JPH03117269A - Halftone picture estimation device - Google Patents

Abstract

PURPOSE:To avoid hardware constitution from being complicated by providing plural decoding use threshold matrix storage means with different aperture and a density pattern decision circuit for scanning aperture selection having a generating means of density pattern for halftone picture estimate. CONSTITUTION:Selection information of a threshold matrix MTX and position information of scanning aperture are inputted to a storage means 68. The threshold MTX used from now is designated and an MTX data with the same aperture as scanning aperture D-A is selected and output relating the scanning position of the scanning aperture among the MTX whose position information is used and designated. The reference threshold MTX is compared for each picture element with white level picture element numbers d-a in the scanning aperture at comparators 66d-66a and based on the comparison output, a decoding density pattern is formed by density pattern generating means 67d-67a. Together with an original dither picture, both pattern picture elements are compared by coincidence decision circuits 62d-62a to decide the coincidence. Thus, even when number or aperture of the decoding threshold MTX is increased, the cost increase is suppressed.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a halftone image estimating device that can satisfactorily estimate an original halftone image from, for example, a halftone displayed binary image, and particularly relates to This is a simplified circuit configuration for scanning aperture selection used in

[Background of the Invention] Currently, many output devices such as display devices and printing devices that are in practical use can only be expressed in binary values of white and black.

A dither method, which is a type of area gradation method, is known as a method of expressing a halftone image in a pseudo manner using such an output device.

As shown in Figure 6, the dither method uses a predetermined threshold value and a large size (for example, about 8 x 8 pixels) as shown in Figure 6 (B) for the original halftone image (Figure 6 (A)). A dithered image (original dithered image) that is a pseudo halftone image as shown in FIG. It was designed to do so.

If it is possible to estimate the original halftone image from such a dithered image, it is possible to perform various data processing based on the estimated halftone image, and it is possible to have a degree of freedom in image conversion, which is convenient. Good.

To estimate a halftone image from a dither image, a predetermined large matrix (hereinafter referred to as a scanning aperture) is prepared, and the number of white or black pixels existing within this scanning aperture and predetermined requirements (described later) are calculated. Under the condition that the condition is satisfied, the halftone level of the pixel of interest is estimated, and this is sequentially executed for each pixel.

At this time, the scanning aperture is estimated by moving one pixel at a time in the row and column directions (main scanning and sub-scanning directions), and this is executed up to the last pixel to estimate a pseudo halftone image relative to the original halftone image. .

When estimating a halftone image using a scanning aperture in this way, rather than using scanning apertures with the same area, it is better to prepare multiple scanning apertures with different aperture areas in advance and adjust the frequency characteristics of the original dithered image. If the scanning aperture with the most suitable aperture area is selected accordingly and the halftone image (halftone level) of that pixel is estimated, a halftone image closer to the original halftone image can be estimated.

This has a high pixel level gradation discrimination ability in the low spatial frequency region (region with few pixel level changes), but only a low pixel level gradation discrimination ability in the high spatial frequency region (region with many pixel level changes). The scanning aperture is selected such that a large scanning aperture is used in the low spatial frequency region and a small scanning aperture is used in the high spatial frequency region.

A specific example of estimating a halftone image based on such a principle has already been proposed by the applicant (for example, in Japanese Patent Laid-Open No. 63-267571, etc.), so a detailed explanation thereof will be omitted. As the scanning apertures, scanning apertures A to G and Z having eight different aperture areas as shown in FIG. 7 are used.

The black circles shown in each scanning aperture of A-Z are the centers of movement when moving on the dithered image in FIG. 6(c). Although the scanning aperture Z can be selected at the upper left pixel with respect to the center of movement, it may be selected at any position at the lower left, upper right, or lower right. By specifying the position of the scanning aperture in this manner, the position of the estimated pixel is also specified.

When a halftone image is estimated from the dithered image of FIG. 6(C) using such a scanning aperture, a halftone image as shown in FIG. 8(B) is obtained. This is very close to the original halftone image shown in FIG. 6(a).

The scanning aperture for each pixel used when estimating the halftone image in FIG. 8(A) is shown in FIG. 8(B). The pixel in the 1st row and 1st column uses the scanning aperture, the pixel in the 1st row and 2nd column uses the scanning aperture D, and the pixel in the 1st row and 3rd column uses the scanning aperture C.

When estimating a halftone image, the aperture area of the scanning aperture can be taken into consideration depending on the spatial frequency domain of the image.

In other words, as mentioned above, a large scanning aperture is used to express high gradations in low spatial frequency areas such as gradation planes, and a small scanning aperture is used to achieve high decomposition in high spatial frequency areas such as line drawings and character drawings. Make it possible to reproduce even images of power.

The specific selection condition for the scanning aperture in the low spatial frequency region is that there is no change in density according to a predetermined conditional expression. On the other hand, in a high spatial frequency region, the optimum aperture is selected based on the coincidence or mismatch of density patterns.

Therefore, the operation for estimating the halftone image is as shown in FIG. 9.

First, a scan aperture with an intermediate aperture area is used as a reference scan aperture (step k). Next, it is checked whether the density patterns of the scanning apertures match or do not match (step 11). When the density patterns match, one of the scanning apertures DG is selected based on a predetermined conditional expression (step m).

When the density patterns do not match, the process moves to step n, and one of the scanning apertures C to A and Z can be selected by density pattern discrimination.

When the selection of the scanning aperture is completed, the level of the halftone image is estimated based on the scan aperture obtained (step 0). Then, this estimated halftone Ll! Image processing such as enlargement/reduction is performed based on the level of image quality, and the image is binarized to complete the halftone image estimation process (steps p, q).

The stepwise processing operation described above is as follows.

The criterion for selecting a scan aperture is the condition that there is no change in pixel level within the selected scan aperture. Particularly in the case of a high spatial frequency region, that is, an image such as a line drawing or character image, the scanning aperture A to C is smaller than the reference scanning aperture.

Z is selected. At this time, which scanning aperture to select can be determined depending on whether the condition that there is no change in the density pattern in that scanning aperture is satisfied.

First, move the scanning aperture to the right from the initial position in FIG. 6(C) (see FIG. 10).

For example, when the center of movement of the scanning aperture is (4, 6), the process is as shown in step (1) in FIG. If the scanning aperture is overlapped with the center of the aperture aligned with the moving position (4°6) in Fig. 6(c), a dithered image as shown in (b) will be obtained, and the self-image in this dithered image will be The number of pixels is 4.

Since the gain of the scanning aperture is 4, the value multiplied by the gain is 1
Assuming that 6 is the average pixel level, each pixel is compensated with 16 as shown in (b).

Here, the gain corresponds to the area ratio of the scanning aperture, and when the gain of the scanning aperture G is 1, the scanning aperture E,
F is gain 2, scanning aperture is gain 4, scanning aperture B, C
has a gain of 8, scanning aperture A has a gain of 16, and scanning aperture Z has a gain of 64.

Then, when the average pixel level shown in (0) is binarized using the threshold value matrix shown in (c), it becomes as shown in (d).

Here, when comparing the dither image (A) and the binary image (D), they do not have the same density pattern.

Therefore, the process proceeds to step (2) in FIG. The scanning aperture used at this time is C.

The number of white pixels within this scanning aperture C is counted as 2. Since the gain of the scanning aperture C is 8, assuming that 16, which is the value multiplied by the gain, is the average pixel level, each pixel is compensated with 16 as shown in (b).

When the average pixel level image shown in (b) is binarized using the restoration threshold matrix shown in (c), it becomes as shown in (=).

Here, when the dithered image (A) and the binary image (D) are compared, they have the same density pattern.

Therefore, in this case, scanning aperture C is selected as the optimal aperture, and the estimated value of the halftone image at that time is 16.

Note that when the scanning aperture is not selected in the -th step,
The processing operations described above can be performed sequentially using the next scan aperture. The scanning aperture selection order is as shown in FIG.
Take the order of B-+A-+z.

When the density patterns do not match even with the scanning aperture A, the smallest scanning aperture Z is selected as the scanning aperture at that time, and the gain multiplied value 64 or O is used as the estimated value of the halftone image.

In the estimation process described above, the density pattern shown in (d) may be calculated each time from the conditions (b) and (c) as described above, but the restoration threshold matrix is known information and The scanning position information of the aperture is also evident. Furthermore, since the number of white pixels in (a) only has values from 0 to 16, the type of density pattern shown in (d) can be known in advance from these combinations.

Therefore, if this density pattern information is stored in advance, the number of white pixels and position information of the scanning aperture can be used as an address.
The corresponding density pattern can be easily referenced.

An example of a density pattern obtained from the number of white pixels and the position information of the scanning aperture will be explained with reference to FIG. 13 and subsequent figures. The reference scanning aperture can be used as the scanning aperture.

The threshold matrix shown in FIG. 6(b) is also used as a restoration threshold matrix for restoring a halftone image. If this threshold matrix is simply arranged vertically and horizontally, the result will be as shown in FIG.

Then, set the position (4, 4) to the origin (Hx * V y
= O*0), and if the scanning aperture is moved by 4 pixels to the left and right, the respective scanning positions (Hx, Vy = Ω).

0) to (Hx, Vy=3.3), the restoration threshold matrix is as shown in FIG.

Similarly, scanning position (Hx, Vy"0.4) ~ (Hx.

The restoration threshold matrix at Vy=3.7) is as shown in FIG. Scanning position (Hx, Vy=4°4) ~(H
x, Vy = 7.7), the restoration threshold matrix is the 16th
as shown, and the scanning position ()-1x, Vy=4
．． O) ~(Hx, Vy=7.3) and f) The threshold matrix becomes as shown in FIG.

By scanning from FIG. 14 to FIG. 17, the entire 8×8 threshold matrix has been scanned.

Next, halftone levels as shown in FIG. 11(b) are determined for the restoration threshold matrix at each of these scanning positions. In the case of a scanning aperture, the value of the number of white pixels is any one of 0 to 16 (0 to 64 when multiplied by the gain). When the dithered image shown in FIG. 11 (=) when the halftone level is 4 (the number of white pixels is 1) is created, it will be as shown in FIG. 18.

This FIG. 18 shows only the pattern obtained when the level is compared with the restoration threshold matrix 1611 shown in FIG. 14.

The pattern shown in FIG. 18 in which the levels have been compared in this manner is used as the density pattern for restoration. A similar density pattern can be created using each of the restoration threshold matrices shown in FIGS. 15 to 17.

Among these density patterns, there are also the same density patterns, so regarding the scanning aperture, the number of white pixels is 0 to 16017, so 17 x 4 (main scanning direction) x 4 (sub-scanning direction). (scanning direction) = 272 density patterns can be created.

Density patterns for restoration are obtained for scanning apertures C, B, and A in a similar manner. That is, scanning aperture C: 9X
8=72 scanning apertures B: 9×8=72 scanning apertures C: 5×4=20 density patterns are created.

In addition, as the position information of the restoration threshold matrix, Hx force direction Q, i, 2.3 HY direction’On 1.2.3 It is sufficient to repeatedly use the position information (2 bits).

Therefore, if the position information of the restoration threshold matrix synchronized with the scanning of the scanning aperture and the number of white pixels at that time are known, the restoration threshold matrix can be specified, and the density pattern related to that number of white pixels can be immediately referred to. becomes possible.

Therefore, the detailed explanation will be left to the description of the present invention, but the restoration density pattern and original density pattern determination circuit 60 can be configured as shown in FIG. 19.

Reference numerals 61a to 61d represent storage means including a ROM in which restoration density patterns used for each scanning aperture, C, B, and A are stored. The number of white pixels d, c, and be a in each scanning aperture, C, B, and A, and position information are supplied to this as a reference address.

The position information generating means 65 is composed of two 2-pit counters 66 and 67 as shown in the figure, and one of the counters 66 is fully supplied with the pixel clock CK and the main scanning signal HV, and from this the position information in the main scanning direction ( 2 bits) is obtained.

The other counter 67 is supplied with the main scanning signal HV and the sub-scanning signal 43 vv, and from this the positional information in the sub-scanning direction (2
bit) is obtained.

The density pattern of each scanning aperture DA specified by the position information is supplied together with the original dithered image to match determination circuits 62d to 62a, where it is determined whether the patterns match or do not match.

[Problem M to be Solved by the Invention] Incidentally, in the above description, an 8×8 matrix is shown as the threshold matrix for restoration, as shown in FIG. 6(b).

When only one threshold matrix is used in this way, the number of levels of the estimated halftone image obtained is also fixed.

However, as threshold matrices for this restoration, multiple threshold matrices with different areas are prepared, and this threshold matrix is selected depending on the situation or specified from outside, and the selected threshold matrix is used to create a halftone image. By estimating , the number of levels of the halftone image can be set according to the area of the threshold matrix used.

As a plurality of threshold matrices with different areas, the 20th
In addition to the 8x8 threshold matrix shown in Figure B (same as Figure 6 (b)), the 4x4 threshold matrix shown in Figure 20A and four 8x8 threshold matrices are used to create a 16x1
6 may be used as a threshold matrix (C in the same figure).

When a plurality of threshold matrices are selectively used in this manner, density patterns corresponding to the respective threshold matrices must be created and stored in the density pattern ROMs 61a to 61d described above.

However, as the number of threshold matrices increases, the number of density patterns to be stored also increases, and accordingly the capacity of the ROM must be increased accordingly. This is difficult from a cost standpoint.

Therefore, this invention takes these points into consideration, and is devised so that even when multiple threshold matrices are selectively used, it can be realized without complicating the hardware configuration too much. .

[Means for Solving the Problems] In order to solve the above-mentioned problems, in the present invention, a plurality of scanning apertures having different aperture areas are selected according to predetermined density pattern information, and an n-value image such as a dither image is generated. (
n≧2), the halftone image estimating device estimates an original halftone image based on n≧2), is provided with a density pattern determination circuit for scanning aperture selection, and the density pattern determination circuit has an area of at least 5. a means for storing a plurality of restoration threshold matrices with different values; and a means for storing a plurality of restoration threshold matrices with different values; It has means for creating a density pattern for image estimation, and means for determining whether the estimation density pattern matches the original density pattern, and when the estimation density pattern matches the original density pattern, the scanning aperture is is selected, and the plurality of restoration threshold matrices having different areas can be selected from the outside.

[Function] Until now, the density pattern for restoration was based on the position information of the scanning aperture and the number of white pixels (or number of black pixels) within the scanning aperture.
Since the obtained data is stored in a predetermined ROM as density pattern data, in the present invention, this series of processing is performed by hardware.

Then, instead of storing the density pattern data in the ROM, a plurality of restoration threshold matrices with different areas are stored in the ROM.
A storage means 68 (FIG. 1) such as the one shown in FIG.

The storage means 68 can input threshold matrix selection information and position 2 information, and from these, the scanning aperture D-1 associated with the scanning position of the scanning aperture in the specified threshold matrix.
Matrix information regarding A is selectively output.

The referenced threshold matrix data is sent to the comparators 66d to 66.
6a, the number of white pixels within the scanning aperture is compared for each pixel, and the comparison output is used to create a density pattern 8.
In 7d to 67a, a density pattern for restoration is formed.

The area of the density pattern is the same as the area of the preselected threshold matrix.

This density pattern is supplied together with the original dithered image to match determining circuits 62d to 62a, and the two patterns are compared pixel by pixel to determine whether they match or do not match.

In this way, even if the number of restoration threshold matrices and their areas become large, this can be processed by hardware, so that no increase in cost occurs.

[Example] Hereinafter, a halftone image estimation device according to the present invention will be described in detail with reference to FIG. 1 and subsequent figures.

For convenience of explanation, a specific example of the image processing apparatus will first be described with reference to FIG. 2 and subsequent figures.

In FIG. 2, an image reading device 1 reads an original image and converts it into binary data.

The original image is read using a photoelectric conversion element such as a CCD and converted into an electrical signal. The converted electrical signal is converted to digital data, and after applying shading correction (correction for equalizing CCD output) to this digital data,
The data is converted into binary data (binary data constituting a dithered image) as shown in FIG. 6(c).

The digitized binary data is supplied to halftone image estimation 1[2 and, if necessary, to image memory unit 6.

2 is a halftone image estimating device according to the present invention, which is a means used when reprocessing binary data that has already been binarized. In the case of binary data from the image reading device 1, since it has already become halftone data at the stage of being digitized, it is output to the subsequent stage without any particular processing. In addition to the binary data, the halftone image estimation device 2 is supplied with timing signals necessary for estimation processing.

The halftone image signal is supplied to the image processing means 3 together with the timing signal, and is subjected to enlarging/reducing, filtering processing, etc.
Image processing corresponding to the specified processing mode is executed.

The image-processed halftone image signal is supplied to the binarization means 4, where it is re-binarized using the threshold selected by the threshold selection signal. The threshold selection signal is designated from a control terminal, keyboard, or the like.

Reference numeral 5 denotes a recording device such as a laser printer, which reproduces an image based on binary data output from the binarization means 4 or the image memory unit 6.

The image memory unit 6 is configured so that it can also store the binary data itself obtained from the image reading device r111.

FIG. 3 shows an example of the configuration of FIG. 2 as an image processing system.

The image reading device 1, the halftone image estimation device 2, the image processing means 3, the binarization means 4 and the recording device 5 are connected to the control terminal 13 via first and second interfaces 11 and 12, respectively. Further, the image memory unit 6 is configured to be connectable to the first interface 11 via the system bus 14. 15 indicates an external device.

FIG. 4 shows an example of the halftone image estimation device 2. As shown in FIG.

Since a part of the configuration disclosed in the above-mentioned publication can also be used in this case, a detailed explanation of the used part will be omitted.

Binary data from the image reading device 1 can be supplied to the terminal 2a, and this binary data can be supplied to the line memory section 22 via the first selector 21.

The line memory section 22 is for receiving the binary data sent from the first selector 21 and storing the binary data for each line. Can be configured with line memory. Then, the second selector 23 sequentially selects and stores binary data corresponding to each line in each of these nine line memories Ll-L9.

The line memory for 9 lines is prepared here because the maximum number of lines of the scanning aperture G to be used is 8 lines, and one more line memory is required for real-time processing. be.

Therefore, the second selector 23 selects eight line memories necessary for the current image processing among the nine line memories.

When using eight line memories, it is sufficient to combine them with one latch circuit. At this time, the input binary data is completely supplied to the latch circuit.

In this case, eight lines of binary data can be obtained using the output of the latch circuit and the outputs of the remaining seven line memories excluding the line memory currently being written.

The binary data of each of the eight selected line memories is supplied to the halftone image estimating section 30, and based on this binary data, a unique scanning aperture is selected from among the plurality of types of scanning apertures.

Data 1g-64z indicating the selected scanning aperture is supplied to a selection circuit 24 to estimate the value of the halftone image determined by the pixel level and gain within the scanning aperture.

Various timing signals obtained from the timing generation circuit 25 are supplied to the above-mentioned selectors 21 and 23, as well as the line memory section 22, halftone image estimation section 30, and selection circuit 24, and the data is processed at the necessary timing. Selection and address sending are controlled.

FIG. 5 shows a specific example of the halftone image estimation section 30.

Means 40 consisting of 841 shift registers (for 8 pixels) is provided for setting a plurality of scanning apertures.

50 is a counting circuit, which is provided in the same number as the scanning numerical aperture. This counting circuit 50 (50a to 50g).

50z), when each scanning aperture is set to the pixel of interest whose halftone image level is to be estimated, the number of white or black pixels within the scanning aperture is counted.

Among them, the outputs of the counting circuits 50a to 50c are supplied to the density pattern determination circuit 60, which outputs two images created based on the counted values and two original images within the scanning aperture (FIG. 6(c)). ) are compared.

The outputs of the remaining counting circuits 50d to 50g are supplied to a conditional expression determination circuit 70.

The same number of conditional expression determination circuits 70 as the number of counting circuits are provided, and the determination outputs from each of them and the determination output of the density pattern determination circuit 60 are supplied to the scanning aperture determination circuit 80 to select the only scanning aperture. Ru.

Reference numeral 32 denotes a register for holding the counted number of white pixels a ``-'' z, and its output is supplied to a multiplier 33 where the counted white pixel number is multiplied by the gain of each scanning aperture. The number of white pixels, which is the result of the multiplication, is used as an estimated value of the desired halftone image.

Data indicating the estimated value of the halftone image is supplied to the selection circuit 24, and the estimated value of the halftone image is selected based on the inter-scan selection 8 outputted from the aperture determination circuit 80.

For example, when a scanning aperture is selected, a halftone image estimate value (=4d) obtained by multiplying the number d of white pixels within this scanning aperture by a gain (=4) can be selected.

FIG. 1 shows an example of a density pattern determination circuit 60 which is a main part of the halftone image estimating device according to the present invention, and the storage means 68 stores data of a plurality of threshold matrices having different areas used for restoration. Stored.

For example, the threshold matrix used is as shown in FIG. 20A.
Let the matrices have areas of 4X4, 8X8 and 16X16 as shown in ~C.

If using a 16x16 threshold matrix, 8x
The data obtained using the threshold matrix of 8 is
By combining the four, the pixel shown in FIG. 20C can be estimated.

The selection information of the threshold matrix and the position information of the scanning aperture (the position information of its center pixel) are input to the storage means 68. Then, the threshold matrix to be used is specified from these, and among the threshold matrices that have been specified using the position information, those related to the scanning position of the scanning aperture,
Matrix data having the same aperture area as the scanning apertures D to A are selected and output.

The referenced threshold matrix data is sent to the comparators 66d to 66.
At 6a, each pixel is compared with the number of white pixels d to a within the scanning aperture, and the comparison output (binary data of "1" and "O") is supplied to density pattern creation means 67d to 67a.

The density pattern creating means 67d to 67a form density patterns for restoration based on the comparison outputs. The area of the density pattern is the same as the area of the threshold matrix output from the storage means 68.

This density pattern is supplied together with the original dithered image to match determining circuits 62d to 62a, and the two patterns are compared pixel by pixel to determine whether they match or do not match.

In this way, even if the number and area of the restoration threshold matrices becomes large, it can be processed relatively easily in terms of hardware.

In the above example, a binary dither image was used as the original image, but instead of a binary dither image, a multi-value image (
It is easy to understand that the present invention can also be applied to the case where a ternary dither image, etc.) is used and the halftone image level is estimated.

[Advantageous Effects of the Invention 1] As explained above, in this invention, instead of storing the density pattern data in the ROM, a plurality of restoration threshold matrices with different areas are stored, and the matrix information regarding each scanning aperture outputted from this is stored. A density pattern for restoration is created based on the number of white pixels or the number of black pixels.

According to this, since the density pattern for restoration is created by logical operation processing, even if the number of threshold matrices for restoration increases or the area of the threshold matrix increases, the cost will not increase. It has the feature that it is possible to realize a halftone image estimation device without any trouble.

[Brief explanation of drawings]

FIG. 1 is a system diagram showing an example of a density pattern determination circuit used in a halftone image estimation device according to the present invention, FIGS. 2 and 3 are system diagrams showing an example of a halftone image estimation device, and FIG. The figure is a system diagram of the halftone image estimation device, Figure 5 is a system diagram of the halftone image estimation unit, Figure 6 is an explanatory diagram when creating a dither image from the original halftone image, and Figure 7 is the scanning used. Figure 8 is a diagram showing an example of an aperture, Figure 8 is a diagram showing an example of an estimated halftone image obtained when a scanning aperture is selected, and Figure 9 is an example of density pattern discrimination. FIG. 10 is a diagram showing the relationship between the dither image and the scanning aperture, FIG. 11 is a diagram showing an example of density pattern discrimination, FIG. 12 is a diagram showing the selection order of the scanning aperture at that time, and FIG. 18 are diagrams for explaining density patterns, FIG. 19 is a system diagram showing an example of a density pattern determination circuit, and FIG. 20 is a diagram of a restoration threshold matrix. 1 Red 2 ・ 3 ・ 4 ・ 5 ・ 22 ・ 24 ・ 30 ・ 60 @ 62a-62d ・ 65 ・ 66a-66d ・ 67a-67d fist 68 ・ ・Image reading device・Halftone image estimation device・Image processing means・2 Value conversion means, recording device, line memory section, selection circuit, halftone image estimation section, density pattern judgment circuit, coincidence judgment circuit, position information generation means, comparator, density pattern creation circuit, threshold value matrix storage means

Claims (1)

[Claims] (1) Select multiple scanning apertures with different aperture areas according to predetermined density pattern information to
A halftone image estimating device that estimates an original halftone image from a value image (n≧2) is provided with a density pattern determination circuit for scanning aperture selection, and the density pattern determination circuit includes at least a plurality of density pattern determination circuits having different areas. means for storing a threshold matrix for restoration; and a means for estimating a halftone image in the scanning area by comparing matrix data of a designated scanning area in the restoration threshold matrix with the number of white pixels or the number of black pixels of the original image. and means for determining whether the estimated density pattern matches the original density pattern, and when the estimated density pattern matches the original density pattern, the scanning aperture is selected. A halftone image estimation device characterized in that the plurality of restoration threshold matrices having different areas can be selected from outside.