JPH10275229A - Filtering device and method for color image - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a color image filtering device and method capable of relieving the burden of the arithmetic processing needed for the filter processing and also significantly accelerating the filter processing speed. SOLUTION: This filtering device is provided with an achromatic signal production part 20 which extracts achromatic components by applying the weighted mean to the value of pixels having the color components corresponding to each other to form a color image, a filter part 24 which extracts the achromatic component having the intermediate value as the central value among those achromatic components of pixels included in a filter area set on the color image, the subtraction processing parts 21, 22 and 23 which subtract the achromatic component value of an attentional pixel from the value of this attentional pixel that should filter every color component, and the addition processing part 25, 26 and which and the center value extracted at the part 24 to the calculated value of the parts 21 to 13. In such a constitution, the attentional pixel and the filter area are successively scanned on the color image and the noises are eliminated from an entire color image.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a filtering apparatus and a filtering method for a color image such as a color scan image of a positive or negative or transmissive or reflective original, or a color digital image of a digital camera or the like.

[0002]

2. Description of the Related Art A color image read by a scanner or the like contains various kinds of noise due to the influence of irregularities (granularity) on a read document and electric noise. The noise of an achromatic component such as a gray component mixed into all of the respective color components such as R, G, B, etc., which constitutes the image, is particularly conspicuous.

[0003] In the outline enhancement processing of a color image, a blurred image is created based on an achromatic image created by extracting an achromatic component from a monochromatic image of each color component, and each pixel in the blurred image is created. Since the edge-enhancement signal is generated by subtracting the pixel value of each pixel in the corresponding achromatic image from the pixel value, noise of the achromatic component is easily mixed into the edge-enhancement signal, and is enhanced by the edge enhancement process. Easy to be.

A first prior art for removing such achromatic color noise is a filtering device shown in FIG. In this filtering device, median filters 1, 2, and 3 are provided for each of R, G, and B color signals forming a color image, and median filter processing is individually performed for each color signal.

[0005] The median filter is such that a matrix-like filter region is set in an image so as to include a target pixel to be subjected to filter processing, and a pixel value of the target pixel in the filter region is set in the matrix region. The noise of the high-frequency component is removed by adopting an intermediate value among the pixel values of a plurality of pixels to be processed. By sequentially scanning the pixel of interest and the filter area on the image, noise of the entire image is reduced. It has been removed.

[0006] In the median filter processing, in the process of selecting an intermediate value in a filter area, pixel values of all pixels located in the filter area are compared with each other, and the pixel values of each pixel are arranged in ascending order of value. In this case, a rearrangement operation is performed in which a pixel value located at an intermediate position in the sequence is set as an intermediate value, and such a rearrangement operation has a property that processing takes time.

In addition, as a second conventional technique, a pixel value of a pixel located in a filter area is weighted and averaged, and an average value obtained by the weighted averaging is used as a pixel of a pixel of interest located in the center of the filter area. Image filters that output values are replaced by R, R instead of the median filters 1, 2, 3.
There is a filtering device provided for each of the G and B color signals.

[0008]

However, in the filtering device of the first prior art, the median filter processing is performed individually for each color signal, so that the processing load is large and the processing is enormous. There is a problem that it takes time. As a method to shorten the processing time,
For example, a method that employs a processing device capable of performing three times the parallel arithmetic processing can be considered, but there is a problem that the cost of the device is increased.

Also, in the filtering device of the second prior art, since the filtering process is individually performed for each color signal, an averaging process having a large processing load must be performed for each color signal. There is a similar problem that it takes a lot of time.

The present invention has been made in view of the above problem, and provides a color image filtering apparatus and a filtering method capable of reducing the load of arithmetic processing required for filter processing and greatly speeding up the filter processing. The purpose is to:

[0011]

The technical means for achieving the above object is to provide a color image obtained by superimposing a plurality of color component monochromatic images such that a target pixel to be subjected to a filtering process is included in the color image. By performing a filtering process on the target pixel in the filter region and sequentially scanning the target pixel and the filter region on the color image, a noise component included in the color image is removed. A color component filtering device for extracting an achromatic component for each pixel based on a pixel value of each pixel in each of the monochrome images; and By performing a predetermined filtering process based on the component values of the achromatic components, the component values of the achromatic components of the pixels in the filter area are represented. Filter means for extracting a table value; subtraction means for subtracting, from the pixel value of the pixel of interest of each color component, the component value of the achromatic component extracted by the achromatic component extraction means of the pixel of interest; Adding means for adding the representative value extracted by the filter means to the calculated value.

Preferably, the filter means sets, as the representative value, a component value having an intermediate value in the filter region among component values of the achromatic component of a pixel located in the filter region. Good to extract.

[0013] Preferably, the filter means includes:
It is preferable that an average value obtained by averaging the component values of the achromatic components of the pixels located in the filter region is extracted as the representative value.

Further, preferably, the achromatic component extracting means extracts the achromatic component by averaging the pixel values of the pixels of the respective color components corresponding to each other.

[0015] Further, a technical means for achieving the above object is to form a predetermined filter area on a color image formed by superimposing a plurality of color component monochromatic images so as to include a target pixel to be filtered. Setting, performing a filtering process on the target pixel in the filter region and sequentially scanning the target pixel and the filter region on the color image, thereby removing a noise component included in the color image. A filtering method, wherein an achromatic component extracting step of extracting an achromatic component for each pixel based on a pixel value of each pixel in each of the monochromatic images, and the achromatic component of a pixel located in the filter area By performing a predetermined filtering process based on the component values of the components, a representative value representing the component values of the achromatic components of the pixels in the filter area is extracted. A filtering step, a subtraction step of subtracting, from the pixel value of the target pixel of each color component, the component value of the achromatic component extracted in the achromatic component extraction step of the target pixel, and a calculated value of the subtraction step, An adding step of adding the representative value extracted in the filtering step.

[0016]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An example of the overall configuration of an apparatus to which an embodiment of the present invention is applied will be described. FIG. 1 is a diagram showing an overall configuration of an apparatus to which an embodiment of the present invention is applied. The image input unit 100 optically reads a color original, such as an input scanner, and generates color digital image data. Then, the generated image data is transferred to the image processing unit 200. The image data is output to an image output unit 300 such as an output scanner after a predetermined process is performed in the image processing unit 200. In the output scanner 300, the image is recorded on a recording medium such as a film. Note that the image input unit 100,
In order to cause the image processing unit 200 and the image output unit 300 to perform an operation desired by the operator, the operation input unit 2
01 and an information display unit 202 are provided.

The image processing unit 200 includes a filtering unit 210 for removing noise contained in the color image, and an outline emphasis processing unit 220 for performing an outline emphasis process on the color image that has passed through the filtering unit 210. Are provided, and the embodiment of the present invention is applied to the filtering unit 210.

Next, an outline of the filtering process in the color image filtering apparatus according to the first embodiment of the present invention will be described. First, as shown in FIG.
, A single-color image 10a, 1 of each color component of R, G, B
(2m + 1) × (2n + 1), in this case, a 3 × 3 matrix filter area F is set on 0b and 10c so as to include the target pixels OPa, OPb and OPc of each color component to be subjected to the filter processing at the center. And the filter area F
The achromatic color component is extracted based on the pixel value of the pixel of each color component located inside the achromatic color component.

The achromatic color component is extracted by multiplying the pixel values Dr, Dg, Db of the pixels corresponding to each color component by weighting factors Ar, Ag, Ab, respectively.

[0020]

(Equation 1)

The weighted averaging is performed as described above to obtain a component value D of an achromatic component for each pixel. In addition,
Here, Ar, Ag, and Ab are set to 1, and the average value of Expression 1 is substantially a simple average value. The term “average” in this specification refers to a concept that includes both weighted average and simple average.

As described above, the achromatic components of the pixels in the filter area F are sequentially extracted. When the achromatic components are extracted from all the pixels in the set filter area F, the filter area is extracted. Among the component values D of the achromatic components of the pixels located in F, one having an intermediate value is selected as a representative value D ′.

The pixel values Dr, D of the target pixels OPa, OPb, OPc of each color component to be subjected to the filtering process
The component value D of the achromatic component of the target pixel OPa, OPb, OPc is subtracted from g, Db, and the result of the operation is represented by the representative of the achromatic component of the pixel in the filter region F including the target pixel OPa, OPb, OPc. By adding the value D ',
Noise of the pixels of interest OPa, OPb, OPc, particularly noise of achromatic components, is removed. FIGS. 6 to 8 show how the noise removal is actually performed.
Will be described later.

When the filtering process for one target pixel OPa, OPb, OPc is completed, the target pixel OPa, OPb
OPb, OPc, and the filter area F are sequentially scanned on the single-color images 10a, 10b, and 10c, as shown in FIG. 2, and all the pixels in the single-color images 10a, 10b, and 10c are similarly scanned as described above. By performing an appropriate filtering process, a noise component of the entire color image 10 is removed. That is, the feature of the present invention is that R,
An achromatic color component is extracted from the pixel values of the G and B color components, and a filtering process is performed based on the extracted achromatic component values.

FIG. 3 is a block diagram showing the electrical configuration of the color image filtering apparatus according to the first embodiment of the present invention. The input signals to the filtering device are R, G, and B color signals indicating the pixel values of each pixel constituting the single-color images 10a, 10b, and 10c. The signals are sequentially input to a chromatic signal generation unit 20 (achromatic component extraction unit) and subtraction processing units (subtraction units) 21, 22, and 23 provided for each color signal.

The achromatic signal generator 20 generates an achromatic signal S by weighting and averaging the signal values (pixel values) of the input color signals R, G, and B as described above. The achromatic signal S of the target pixels OPa, OPb, and OPc to be subjected to the filtering process in the chromatic signal S is output to the subtraction processing units 21, 22, 23, and the target pixel OP
The achromatic signal S of all pixels in the filter area F including a, OPb, and OPc is output to the filter unit 24. That is, as shown in FIG. 4, the achromatic signal S _ij of nine pixels in the filter area F is sequentially output from the achromatic signal generation unit 20 to the filter unit 24, and the subtraction processing units 21, 22, 23, the target pixels OPa, OPb, O of the nine signals S _ij
Achromatic signal S ₂₂ of the Pc is to be outputted.
Here, the signal value of the achromatic signal S indicates the component value D of the achromatic component.

Each of the subtraction processing units 21, 22, and 23 outputs a color signal R, G,
By subtracting the achromatic signal S from B, signals r, g, b corresponding to each color component are generated, and the corresponding addition processing units 25, 26, 27 provided for each color component.
(Adding means). Here, the signal values of the signals r, g, b are values obtained by subtracting the signal value of the achromatic signal S from the signal values of the color signals R, G, B.

On the other hand, the filter unit 24 performs a magnitude comparison between the signal value s _ij of the nine achromatic signal S _ij input, a signal value s _ij of the nine signal S _ij, shown in FIG. 5 _So that the signal value s _ij located at the middle position of the sequence is extracted as a representative value D ′, and the signal S ′ having the representative value D ′ as a signal value is added to each addition processing unit. 2
5, 26, 27.

Each of the addition units 25, 26, 27 receives the signals r, g, b and the signal S 'for the target pixel OPa, OPb, OPc from the subtraction units 21, 22, 23 and the filter unit 24. And the noise-removed color signals R ′, G ′, B ′ having the signal value of the signal S ′ added to the signal values of the signals r, g, b, and having the operation result as the signal value.
Is output to the outside.

As described above, one attention pixel OPa, OP
When the filtering process on the color signals R, G, B of b, OPc is completed, the target pixels OPa, OPb, OPc and the filter area F are sequentially updated, and noise removal is performed on all the pixels in the color image 10. The color signals R ′, G ′, and B ′ from which noise has been removed are sequentially input to an outline emphasis processing unit 220 described later.

In the filtering device having such a configuration, the achromatic signal generation unit 20 is provided with two filters in order to eliminate the waste of generating the achromatic signal S for one pixel many times.
The achromatic signal generation unit 20 is provided with a memory capable of storing the achromatic color formation signals S of the pixels of the (n + 1) th line, here three lines, and the achromatic signal of the pixels in the filter region F where the filter unit 24 is to perform the filtering process. S may be sequentially read out from the memory of the achromatic signal generator 24. Correspondingly, in order to synchronize the signals r, g, b, etc. output from the subtraction processing units 21, 22, 23 and the signal S ′ output from the filter unit 24, the subtraction processing unit A delay processing unit may be provided between 21, 22, 23 and the addition processing units 25, 26, 27, and the like.

Next, based on FIG. 6 to FIG. 8, each target pixel OPa, O
The manner in which the noise of Pb and OPc is removed will be described. FIG. 6A shows the color signals R, G, and B of the target pixels OPa, OPb, and OPc of the respective color components to be subjected to the filter processing, and FIG. 6B shows the subtraction processing units 21, 22, 23.
6 show signals r, g, and b output from FIG.
(C) shows the color signals R ′, G ′, and B ′ from the addition processing units 25, 26, and 27 from which noise has been removed.
FIG. 7 shows the achromatic signal S _ij of each pixel in the filter area F generated by the achromatic signal generation unit 20.

As shown in FIG. 6A, the target pixel O
If the color signals R, G, and B of Pa, OPb, and OPc include noise components Nr, Ng, and Nb of achromatic components such as gray components, the noise components Nr, Ng, and Nb cause the noise components in FIG. As shown in the figure, the target pixels OPa, OPb,
Also it becomes possible to include noise components N achromatic signal S ₂₂ of OPc, a large outlier from the signal values s _ij of the achromatic signal S _ij of the surrounding pixels whose signal value s ₂₂ contains no noise. Thus, the pixel of interest OPa, OPb, the signal value S ₂₂ of the achromatic signal OPc, without being extracted in the extraction process of the representative value by the filter unit 24, are excluded, proper other pixels in the filter region F a signal value s _ij of the achromatic signal S _ij, for example, the signal value s ₁₃ of the signal S ₁₃ is extracted as the representative value D '.

[0034] The color signals R, G, the pixel of interest in FIG. 7 B OPa in FIG 6 (a), OPb, when subtraction processing achromatic signal S ₂₂ of the OPc, signal r of FIG. 6 (b), g, b
FIG. 6C shows the result of adding a signal S ′ having a representative value D ′, that is, a signal S _{13 in} this case, to the signals r, g, and b in FIG. 6B. The color signals R ′, G ′, B ′ from which noise has been removed are obtained.

In other words, the processing units 21, 22, 23
In the subtraction processing and addition processing of 25, 26 and 27, as shown in FIG. 8, achromatic signal S of the peripheral pixels including the noise component N from the achromatic signal S ₂₂ does not include the noise
_ij (for example, S ₁₃ ), a noise component N is extracted, and the extracted noise component N is
It can be said that the subtraction processing is performed from the color signals R, G, and B in (a).

Strictly speaking, the value of the noise component N included in the achromatic signal S _ij is the value of the noise component Nr, Ng, Nb included in each of the color signals R, G, B. Although the average value corresponds to the average expression, in the drawing, the values of the noise components N, Nr, Ng, and Nb are treated as the same value for simplification.

Next, a case where noise is mixed in only one or two component color signals R, G, B of the three component color signals R, G, B will be described. As shown in FIG. 9, the color signals R, G,
If it contains noise component Nb only the color signal B of B, the target pixel OPa, OPb, the achromatic signal S ₂₂ of OPc, due to the influence of the noise component Nb, 9
As shown in (b), the noise component N 'is included. Here, the value n ′ of the noise component N ′ is one third of the noise component Nb.

Then, the noise component N 'is extracted by the above-described subtraction and addition processing, and is subtracted from the color signals R, G, and B of FIG. 9A, as shown in FIG. 9C. Color signals R ', G', B 'are obtained. Among the color signals R ', G', and B ', regarding the color signal B', the noise component Nb 'included in the color signal B' is replaced with the original noise component Nb.
b is about two-thirds of the value, and the influence of noise is improved. However, as for the color signals R ′ and G ′ which originally do not include noise, the color signal R ′ is obtained as a result of the subtraction processing of the noise component N ′. The signal values of 'and G' deviate from the original values by the value n 'of the noise component N'.

However, the color signals R ',
As shown in FIG. 9D, when the achromatic signal S ₂₂ ′ is generated as shown in FIG. 9D based on G ′ and B ′, the noise in the color signal B ′ of FIG. and 'value nb' of components Nb, color signals R ', G' decrement n signal values 'are canceled, an appropriate achromatic signal noise component is removed S _22' is made so as to obtain .

For this reason, the outline emphasis processing section 220 performs the processing shown in FIG.
The achromatic signal generated based on the color signals R ′, G ′, and B ′ of FIG. 9C does not contain the noise component Nb included in the color signal B of FIG. One or two of the component color signals R, G, B,
Even when noise is mixed only into B, the noise is not enhanced by the outline enhancement processing unit 220. Therefore, the mixing ratio of the color signals in the achromatic color component extraction processing of the filtering device of the present embodiment and the contour enhancement processing unit 2
It is desirable that the mixing ratio of the color signals in the achromatic component extraction processing of No. 20 is the same.

As described above, according to the present embodiment, the noise contained in the color image 10, particularly the noise of the achromatic component, can be removed, and the achromatic component is extracted from the R, G, B color components. In the filter processing, the extraction of the intermediate value having a large processing load is performed based on the component value D of the achromatic color component.
Compared to the case where the median filter processing is individually performed for each of the G and B color components, the load of the arithmetic processing can be greatly reduced, and the speed of the filter processing can be nearly tripled.

Since the noise of the achromatic component has been removed, the noise is not emphasized by the outline emphasis processing in the outline emphasis processing section 220.

Further, the filter unit 24 extracts a representative value D 'by a median filter process for extracting an intermediate component value D as a representative value D' from the achromatic component values D of the pixels in the filter area F. Accordingly, noise can be removed while preventing blurring of the color image 10 due to the filter processing.

Since the achromatic component is extracted by averaging the pixel values of each of the R, G, and B color components, a simple averaging or a weighted averaging is performed according to the noise included in the color image. The achromatic color component can be extracted by changing the weighting factors Ar, Ag, and Ab multiplied by the pixel value of each color component, and can cope with the removal of various types of noise.

Next, a filtering device according to a second embodiment of the present invention will be described. The filtering device of the present embodiment is a filter that extracts a representative value D ′ by averaging the component values of the achromatic components of the pixels in the filter area F instead of the filter unit 24 of the filtering device of FIG. Except that a unit is adopted, it is the same as the above-described filtering device, and the corresponding parts are denoted by the same reference characters and their illustration and description are omitted.

The filter unit of the present embodiment averages the achromatic signal S _ij of the pixels in the (2m + 1) × (2n + 1) filter area F input from the achromatic signal generation unit 20 to obtain the achromatic component. A representative value D ′ is extracted, and a signal S ′ having the representative value D ′ as a signal value is added to each addition processing unit 2.
5, 26, 27.

[0047] That is, each pixel in the filter region F, as shown in FIG. 10, k _-mn to k _mn and weighting factors assigned, the achromatic signal S of each pixel in the filter region F is multiplied by the weighting factor k _-mn to k _mn corresponding to the signal value s _ij of _ij,

[0048]

(Equation 2)

As described above, the representative value D 'is extracted by averaging, in this case, weighted averaging. FIG.
1 shows an example of the weighting factor k _-mn to k _mn when the filter region F is a matrix of 3 × 3.

The signal S 'output from the filter unit is free from the influence of noise by the averaging process of the above equation (2). By sending this signal S' to each of the adding units 25, 26 and 27, Attention pixels OPa, OPb, OP
The noise components included in the color signals R, G, B of c are removed.

Also in the present embodiment, the noise contained in the color image 10, particularly the noise of the achromatic component, can be removed, and the achromatic component is extracted from each of the R, G, B color components, and the processing load in the filtering process is increased. Is performed on the basis of the component values of the achromatic color components. Therefore, compared to the case where the averaging process is individually performed for each of the R, G, and B color components as in the conventional example, the calculation process of the arithmetic process is large. The same effects as those of the first embodiment can be obtained such that the burden can be greatly reduced, and the speed of the filter processing can be nearly tripled.

Further, in the present embodiment, the filter section extracts the representative value D 'by averaging, so that the processing is performed in comparison with the method of extracting the intermediate value by the reordering operation of the first embodiment. This can be performed at high speed, and the occurrence of unnatural occurrence of unnatural pixel values between adjacent pixels due to the filter processing can be prevented.

In each of the above embodiments, R, G, B
Although the achromatic color component is extracted by averaging the pixel values of the respective color components, the maximum value or the minimum value among the pixel values of the three component pixels of R, G and B corresponding to each other is extracted. A pixel value as a value may be selected, and the selected pixel value may be extracted as an achromatic component.

In each of the embodiments described above, noise removal of a color image formed by additive color mixing has been described. However, noise removal of a color image formed by subtractive color mixing is also performed based on the pixel value of each color component. The same can be done by extracting an achromatic component with each method.

[0055]

As described above, according to the first and fifth aspects of the present invention, it is possible to remove noise contained in a color image, particularly noise of an achromatic color component, and to remove noise from R, G, and B color components. Since the achromatic components are extracted and the filtering process with a large processing load is performed based on the component values of the achromatic components, the median filtering process is individually performed for each of the R, G, and B color components as in the conventional example. , Or the load on the arithmetic processing can be greatly reduced as compared with the case of performing the averaging processing, and the filtering processing can be speeded up nearly three times.

Further, since the noise of the achromatic component is removed, the noise is not emphasized by the subsequent outline emphasis processing.

According to the second aspect of the present invention, in the filtering process of the filter means, an intermediate component value is extracted as a representative value from the achromatic component values of the pixels in the filter area. In addition, noise can be removed while preventing blurring of a color image due to filter processing.

According to the third aspect of the present invention, in the filtering process of the filter means, the representative value is extracted by the averaging process, so that the process is performed at a higher speed than the method of extracting the intermediate value as the representative value. In addition to this, it is possible to prevent the occurrence of an unnatural drop in pixel values between adjacent pixels due to the filtering process.

According to the fourth aspect of the present invention, since the achromatic component is extracted by averaging the pixel values of each color component, a simple averaging is performed according to the noise included in the color image. Achromatic components can be extracted by changing the weighting factor applied to the pixel value of each color component by employing a weighted average, and can cope with the removal of various types of noise.

[Brief description of the drawings]

FIG. 1 is a diagram showing an overall configuration of an apparatus to which an embodiment of the present invention is applied.

FIG. 2 is an explanatory diagram illustrating an outline of a filtering process of the color image filtering device according to the first embodiment of the present invention.

FIG. 3 is a block diagram of a color image filtering device according to the first embodiment of the present invention.

FIG. 4 is a diagram showing an achromatic signal of a pixel in a filter area.

FIG. 5 is a diagram showing a state in which the signal values of the achromatic signal of FIG. 4 are arranged in order based on the magnitude relation.

6A is a diagram illustrating a color signal of a target pixel of each color component input to the device of FIG. 3, and FIG. 6B is a diagram illustrating a signal output from each subtraction processing unit; And FIG.
(C) is a diagram showing a color signal from which noise has been removed and output from each addition processing unit.

FIG. 7 is a graph showing achromatic signals of pixels in a filter area.

FIG. 8 is a diagram showing how noise components included in an achromatic signal are extracted.

9A is a diagram illustrating a color signal of a target pixel of each color component input to the apparatus of FIG. 3, and is a diagram illustrating an achromatic signal of the target pixel; FIG. FIG. 9D is a diagram illustrating a color signal output from each addition processing unit, and FIG.
FIG. 3C is a diagram illustrating an achromatic signal generated from the color signal of FIG.

FIG. 10 is a diagram illustrating weighting factors assigned to each pixel in a filter area.

FIG. 11 is a diagram illustrating a specific example of the weighting factors of FIG. 10;

FIG. 12 is a diagram showing a conventional filtering device.

[Explanation of symbols]

 Reference Signs List 20 achromatic signal generation section 21, 22, 23 subtraction processing section 24 filter section 25, 26, 27 addition processing section F filter area OPa, OPb, OPc Target pixel

Claims (5)

[Claims] 1. A predetermined filter area is set on a color image formed by superimposing a plurality of color component monochromatic images so as to include a target pixel to be subjected to a filtering process. A color image filtering device that performs a filtering process and sequentially scans the target pixel and the filter region on the color image, thereby removing a noise component included in the color image. Achromatic component extraction means for extracting an achromatic component for each pixel based on the pixel value of each pixel; and a predetermined filtering process based on the achromatic component values of the pixels located in the filter area. Filter means for extracting a representative value representing a component value of an achromatic component of a pixel in the filter area; Subtraction means for subtracting, from the elementary pixel value, the component value of the achromatic component extracted by the achromatic component extraction means of the target pixel; and calculating the subtraction means by adding the representative value extracted by the filter means to the calculated value of the subtraction means. A color image filtering apparatus, comprising: an adding unit that performs addition. 2. The filter unit extracts, from the component values of the achromatic component of a pixel located in the filter area, a component value having an intermediate value in the filter area as the representative value. The color image filtering apparatus according to claim 1, wherein the filtering is performed. 3. The method according to claim 1, wherein the filter unit extracts, as the representative value, an average value obtained by averaging component values of the achromatic components of pixels located in the filter area. A filtering device for a color image as described in the above. 4. The color image extracting apparatus according to claim 2, wherein the achromatic component extracting means extracts an achromatic component by averaging pixel values of pixels of the respective color components corresponding to each other. Filtering device. 5. A predetermined filter region is set on a color image formed by superimposing a plurality of color component monochromatic images so as to include a target pixel to be subjected to a filtering process. A color image filtering method for performing a filtering process and sequentially scanning the target pixel and the filter region on the color image, thereby removing a noise component included in the color image. Based on the pixel value of each pixel, an achromatic component extraction step of extracting an achromatic component for each pixel, and a predetermined filtering process based on the achromatic component values of the pixels located in the filter area, A filtering step of extracting a representative value representing a component value of an achromatic component of a pixel in the filter area; A subtraction step of subtracting, from the elementary pixel value, the component value of the achromatic component of the target pixel extracted in the achromatic component extraction step, and calculating the subtraction step by adding the representative value extracted in the filtering step to the calculated value of the subtraction step A color image filtering method, comprising: an adding step of adding.