 are used to designate and extract the color to be shifted, but the replacement color is not designated and instead the shift data (ΔH,
ΔS, ΔL) are input using the numeric keypad or the like on the operation unit.

The data of each component and the shift data are sent to the adder Add21.
．． They are added in Add22 or Add23, and selector 5e121.5e122 or 5e123 selects and outputs unshifted data H, S, L or shifted data H+ΔH, S+ΔS, L+ΔL according to the extraction signal. Among the color components, components that are not desired to be modified may be set to '0' as the shift amount.

Note that the system may be configured to include both the color conversion circuit 26 and the color correction circuit 2E, and to select and use one of them by switching modes. Further, it may be configured such that color conversion can be performed for some colors and color correction can be performed for other colors at the same time.

For example, in FIG. 1d, instead of the output data Hs l ss f t, s of the third replacement color specifying circuit, the correction value H+
If ΔH, S+ΔS, and L+ΔL are input to the color conversion circuit, color correction is performed on pixel data of a color matching the third designated extraction color.

Further, in each of the above embodiments, the KGB data is smoothed, but the same effect can be obtained even if the H8L data subjected to color system conversion is smoothed for each component.

Furthermore, in color specification and extraction/replacement, (1)
Equations, Equations (2.1) to (2.3) and Equations (3.1) to
Although the H8L color system defined by equation (3.3) is used, Lm defined by equations (2.1) to (2.3)
a'', bI' can be used as is, L'', U'', vl
Considerable effects can be expected by using the Munsell color system, the YIQ color system, or the simple HSL color system introduced at the 1st Color Engineering Conference J (1984; p. 91). It is possible.

■Detailed Description of the Invention According to the present invention, a microscopically discontinuous color area is transformed into a continuous color area by smoothing the sampled color signal. Colors can be stably and appropriately specified even for halftone originals, and the applicability of the color conversion device is improved.

Further, as explained in the embodiment, by processing the data separated into one color into the lightness component, hue, and saturation components, an extremely user-friendly color conversion device was obtained.

[Brief explanation of the drawing]

FIG. 1a is a block diagram showing the configuration of a device according to a first embodiment of the present invention, FIG. 1b is a block diagram showing a configuration of a device according to a second embodiment of the present invention, and FIG. 1C is a block diagram showing a configuration of a device according to a third embodiment of the present invention. FIG. 1d is a block diagram showing the structure of an apparatus according to a fourth embodiment of the present invention. FIG. 2a is a schematic diagram in which a 3×3 pixel matrix used for smoothing processing is developed on a plane. FIGS. 2b and 2c are block diagrams showing the detailed configuration of the smoothing circuit 21. FIG. Figure 3a, Figure 3b, Figure 3C, Figure 3d, Figure 3e,
Figures 3f, 3g, 3h, 31 and 3j
The figure is a schematic diagram of a digital filter used in the smoothing circuit 21 developed on a plane. FIG. 4a is a block diagram showing a detailed configuration of the first color system conversion circuit 22, and FIGS. 4b and 4c are block diagrams showing modifications thereof. . FIG. 5 is a block diagram showing the detailed configuration of the extraction color designation circuit 23. FIG. 6a is a block diagram showing a detailed configuration of the color recognition circuit 25, and FIGS. 6b and 6c are block diagrams showing modifications thereof. FIG. 7 is a block diagram showing the detailed configuration of the color conversion circuit 26. FIG. 8 is a block diagram showing the configuration of the color recognition circuit 25 of the fourth embodiment of the present invention, and FIG. 9 is a block diagram showing the configuration of the color conversion circuit 26. FIG. 10 is a block diagram showing the configuration of a color correction circuit of a fifth embodiment of the present invention. FIG. 11 is an explanatory diagram for explaining the drawbacks of the prior art. 1: Scanner (color signal generation means) 2: Image processing section 21: Smoothing circuit (smoothed signal generation means) 22.27
: Color system conversion circuit 23 : Extraction color designation circuit (extraction color signal generation means) 24 : Replacement color designation circuit (replacement color signal generation means) 25 : Color recognition circuit 26 : Color conversion circuit 25. 26 : (Conversion means) 28: OCR image correction circuit 29: Gradation processing circuits 2A, 2C, 2D: Selector 2B: Area determination section (area specifying means) 3: Printer 4: Digitizer 5: Operation section 4.5: (minimal area specifying means) 6 :System controller Figure 3a Figure 3b Figure 3C Figure 3d Figure 3e Figure 3f Figure 3g Figure 3h Figure 31 Figure 31 Figure [-1 Figure 5 L-1 Figure 6a Figure 1 b Figure C

Claims (4)

[Claims] (1) Color signal generation means that generates color signals of multiple components in association with each micro-region into which the original image is micro-divided; specifying an arbitrary first micro-region and second micro-region from the micro-segmented original image. Microregion specifying means; Smoothing color signals corresponding to each component of the microregion of interest and microregions in a predetermined vicinity thereof, and smoothing them in association with the microregion of interest. Smoothed signal generation means for generating a signal; Extracted color signal generation means for generating an extracted color signal specifying at least one component from the smoothed signal corresponding to the first minute area; Smoothed signal generation means for generating a smoothed signal corresponding to the second minute area Replacement color signal generation means for generating a replacement color signal that specifies at least one component from the smoothed signal; the smoothed signal corresponding to the minute area of interest and the extracted color signal are the components specified by the extracted color signal; a conversion means for replacing a component specified by the replacement color signal of the smoothed signal corresponding to the minute area with the replacement color signal if the two match with each other. (2) The converting means performs the replacement only in the area specified by the area specifying means, further comprising an area specifying means for specifying a region including a plurality of micro-areas obtained by subdividing the original image. A color conversion device according to claim (1). (3) Color signal generation means for generating color signals of multiple components in association with each micro-region into which the original image has been micro-divided; micro-area specifying means for specifying any micro-region from the micro-segmented original image; at least one Shift setting means for setting the shift amount of two components; Smoothing the color signals associated with each of the minute areas of interest and minute areas in a predetermined vicinity thereof for each component; Smoothed signal generation means that generates a smoothed signal in association with the region; Extracted color signal that generates an extracted color signal that specifies at least one component from the smoothed signal corresponding to the minute area designated by the minute area designation means. Generating means: When the smoothed signal corresponding to the minute area of interest and the extracted color signal match with respect to the component specified by the extracted color signal, the shift amount of the smoothed signal corresponding to the minute area is set. A color conversion device comprising: a shift means for shifting the component by a set shift amount; (4) The converting means further comprises a region specifying means for specifying a region including a plurality of micro-regions obtained by subdividing the original image, and the converting means performs the shift only in the region specified by the region specifying means. A color conversion device according to claim (3).