JPH0511702B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention reads a color original and performs color separation of the read original image (color of each pixel of the original image is set to one of a plurality of predetermined colors). The present invention relates to an image processing apparatus that performs color classification (classification into colors) and then performs multivalue processing, and more specifically relates to an image processing apparatus that is capable of separating an achromatic color in addition to two types of colors.

(Prior Art) Various methods are conventionally known as methods for reading a color document with a scanner and color-separating the read color images.

(1) First method As shown in Figure 8, a red light source 2 is placed on a color original 1.
A red illumination from a blue light source 3 and a blue illumination from a blue light source 3 are performed, and the respective optical information is received by a photoelectric conversion means 4 such as a CCD and converted into an electric signal. The values obtained by normalizing the output of the photoelectric conversion means 4 with the blank paper output value are V _R and V _B, respectively.
Then, these two signals are processed to obtain a color separation map. In 1107, the proceedings of the 1981 National Conference of the Institute of Electronics and Communication Engineers, they created a color separation map as shown in Figure 9 and suggested that it would be possible to separate multiple types of colors based on this color separation map. . The horizontal axis in FIG. 9 shows the normalized output (%) of the photoelectric conversion means 4 when the red light source is on, and the vertical axis shows the normalized output (%) of the photoelectric conversion means 4 when the blue light source is on.

(2) Second method Two types of photodetectors with different spectral sensitivity characteristics are provided for each pixel on the document, and the output of these photodetectors is
This method performs arithmetic processing on V _A and V _B to separate colors (Japanese Unexamined Patent Publication No. 44825/1983). For example, with respect to the vertical luminance axis (V _A + V _B ), when V _A + V _B ≧ a ₁ , white a ₂ < V _A + V _B < a ₁ , chromatic color, when V _A + V _B ≦ a ₂ , and black When logV _A −logV _B ≧ _{b 1} Red b ₂ <logV _A −logV _B < b ₁ Green When logV _A −logV _B ≦ _{b 2} When it is judged as blue. However, a ₁ , a ₂ , b ₁ , and b ₂ in the formula are certain constants. FIG. 10 shows the color separation map obtained in this manner.

(3) Third method This is a method of separating optical information into the three colors of red, green, and blue using a dichroic mirror, prism, or R, G, and B filters (Japanese Patent Laid-Open No. 50-62320) It is described in).

FIG. 11 is a diagram showing various methods of color separation. A is an image 12 of a photographic lens 11 that is transferred to a plurality of relay lenses 13 to 16 and a dichroic mirror 1.
7 and 17', the images are separated into three colors and re-imaged on the image pickup tubes 18 to 20, respectively. In the example shown in B, a plurality of prisms 21 to 24 having a special shape are arranged between the photographing lens 11 and each of the image pickup tubes 18 to 20,
The dichroic mirrors 17 and 17' between the prisms 21 and 22 and between the prisms 23 and 24 are eliminated, so that the colors can be separated into three colors.

C is three prisms 25, 26 with acute apex angles,
26' are fitted together to form a triangle ABC as shown in the figure, and dichroic mirrors 27 and 28 are formed on the boundary surface of each prism to perform three-color separation. This is a configuration in which the prism of the example shown in C is exactly turned over. Dichroic mirrors 29 and 30 are formed at the boundaries of each prism, respectively.

(Problems to be Solved by the Invention) In the field of conventional image processing methods and devices as described above, developments have been made up to the stage of color separation, but how to process the separated image data is still unclear. To this point, it is still under development.

The present invention has been made in view of these points, focusing on the drawbacks of the conventional image processing methods and devices described above, and its purpose is to improve the ability to obtain well-balanced output images. The goal is to realize an image processing device that can.

(Means for Solving the Problems) The present invention solves the above-mentioned problems by providing an image reading means for separating reflected light or transmitted light from a color document into at least two colors and reading each pixel of the document image. a color information storage means storing in a memory a color separation map for classifying the color into one of a plurality of predetermined colors, the color separation map having a density corresponding value set at each address; Color separation information creation means for creating an address signal for the memory storing the color separation map from the output signal of the image reading means and outputting it to the color information storage means; and an address signal output from the color separation information creation means. and a multi-value converting means for converting the density corresponding value read from the color information storage means into multiple values, and the multi-value converting means converts a threshold value determined for each of the plurality of predetermined colors into multiple values. The apparatus is characterized in that it comprises a threshold value generation means, and a comparison means for comparing the threshold value outputted by the threshold value generation means and the density corresponding value read from the color information storage means.

(Function) In the present invention, reflected light or transmitted light from a color document is separated into at least two colors and read, and from this signal, the color separation information creation means generates an address signal to the memory in which the color separation map is stored. Create and output. The multi-value conversion means performs multi-value conversion by comparing the density corresponding value read out from the color information storage means using this address signal with a threshold value determined for each of a plurality of predetermined colors.

Conventionally, in monochrome analog copying machines, the bias voltage of the developing section is adjusted in order to copy colors such as red and blue of a color original with sufficient density. For example, the developing bias voltage is lowered for documents with similar reflection densities, such as orange and light blue. On the other hand, for blue writing instruments (for example, ballpoint pens), in order to obtain sufficient copy density, writing instrument manufacturers use measures such as adding a small amount of carbon black. In addition, in the case of a document containing a mixture of characters written with black, blue, and red writing instruments, even if the black and blue characters are copied with sufficient density, the red characters have a problem with the spectral sensitivity characteristics of the photoreceptor. However, it may not be copied with sufficient density. The reason for this is that the reflection density differs depending on the color. In such a case, according to the present invention, since the threshold value is changed depending on the color, sufficient copy density can be obtained for all colors without changing the developing bias voltage. .

(Example) Hereinafter, the present invention will be described in detail with reference to the drawings.

First, prior to a specific explanation, in order to facilitate understanding of the process performed by the apparatus of the present invention, the flow of the process will be explained using the flowchart shown in FIG.

Step 1 Scan the color original. In order to read a color original, it is necessary to illuminate the original with a light source. As the light source, for example, one having a spectrum as shown in FIG. 2A is used. In the figure, the horizontal axis indicates wavelength (nm), and the vertical axis indicates relative intensity (%). A color original is irradiated with a light source having a spectrum as shown in FIG. 2A. The reflected light from the color original is incident on an optical means as shown in FIG. 3, for example, and is decomposed into red and cyan components. In the figure, light reflected from a color original is incident on a prism 31. The incident light is reflected in a cyan color by a dichroic mirror 33 formed at the interface between the prisms 31 and 32.
The light is emitted from the prism 31. On the other hand, red colors are transmitted through the dichroic mirror 33 and are transmitted through the prism 3.
It is emitted from 2. That is, the dichroic mirror 33 used in the present invention separates incident light into two complementary colors. FIG. 2B is a diagram showing the characteristics of a dichroic mirror. In the figure,
The horizontal axis is wavelength (nm), and the vertical axis is transmittance (%).
From the figure, it can be seen that red wavelengths with longer wavelengths are transmitted, and cyan wavelengths with shorter wavelengths are reflected. Here, the term "complementary color relationship" refers to a relationship between colors X and Y such that when two colors are X and Y, respectively, X+Y=white.

Step 2 Separate the two colors red and cyan.
Convert it into an electrical signal using a photoelectric conversion element such as a CCD. FIG. 2C is a diagram showing the spectral sensitivity characteristics of the CCD used in the present invention. In the figure, the horizontal axis represents wavelength (nm), and the vertical axis represents relative sensitivity (%). As is clear from the figure, this CCD has a peak around the wavelength of 600 nm. These photoelectrically converted signals are normalized by the output value of the reference color (white). Let the normalized red photoelectric conversion signal be V _R and the cyan photoelectric conversion signal be V _C . In this example, these photoelectric conversion signals are converted into 6-bit digital data by an A/D converter. This is to make it easier to perform arithmetic processing on a computer (or microcomputer).

Step 3 Digital image data V _R obtained in Step 2,
A coordinate system is created using V _C , and color separation is performed based on the created color separation map. To determine this coordinate axis, consider the following points.

In order to be able to express halftones, we adopt the concept of the reflectance (reflection density) of the document, which corresponds to the brightness signal of a television.

Introduce the concept of pigments (including hue and saturation) such as red and cyan.

From the above, it is preferable to use, for example, the following as the luminance signal information and color difference signal information.

Luminance signal information (5 bits) = V _R + V _C (1) Sum of V _R , V _C (0≦V _R ≦1.0, 0≦V _C ≦1.0) V _R
+V _C (0≦V _R +V _C ≦2.0) corresponds to the black level (=0) and white level (=2.0), and all colors are from 0 to 2.0.
Exists within the range of

Color difference signal information (5 bits) = V _R / (V _R + V _C ) or V _C / (V _R + V _C ) (2) In the case of achromatic color, V _R component included in the whole (V _R + V _C ) , the proportion of the V _C component is constant. Therefore, V _R /(V _R +V _C )=V _C /(V _R +V _C )≒0.5. On the other hand, in the case of chromatic colors, V _R /
The value of (V _R +V _C ) or V _C /(V _R +V _C ) is one measure of the hue and saturation of the original. That is, (1) Red color 0.5<V _R /(V _R +V _C )≦1.0 0≦V _C /(V _R +V _C )<0.5 (2) Cyan color 0≦V _R /(V _R +V _C ) It can be expressed as <0.5 0.5<V _C /(V _R +V _C )≦1.0. From this, the coordinate axes are V _R + V _C and V _R / (V _R + V _C ) or V _C / (V _R +
By using a coordinate system with two axes of V _C ),
It becomes possible to clearly separate chromatic colors (red, cyan) and achromatic colors.

FIG. 4 is a diagram showing an example of a color separation map in which the color gamut is divided according to the color separation method described above.
Note that in this specification, the term "color gamut" is used to mean classified specific colors. In the figure, the horizontal axis represents color difference signal information V _C /(V _R +V _C ), the left vertical axis represents luminance signal information V _R +V _C , and the right vertical axis represents reflection density due to achromatic color. Color difference signal information =
There is an achromatic color in the vicinity of 0.5 and the area where the luminance signal information is small (shaded area in the figure), the area smaller than 0.5 is redish, and the area larger than 0.5 is cyanish. or,
Since there is a correspondence relationship as shown in the figure between the reflection density and the luminance signal information V _R +V _C , it is easy to directly link it to the output value. In the example shown in the figure, V _C /(V _R + V _C ) is plotted as color difference signal information on the horizontal axis, but V _R /
The same is true for (V _R +V _C ).

In an actual image processing apparatus, the color separation map shown in FIG. 4 is created and stored in a ROM table. FIG. 5 shows an example of such a ROM table, which has a capacity of 32×32 dots.
Therefore, the number of address bits is: Row address (V _R + V _C ) 5-bit column address V _C / (V _R + V _C )
It is 5 bits. This ROM table stores quantized density correspondence values (patterns) obtained from the reflection density of the original. Although the coordinates shown in FIG. 4 are created as a ROM table as shown in FIG. 5, they do not necessarily have to be a ROM table, and the means for realizing this is not limited to hardware.

Step 4 The image data separated in Step 3 is multivalued using independent threshold values for each color gamut. That is, when outputting from the output device, the above-mentioned
The density-corresponding values stored in the ROM table are read using the address signal, and the read-out density-corresponding values are binarized using the threshold values specified for each color gamut of red, cyan, and achromatic colors, and the result is output data. . still,
It does not necessarily have to be a binary output, but may be a multi-valued output.

For example, when using a fixed threshold value, there is a clear correspondence between the luminance signal V _R +V _C and the reflection density of the document, as explained with reference to FIG. D, E, F
It is mainly distributed within the range with values such as (hexadecimal). On the other hand, red images are mainly distributed within a range having values such as 8, 9, A, B, C, and D. Now, if we set "C" as a fixed threshold value without changing the threshold value for each color gamut, all blue images will be "1" when binarized.
The output is expressed with sufficient density. On the other hand, in a red image, 8, 9, A, and B become "0", and sufficient density cannot be obtained. In order to eliminate such drawbacks of fixed threshold values, the threshold values are changed for each color. For example, in the case of the above-mentioned example, "8" is used as the threshold value in the case of red color binarization. When "8" is used as the threshold value, all of 8 to D become "1", and output expression can be achieved with sufficient density.

FIG. 6 is a block diagram showing an embodiment of the apparatus of the present invention. In the figure, 41 is the first CCD that receives red optical information, 42 is the second CCD that receives cyan optical information, and 51 is the first CCD 4.
a first amplifier that amplifies the photoelectric conversion output of 1, 52;
is a second amplifier that amplifies the photoelectric conversion output of the second CCD 42. first and second CCD41,
42 constitutes the photoelectric conversion means 40, and the first and second
Amplifiers 51 and 52 constitute an amplifying section 50.
61 is a first A/D converter that converts the output of the first linear amplifier 51 into digital data at equal intervals, and 62 is a second A/D converter that similarly converts the output of the second linear amplifier 52. In the D converter, these first
and second A/D converters 61 and 62 constitute an A/D conversion section 60. The number of bits used for the A/D converters 61 and 62 is, for example, about 6 bits.

71 is the output of the first A/D converter 61 and the second
an arithmetic processing circuit that receives the output of the A/D converter 62 and performs various arithmetic processing to obtain color separation information; 72 is a first memory that stores the luminance signal data V _R +V _C of the arithmetic processing circuit 71; 73 is an arithmetic processing circuit 71
A second memory 81 stores the color difference signal data V _C /(V _R +V _C );
A third memory receives the output of 73 as an address and outputs chromatic color (red, cyan) data (density corresponding value); 82 is the same as the first and second memory 72;
73 output as address and achromatic color (black,
This is the fourth memory that outputs (gray, white) data.
an arithmetic processing circuit 71 and first and second memories 72;
73 constitutes the color separation information creation means 70, and the third
and fourth memories 81 and 82 constitute color information storage means 80. As the arithmetic processing circuit 71, for example, a microcomputer is used. A combination of the third and fourth memories 81 and 82 becomes the ROM table shown in FIG.

43 is a first buffer that temporarily stores the output of the third memory 81, and 44 is a second buffer that temporarily stores the output of the fourth memory 82.
45 is a color selection circuit that receives B (black), B (blue), and R (red) selection signals, and its output is sent to the first and second buffers 4.
3 and 44. 46 is the first and second
47 is a comparison circuit that compares the image data (density corresponding value) outputted from either one of the buffers 43 and 44 with a threshold value to obtain multivalued data (including binary data); This is a threshold circuit that outputs independent threshold values for each color gamut in response to a color selection signal. Comparison circuit 46
As the threshold value circuit 46, for example, a ROM in which threshold values for each color gamut are stored in advance is used. or,
In addition to the output from the color select circuit 45, a density regulation signal is also input to the threshold circuit 46, so that the threshold value can be changed by the density regulation signal as well as the color selection signal. The numbers in the figure indicate the number of bits of the signal line. The operation of the device configured as described above will be explained as follows.

The optical information of the color original is incident on an optical means as shown in FIG. 3, and is separated into red and cyan. The decomposed red and cyan optical information enters CCDs 41 and 42, respectively, and is converted into electrical signals. The converted image signals enter amplifiers 51 and 52, where they are amplified to a predetermined level, and then converted into digital data by subsequent A/D converters 61 and 62. The arithmetic processing circuit 71 receives the red and cyan image data converted into digital data and normalizes it using the output value of the reference color (white). That is, the image data of the reference color is set to 1.0, and the image data of each of the red and cyan colors are normalized. The image data normalized in this way is
Let them be V _R and V _C , respectively.

Next, the arithmetic processing circuit 71 performs the calculations expressed by equations (1) and (2), and stores the results in the first and second memories 72 and 73. That is, the luminance signal V _R +V _C
is stored in the first memory 72, and the color difference signal V _C /
(V _R +V _C ) is stored in the second memory 73. The outputs of these first and second memories 72 and 73 are given to third and fourth memories 81 and 82 as address signals. 3rd and 4th memory 8
Density correspondence data (density correspondence values) stored at addresses corresponding to the input addresses are output from 1 and 82 and held in buffers 43 and 44, respectively.

On the other hand, the color select circuit 45 receives the BBR signal and provides a select signal to either the first or second buffer 43 or 44. For example, the first
When the second buffer 43 is selected, data corresponding to red or cyan density is output, and when the second buffer 44 is selected, data corresponding to black density (white, white, etc.) is output.
Gray, black) density correspondence data is output. The output density-corresponding data enters the comparator circuit 46. On the other hand, a threshold circuit 47 that provides a threshold value to the comparison circuit 46
The color selection signal and the density regulation signal from the color selection circuit 45 are inputted to the color selection circuit 45, and a threshold value corresponding to the color gamut is outputted. The comparison circuit 46 uses the threshold value set for each color gamut to divide the density corresponding data into two.
Convert to value data (or multi-value data in some cases). By using this binary data as input data to a printer, copying machine, etc., it can be output and displayed externally. The means for recording the output value may be optical fiber (OFT), liquid crystal (LCD), exposure on the photoreceptor surface using a laser, inkjet, thermal transfer, recording on silver salt or non-silver salt material, or
There are no limitations on output to CRT, etc. The above operations are
This will be repeated each time the CCDs 41, 42 receive new optical information.

In the embodiment shown in FIG. 6, the color separation information creation means 70 and the color information storage means 80 are provided separately, but they can also be formed integrally. FIG. 7 is a block diagram showing another embodiment of the apparatus of the present invention, in which color separation information creation means and color information storage means are integrally formed. Components that are the same as those in FIG. 6 are designated by the same numbers. Digital image data from the A/D converter 60 enters an arithmetic processing circuit 71. The arithmetic processing circuit 71 receives the input image data and calculates V _R and V _C , and then calculates the luminance signal V _R +
Find V _C and the color difference signal V _C /(V _R +V _C ). These values are immediately stored in memories 81 and 82 as address data.
It's in. Since the subsequent operation is the same as that shown in FIG. 6, the explanation will be omitted.

In the above explanation, V _C /(V _R +V _C ) was used on the horizontal axis of the color separation map shown in FIG. 4, but V _R /
It may be (V _R +V _C ). A similar effect on the horizontal axis can also be obtained by using (V _R −V _C )/(V _R +V _C ) or (V _C −V _R )/(V _R +V _C ) as the horizontal axis. For example, if we use (V _R − V _C )/(V _R + V _C ) on the horizontal axis, (V _R − V _C )/(V _R + V _C ) = 0 Achromatic color in the vicinity > 0 Reddish color < 0 Cyan color.

Furthermore, in the above description, the spectral characteristics of the dichroic mirror are those of red transmission and cyan reflection type, but the present invention is not limited to this. Any dichroic mirror having spectral characteristics that separates into two related colors may be used. Further, the color separation means is not limited to a dichroic mirror, but may be any device that can separate colors. For example, it may be a spectral filter or the like. Further, the color separation map is not limited to the T-shape shown in FIG. 4, and may be of any shape.

In the above description, the arithmetic processing circuit 71 includes V _R +V _C and V _C /(V _R +V _C ) calculations and the A/D converter 61,
The case where the normalization operation of the output of 62 is performed is taken as an example. However, the present invention is not limited to this, and the memories 72 and 73 may be addressed immediately with the outputs of the A/D converters 61 and 62. In this case, the memories 72 and 73 have V _R +V _C and V _C /(V _R + V _C ) corresponding to the input addresses V _R and V _C , respectively.
It is preferable to use a ROM in which data having characteristics is written. Furthermore, in this case, the A/D converters 61 and 62 are set to full scale in advance so that full scale (FS) data of 1.0 is output when the image data of the white reference color is input. If the adjustment is made, the arithmetic processing circuit 71 becomes unnecessary.

Next, a modification of the present invention will be described. 12th
13 are block diagrams showing a modification of the present invention. The example shown in FIG. 12 will be described first. Photoelectric conversion means 9 such as CCD converts two types of optical signals.
1,91' is converted into electrical signals V _A and V _B as disclosed in Japanese Patent Application Laid-Open No. 57-44825. The converted V _A and V _B signals are amplified by amplifiers 97 and 97', and then an adder 98 performs the calculation (V _A +V _B ). This calculation output is determined by the black discrimination circuit 99 to be white when V _A +V _B ≧a ₁ , white when a ₂ ≦V _A +V _B < a ₁ , and black when chromatic color V _A +V _B ≦a ₂ . The reason why the output from the black discrimination circuit 99 is input to the color discrimination circuit 94 is that a ₂ ≦V _A +V _B
If <a ₁ (it is a chromatic color), the color discrimination circuit 9
This is to inform 4.

On the other hand, the V _A and V _B signals are sent to logarithmic amplifiers 92 and 9.
2' for logarithmic amplification. Logarithmically amplified signal
logV _A and logV _B are then subtracted by the subtracter 93 (logV _A −
logV _B ) is calculated. This (logV _A −
When logV _A −logV _B ≧b ₁ , red b ₂ <logV _A − logV _B < b ₁ , green when _{logV A −logV B} ≦b _2, blue. I judge that. The color data and density data based on this determination result are processed by a comparison circuit 95 provided for each color.
96 and 97, and is multi-valued (including binarized) using threshold values independently set for each color gamut. Similarly, multi-value conversion is performed for black using threshold values independently set by the comparator circuit 100. The threshold values for each color gamut are provided by a threshold circuit 96 that receives a color select signal and generates an independent threshold value for each color gamut.

Next, the example shown in FIG. 13 will be explained. In the example shown in the figure, color separation is performed based on the condition that optical information can be separated into colors using a color separation map as shown in FIG. Photoelectric conversion means 10
The red signal V _R and the blue signal V _B converted into electric signals by amplifiers 102 and 101'
After being amplified by 102', A/D converter 103,
The data is converted into digital data at 103' and given as addresses to red memory 104, blue memory 105, and black memory 106, respectively. memory 10
4 to 106 constitute a ROM table based on the color separation map shown in Figure 14, and the BBR
The image data selected by the signal is outputted to the corresponding buffer memory 107-1.
09 and is held. Output data from the buffer selected by the color selection circuit 110 is multi-valued by the comparison circuit 111.
The threshold value of this comparison circuit 111 is given by a threshold value circuit 112, and a color select signal is given to the threshold value circuit 112 from a color select circuit 110, and the threshold value circuit 112 generates a threshold value according to the color gamut. Note that in the above embodiments, each threshold value that can be set has been described, but instead of using a fixed threshold value for each color gamut, it is also possible to set a plurality of threshold values for halftone expression. That is, a threshold value group such as a dither matrix may be set for each color.

(Effects of the Invention) As explained in detail above, according to the present invention,
By setting independent threshold values for each color gamut for each color-separated image data, it is possible to obtain an image processing device that can output and express the density of an image of each color in a well-balanced manner.

[Brief explanation of the drawing]

Fig. 1 is a flowchart explaining the processing flow shown to facilitate understanding of the processing in the apparatus of the present invention, and Fig. 2 shows examples of spectral characteristics of the light source, dichroic mirror, and CCD used in the present invention. 3 is a block diagram showing an embodiment of the optical means used in the present invention, FIG. 4 is a diagram showing an example of a color separation map according to the present invention, FIG. 5 is a diagram showing an example of a ROM table, and FIG. The figure is a block diagram showing one embodiment of the apparatus of the present invention, FIG. 7 is a block diagram showing another embodiment of the apparatus of the present invention, FIG. 8 is a diagram showing the concept of the color separation optical system, and FIG. 10 shows an example of a color separation map, FIG. 11 shows an example of a conventional configuration of a color separation optical system, FIGS. 12 and 13 show a modification of the present invention, and FIG. 14 shows an example of a color separation optical system. FIG. 3 is a diagram showing a color separation map. 1...Original, 2...Red light source, 3...Blue light source,
4, 40...Photoelectric conversion means, 11...Photographing lens, 13-16...Relay lens, 17, 17',
27, 28, 29, 30, 33... Dichroic mirror, 18-20... Image pickup tube, 21-26,
26', 31, 32...prism, 41, 42...
...CCD, 43, 44...Buffer, 45...Color selection circuit, 46...Comparison circuit, 47...
Threshold circuit, 50...Amplification section, 51, 52...Amplifier, 60...A/D conversion section, 61, 62...A/
D converter, 70...Color separation information creation means, 71...
...Arithmetic processing circuit, 72, 73, 81, 82...Memory, 80...Color information storage means.

Claims (1)

[Scope of Claims] 1. An image reading means that separates and reads reflected light or transmitted light from a color original into at least two colors, and each pixel of the original image is read in one of a plurality of predetermined colors. a color information storage means storing in memory a color separation map that classifies the color separation map and in which each address is set with a density corresponding value; and a color information storage means storing the color separation map based on the output signal of the image reading means. color separation information creation means for creating an address signal for the color information storage means and outputting it to the color information storage means; and a density corresponding value read out from the color information storage means based on the address signal output from the color separation information creation means. and a multi-value conversion means for converting into multi-value, the multi-value conversion means includes a threshold generation means for generating a threshold value determined for each of the plurality of predetermined colors, and an output of the threshold value generation means. An image processing apparatus comprising: a comparison means for comparing a threshold value read out from the color information storage means with a density corresponding value read from the color information storage means.