JPH04156179A - Picture signal processing method - Google Patents

Abstract

PURPOSE:To obtain a sharp dot picture without moire by deciding a mean mask size in response to a pitch of a regular pattern in picture information and using the mean mask size to implement averaging processing. CONSTITUTION:A mask size M in a mask size decision circuit 28 is decided or revised by reading a pitch of a regular pattern of a picture signal data. An averaging processing (unsharpness processing) circuit 32 implements averaging processing being unsharpness processing to a picture signal stored in a 1st line memory circuit 30 based on the mask size M to obtain an averaging picture signal U. A 1st unsharpness signal being an averaged signal is a signal from which a frequency component of a regular pattern is sufficiently eliminated by averaging. Thus, no moire is caused even in the case of reproduction as a dot picture.

Description

DETAILED DESCRIPTION OF THE INVENTION <Industrial Application Field> The present invention relates to an image signal processing method, and more specifically, when performing shading processing on image information including a regular pattern, the image information Determine an averaging mask size according to the regular pattern or the conversion magnification of the regular pattern and input/output, and obtain a clear image without moiré by performing averaging processing with this averaging mask size. This invention relates to an image signal processing method that allows for.

<Conventional technology> For example, in the fields of printing and plate-making, streamlining work processes,
2. Description of the Related Art Image scanning, reading and reproducing systems that electrically process image information carried on a document to create a film original are widely used for the purpose of improving image quality.

This system basically consists of an image reading section and an image recording section. be transformed. Next, the image information photoelectrically converted in the image reading section is subjected to predetermined image processing according to the plate-making conditions, and then converted into an optical signal such as a laser beam in the image recording section to form an image made of a photosensitive material such as a film. recorded on a record carrier. The military record image record carrier is processed by a predetermined developing device and used as a film original for printing, etc.

By the way, in such an image scanning reading/reproducing system,
In order to reproduce the shading of a continuous tone image, stringing processing is performed to convert the continuous tone image into a halftone image. In this case, if the original 1i4 processing has a regular pattern, such as a printed matter that has already been shaded, the regular pattern and the output net structure will interfere with each other due to the shading process, resulting in the inconvenience of causing moiré in the image. There is.

Therefore, as a method to solve this inconvenience, there is a method in which the screen angle when performing the screen processing is adjusted so that the angle at which the occurrence of moire is minimized is selected.
However, it has been pointed out that in this case, it not only requires a great deal of skill to select the optimum screen angle, but also the troublesome work required to do so. Another method has been proposed that blurs the image by changing the aperture size in the shading process, making it difficult to see regular patterns, but the problem is that the image becomes unclear and the quality deteriorates. There is. Furthermore, this method can only be adopted when the aperture size can be physically changed, and its scope of application is considerably limited.

Therefore, in Japanese Patent Application No. 1-165607, the present applicant applied unsharpness processing by averaging to the image information carried on the manuscript, and then performed sharpness enhancement processing by sharpening, thereby creating a regular pattern. This paper proposes an image signal processing method and apparatus that can reproduce a clear moiré-free image from an original image.

<Problems to be Solved by the Invention> The method disclosed in the above application performs unsharpness processing, that is, averaging IA processing, with a fixed mask size, but the mask size corresponds to a relatively fine mesh. is used, so this method can sufficiently prevent the occurrence of moiré in normal originals with a relatively fine screen pitch.

However, if the input document has rough halftone dots, or if magnification is converted and expanded exposure is performed before averaging processing, the input document screen pitch becomes large, and the fixed mask size becomes larger than the input document screen pitch. It becomes smaller. As described above, when the mask size is small with respect to the manual halftone dot pitch, sufficient averaging is not performed, causing moiré to occur in the output halftone dot image and deteriorating the image quality.

An object of the present invention is to solve the above-mentioned problems of the prior art, and to convert image information including a regular pattern according to the pitch of the regular pattern, or according to the pitch of the regular pattern and the conversion magnification of the hitch blade. An object of the present invention is to provide an image signal processing method that obtains a clear halftone dot image without moiré by performing averaging processing using a mask size and then performing sharpness enhancement (sharpening) processing.

Means for Solving the Problems> In order to achieve the above object, an aspect of the present invention averages image information including regular patterns, sharpens the obtained unsharpness signal, When performing halftone processing on the obtained sharpness signal to form a halftone image signal, an averaging mask size is determined according to the pitch of the regular pattern of the image information, and averaging processing is performed using this averaging mask size. The present invention provides an image signal processing method characterized by performing the following steps.

Further, a second aspect of the present invention is to average image information including regular patterns, sharpen the obtained unsharpness signal, and perform hatching processing on the obtained sharpness signal to form a halftone dot image. An image signal characterized in that when forming the signal, an averaging mask size is calculated according to the pitch of the regular pattern of the image information and a human output conversion magnification, and averaging processing is performed using this averaging mask size. The present invention provides a processing method.

In the first aspect, the mask size is preferably the same as the pitch of the regular pattern.

In the second aspect, the mask size is 0.8 times or more 1 times the pitch of the regular pattern and the conversion magnification.
．． It is preferably 2 times or less, more preferably 0
．． It is preferably 9 times or more and 1.1 times or less.

<Embodiment> Next, a preferred embodiment F of the image signal processing method according to the present invention will be described. ! The details will be described below with reference to the accompanying drawings.

In FIG. 1, reference numeral 10 indicates an image reading and recording apparatus to which an image signal processing apparatus that implements the image signal processing method according to the present embodiment is applied.

In this image reading/recording device 10, image information including a regular pattern of the document G (for example, a halftone m image of a printed matter, etc.)
is converted into an electrical signal and then reproduced on the film F as a halftone image.

That is, the document G is configured to be conveyed in the sub-scanning direction in the direction of the arrow by a conveyance means (not shown), and the image information (halftone image) including the regular pattern of the document G is transmitted through the condensing optical system 12. Main scanning is performed in the direction of arrow B by the CCD 14 which is a photoelectric conversion means. The image signal photoelectrically converted by the CCD 14 and subjected to analog correction such as gain correction is converted into an image signal S as a digital signal by the A/D converter 18 based on the main scanning clock φ8 from the clock generator 16. The signal is supplied to an image processing section 20, which is a device that implements the image signal processing method of the present invention.

The image processing section 20 performs CCD shading correction, dark time correction, etc. on the image signal S based on the main scanning clock signal φ8 and the sub-scanning clock signal φt from the clock generator 16, and then performs logarithmic conversion processing, It is subjected to image processing such as gradation conversion processing, magnification conversion processing, smoothing processing, sharpening processing, and hatching processing, and is output to the image recording section 22 as a binarized halftone image signal R. The image recording section 22 converts the halftone image signal R into an optical signal such as a laser beam and guides it onto the film F, thereby recording a halftone gradation image.

FIG. 1iS2 shows an example of the configuration of an apparatus that implements the first aspect of the image signal processing method of the present invention, including the image processing section 20 of FIG. In this case, the image processing section 20 includes a correction circuit 24, a conversion processing circuit 26, a mask size determination circuit 28, a first line memory circuit 30, an averaging processing (unsharpness processing) circuit 32, a second line memory circuit 34, a sharpening A sharpness processing circuit 36 and a shading processing circuit 38 are provided.

The correction circuit 24 performs offset correction for each line to correct the temperature drift of the base of the image signal after A/D conversion by the CCD 14, which is a solid-state image sensor; CCD defective pixel correction;
Shading correction that corrects fluctuations in the amount of received light due to variations in each pixel of the CCD 14 (including fluctuations in illumination light), and corrects base fluctuations for each pixel (which exist even when no light is incident) This type of shading performs dark correction, etc., and makes the received light signal of each pixel uniform with a uniform base.For example, if the original image density is the same, the image signal (image data) is the same.This kind of shading Correction, dark time correction, etc. may be performed in the analog data before the light reception signal by the CCD 14 is A/D converted by the A/D converter 91B. At this time, the correction circuit 24 is provided downstream of the signal transmission of the A/D converter 18. Further, the correction circuit 24 may be arranged after the logarithmic conversion circuit described later, and may correct the CCD signal after the logarithmic conversion.

The conversion processing circuit 26 includes a logarithmic conversion circuit that logarithmically converts an image signal and l! It consists of a gradation conversion circuit that converts into an image signal corresponding to i-tone characteristics (exposure-density characteristics), and converts it into a signal for image recording.

The mask size (M) determining circuit 28 obtains the pitch of the regular pattern of the input image [s], and uses the same mask size (data) Mo as this pitch, that is, the averaging processing circuit 32 performs averaging. The number of pixels M1 to be used is determined, the number of lines to be stored is transferred to the first line memory circuit 30, and the mask size data M1 is transferred to the averaging processing circuit 32.

In the present invention, the mask size M1 used is preferably the same as the regular pattern of the input image signal S, that is, the pitch of the halftone dots, but when averaging, the target pixel is symmetrical about the pixel. It is preferable that it is an odd number.

Therefore, as long as moiré cannot be visually recognized, it is not necessary to match perfectly. Here, for example, in this embodiment, the halftone dot pitches of 150 lines, 100 lines, and 85jiI are mask sizes of 10 (10xlO) and 15 (15x1O), respectively.
5), corresponds to 25 (25×25) pixels, but in this case, it is preferable to use 11.15.25 or the like as the averaged mask size M1.

The mask size Ml can be determined and changed by the mask size determination circuit 28 by simply reading the pitch of a regular pattern from the image signal data, or by manually inputting the pitch 40 by an external operator. It may be something that you receive.

The first line memory circuit 30 is composed of a plurality of line memories necessary for averaging processing (unsharpness processing),
The image signal processed by the conversion processing circuit 26 is held for M11 lines according to the mask size data M1.

For example, assuming that the image signal obtained from the document S is divided into n×n pixels, first, the first line memory circuit 30 follows the mask size M sent from the mask size determination circuit 28, and then Image signals for M1 lines centered around the line including the pixel (i, j), that is, L lines (L=(M-1)/2) before and after the pixel line i, are held.

The averaging processing (unsharpness processing) circuit 32 includes a first
Averaging processing, which is unsharpness processing, is performed on the image signal held in the line memory circuit 30 using a mask size M to obtain an averaged image signal U.

That is, the image signal 5iJ (ixl, . . . n%j=1, . . . n) held in the first line memory circuit 30 is processed by the mask size M
The signal is averaged with the surrounding pixel signals held in the first line memory circuit 30 according to l, and converted into a first unsharpness signal UIJ, which is an electrically blurred image signal.

This first unsharpness signal UIJ has a mask size M
It is obtained by averaging the image signal SIJ with the number of pixels determined by , and the surrounding image signals. , and if L is defined as (MI-1)/2 as described above, then UIJ is UIJ"Σ Σ S k-/M + '
...(1) It is determined as klI mm-L mmJ-L. The averaging according to the present invention is not limited to this, and may be any type of averaging as long as it is performed using the mask size Ml.

In the present invention, the averaging processing circuit 32 performs averaging with a mask size M according to the pitch of a regular pattern included in the image signal S, that is, with the pixel signal SIJ as the center and surrounding pixel signals of MI XM+. Because I will do it,
Even if the image signal includes a frequency component of a regular pattern as shown in FIG. 3a, that is, a variation in pitch M1, the third
As shown in Figure b, the first unsharpness signal U, which is an averaged signal, becomes a signal from which frequency components of a regular pattern have been sufficiently removed by averaging.

Like the first line memory circuit 30, the second line memory circuit 34 is composed of a plurality of line memories, and holds the averaged signal processed in the averaging processing circuit 32. That is, this circuit 34 stores the above-mentioned first unsharp signal UIJ for each main scanning line.

The sharpening processing circuit 36 emphasizes edges such as contours of the image and performs sharpening processing.

In the present invention, since the sharpening processing circuit 36 performs sharpness enhancement processing on the first unsharpness signal U from which frequency components of the regular pattern have been sufficiently removed, edge portions as shown in FIG. is emphasized, and an output image signal (sharpness signal) So without moiré is obtained.

Specifically, sharpness enhancement processing is performed in the sharpness processing circuit 36 to create a pressure image signal (sharpness signal) S'lJ. That is, in the sharpness processing circuit 36, by averaging the image signals around the 341 unsharpness signal UIJ using a separately determined mask size M2, as in the case of equation (1), U'lJ = Σ Σ U k-/M x
' ... (2) ks+l −L mmJ −
A second unsharpness signal U° that is L is created.

In this case, the mask size data M2 is optimally determined in terms of visual characteristics for sharpness enhancement. Next, a difference signal between the first unsharpness signal UIJ and the second unsharpness signal U'lJ is determined.

Then, by multiplying this difference signal by the sharpness parameter K and adding the first unsharpness signal UIJ, S'IJ = U IJ + K ・ (UIJ - U
'muj)... (3) Output image signal S'1. is obtained.

In this case, the image signal SIJ including the regular pattern has the frequency components of the regular pattern averaged and removed by creating a first unsharpness signal U-th in the unsharpness processing circuit 32. Therefore, the sharpness processing circuit 36 performs sharpness enhancement processing on the first unsharpness signal UIJ that does not include frequency components of the regular pattern, so that the edge portion is emphasized.
Moreover, the image signal S°1 without moiré. is generated.

Note that when the image information carried on the document G is an image that does not include a regular pattern, such as a photographic document, for example,
Unsharpness processing circuit 3 forming image processing section 20
2 is not applied, the sharpness processing circuit 36 can perform normal sharpness enhancement processing on the image information. For example, it is possible to perform unsharp masking by subtracting a constant times the averaged image signal U from the image signal S to increase image sharpness and to perform edge enhancement.

The halftone processing circuit 38 converts the halftone image signal R from the image density signal.
This halftone image signal is used to area-modulate the image density according to the required angle and number of lines.

That is, the output image signal 5lIj becomes a halftone image signal R converted by the halftone processing circuit 38 into an on/off signal having a desired halftone dot size based on a predetermined halftone information signal.

This halftone image signal R is output to the image recording section 22.

As described above, conventionally known circuits can be used as the correction circuit 24, the conversion circuit 26, the first and second line memory circuits 30, 34, the sharpening processing circuit 36, and the shading processing circuit 38.

The image reading device that implements the image processing method according to the present invention is basically configured as described above, and its operation will be explained next.

First, as shown in FIG.
It is converted into an electrical signal by scanning in the main direction. Next, the halftone gradation image converted to electric No. 6 is converted into a digital signal by the tk A/D converter 18 based on the main scanning clock φ× from the clock generator 16, and the image signal 5 (
SIJ) is obtained. This "image signal 5 (SIJ)" is transmitted to the image processing section 20, and as shown in FIG.
After performing conversion logic such as singular conversion, gradation conversion, and magnification conversion, the data is transferred to the 'i41 line memory circuit 30 and held for each main scanning line. At this time, since the pitch 40 of the regular pattern of the original G is being performed or the pitch M has been read from the image signal, the mask size determining circuit 28 generates the mask size data M1 that is the same as the pitch M1. Since it is given to the first line memory circuit 30,
The first line memory is line data (image signal for one line)
is held for Ml lines.

In the averaging processing circuit 32, the first line memory circuit 3
The image signal Slj held at 0 is averaged using the mask size data M from the mask size determining circuit 28 according to the above equation ii (1) at the mask size M,
The first unsharpness signal υ is converted into a first unsharpness signal υ in which the frequency components of the regular pattern are sufficiently removed.

Next, the first unsharpness signal UIJ is stored in the second line memory circuit 34 for M2 lines determined separately for each main scanning line.

In the sharpening processing circuit 36, the first unsharpness signal U for M2 lines held in the 'f%2 line memory circuit 34 calculates the mask size M by the above equation (2).
, the second unshape signal U' is averaged by ,
After the signal is converted into a moiré-free output image signal S', the edges are emphasized by unsharp masking according to equation (3).

The output image signal S° which has been subjected to the sharpness enhancement processing as described above is processed by the hatching processing circuit 34 to correspond to the area of the halftone dot in an optimal and high wI degree according to the image density without causing moiré. converted into a pulse width modulated signal,
It is output as a halftone image signal R to the image recording section 22.

The image processing unit 20 includes a magnification conversion circuit that converts pixel signals into pixel signals corresponding to the pixel density in the main scanning direction for enlargement and reduction exposure.
However, the averaging is performed so that the pitch of the regular pattern of the original G, the pitch of the regular pattern of the image signal S to be averaged, and the pitch of the regular pattern of the image signal S to be averaged do not differ. It is preferable to provide it at a stage subsequent to the processing circuit 32.

Next, in the image recording section 22, the halftone image signal R
A moiré-free halftone dot image is formed on the film F based on this.

In the present invention, in the averaging process, averaging is performed using a mask size M1 that is the same as the pitch of the regular pattern of the original G, that is, the pitch (reciprocal of frequency) Ml of fluctuations superimposed on the image signal S. For this reason, conventionally, when the mesh pitch of the human image signal does not match the pitch of the regular pattern and the fixed average mask size,
In particular, even when the pattern pitch is larger than the average mask size (coarse mesh), averaging is performed sufficiently, so the averaged signal U is a signal from which the frequency components of the pattern have been sufficiently removed. . Therefore, in the present invention, since the sharpening process is performed using a sufficiently averaged signal that does not include the frequency component, moiré does not occur when forming a halftone image. .

When magnification conversion is performed before averaging processing as in the second R of the present invention, it is necessary to change the average mask size by considering not only the pitch of the document G but also the conversion magnification.

FIG. 4 shows an image signal processing apparatus that implements the image signal processing method of the second aspect of the present invention.

This image processing device 50 consists of another configuration example of the image processing section 20 shown in FIG.

In this image processing device 50, a magnification conversion circuit 27 is provided between the conversion circuit 26 and the first line memory 30, and a mask size calculation circuit 29 and an input section 41 are connected to a mask size determination circuit 28 and a pitch manual 40, respectively. The configuration is completely the same as the configuration example shown in FIG. 2 except that , and therefore the same components are given the same numbers and their explanations will be omitted.

In FIG. 344, the magnification conversion circuit 27 converts the signal into an image signal corresponding to the pixel density in the main scanning direction according to the given magnification or reduction magnification after being subjected to conversion processing in the conversion circuit 26 during expansion or reduction exposure. It is something.

The input section 41 inputs the pitch of a regular pattern of the document G, for example, the dot pitch, and the magnification of enlargement or reduction, that is, the conversion magnification between the document image and the recorded image. Size calculation circuit 29
transmitted to.

The mask size calculation circuit 29 calculates the number of pixels for averaging processing, that is, the mask size M, (the number of lines M+ to be held in the first line memory 30) from these pitch and magnification data, and calculates the average conversion processing circuit 32
A mask size Ml that is equal or approximately equal to the pitch of the regular pattern of the image signal Sm after magnification conversion immediately before inputting is calculated.

The mask size M used in the present invention is 0.8 times or more the product of the pitch p of the regular pattern of the document G and the magnification μ, and 1
．． It is preferable to determine the ratio to be within a range of 2 times or less, preferably 0.9 times or more and 1.1 times or less. That is, 0.8≦M+/(p*μ)≦1.2 (4) Preferably, 0.9≦M+/(p*μ)≦1.1 (5) It is good that it is.

As mentioned above, since an odd number is generally used as the mask size M, in the present invention, 3.5.7
．． 9.11.13.15.17.19.21.25, etc. can be used, but the present invention is not limited thereto. Even in the case of equal magnification, the mask size M may be determined by the image signal processing device 5° shown in FIG. 4 from the above equations (4) and (5).

The 342nd aspect of the present invention calculates and determines the mask size M in this way, but since the averaging process and other processes are similar to those shown in FIGS. 1 and 2, the effect thereof is The explanation will be omitted.

Incidentally, as the mask size M increases, the number of pixels (MxM) required for averaging increases, and therefore the number of lines to be held in the line memory circuits 30 and 34 increases. Therefore, it is necessary to increase the memory scale of the line memory circuits 30 and 34, but increasing the scale increases the cost, so there is a limit to how large the memory scale can be increased.

Therefore, if the mask size M1 exceeds the maximum memory size of the line memory circuit 30, it is preferable to thin out the data (image signal) and perform averaging processing.

For example, when the memory buffer size of the line memory circuit 3o is 13 lines, to perform averaging with a mask size of 25, either use 89 thinning rate of 2 and buffer size of 13, or use 50 thinning rate of 3 and buffer size of 9. You can use .

The table below shows the thinning rate and the number of used buffers for mask sizes 17 to 25 when the maximum number of memory lines is 13 lines.

For example, in the case of mask size 17, if the image signal of the picture plane is SIJ, the line memory circuit y30 is i-8,
i-6, i-4, i-2, 1% i+2, i+4, i+
6, i+8 line ζ image signal is held, and the first unsharp signal 4UIJ is U IJ" Σ Σ S 1-2k
J-2+a/9'm5-4 Tsll-
4 The second unsharp signal is U'lJ"Σ Σ U l-2k j-2/92
Therefore, the 4S'th output image signal can be obtained using the above-mentioned equation (3).

By using such a method, it is possible to save memory.

In the image reading and recording apparatus shown in FIG.
If the image information contained in the original 84G read in the step includes a regular pattern such as a halftone image that has been subjected to a halftone process, this image information is further subjected to a halftone process, as will be described later. Moiré may occur in the image output on the film F in the image recording section 22.

) For this reason, a method has been adopted in which the image signals obtained from the original G containing regular patterns are once averaged and then sharpened to suppress the appearance of the regular patterns. However, if the regular pattern of the original G is rough, or if the regular pattern of the image signal S to be averaged is large compared to the mask size M, such as during enlarged exposure, the averaging process will not work. Frequency components of this regular pattern cannot be removed, and moiré may occur in the output image.

However, in the first and second aspects of the present invention, since the pitch of the regular pattern of the image signal S to be averaged and the mask size M1 are made to match or almost match,
Since the averaging is sufficiently performed and the frequency components of the regular pattern can be removed, moiré does not occur in the formed image even if the sharpening process and the shading process are performed.

Although the image signal processing method of the present invention has been described above by citing preferred embodiments, the present invention is not limited to these embodiments, and various improvements and designs may be made without departing from the gist of the present invention. Of course, it is possible to change.

<Effects of the Invention> As detailed above, according to the present invention, when averaging processing of image information including a regular pattern, the pitch of the regular pattern or the conversion magnification of the pitch and the input/output image signal is The mask size is determined according to the image information signal, and the pitch of the regular pattern of the image signal to be averaged matches or almost matches the mask size, so that the image information signal can be sufficiently averaged. It is possible to sufficiently remove the frequency components of the regular pattern that have been superimposed on the regular pattern.

Therefore, after sharpening the averaged unsharpness signal, desired halftone processing is performed to obtain a halftone image signal, and even if this signal is reproduced as a halftone image, no moiré will occur.

As a result, it becomes possible to create a high-quality film original plate for printing plate making from image information containing regular patterns.

[Brief explanation of the drawing]

FIG. 's1 is a schematic configuration diagram of an image reading and recording apparatus in which an image processing apparatus in which an image signal processing method according to the present invention is implemented is commonly used. FIG. 's2 is a block diagram of an embodiment of an image processing apparatus including the image processing section shown in FIG. Figures 3a, fiBb and 3c are graphs of examples of an input image signal, an unsharpness signal and a sharpness signal (output image signal), respectively, to which the image signal processing method of the present invention is applied. FIG. 4 is a block diagram of another embodiment of an image processing apparatus in which the image signal processing method of the present invention is implemented. Explanation of symbols 20... Image processing unit, 24... Correction circuit, 26... Conversion circuit, 27... Magnification conversion circuit, 28... Mask size determination circuit, 29... Mask size calculation circuit , 30.34... Line memory circuit, 32... Averaging processing circuit, 36... Sharpening processing circuit, 38... Shading processing circuit, 40... Pitch manual power, 41... Input section F I 0.1 Hand Thread Selling Yamacorrection) (Spontaneous) October 18, 1991 1990 Patent Application No. 281079 2, Title of Invention Image Signal Processing Method 3, Person Performing Correction and Fig. 4 ) 6. Contents of correction (1) "Manuscript processing" in the 17th line of page 3 of the specification is corrected to "manuscript image." (2) In the 20th line of page 5, "deteriorate" is corrected to "deteriorate". (3) The second unsharp signal in the fifth to third lines from the bottom of page 27 is...
Therefore, the above equations (2) and (3) are corrected. (4) Correct the drawings in Figures 2 and 4 as shown in the attached sheet.

Claims (2)

[Claims] (1) Averaging the image information including regular patterns,
When sharpening the obtained unsharpness signal and performing hatching processing on the obtained sharpness signal to form a halftone image signal, an average mask size is determined according to the pitch of the regular pattern of the image information. and performing averaging processing using this averaging mask size. (2) Averaging the image information including regular patterns,
When the obtained unsharpness signal is sharpened and the obtained sharpness signal is shaded to form a halftone image signal, the average is calculated according to the pitch of the regular pattern of the image information and the input/output conversion magnification An image signal processing method characterized by calculating an average mask size and performing averaging processing using this average mask size.