JPS62117074A - Picture processor - Google Patents

Abstract

PURPOSE:To satisfactorily estimate an original halftone picture out of a multi-value picture by using a means which secures the coincidence between the position of a threshold value matrix set when a multi-value picture is obtained and that set when a density pattern is obtained. CONSTITUTION:The value estimated by the ratio between the white and black areas within a selected opening is turned again into the multiple value by the dither matrix corresponding to the size of the opening. An opening G is put on the original scan position and the binary picture of the part enclosed by an opening frame is extracted. Then the number of white picture elements in the opening frame is defined as the estimated value to decide a mean picture element level within the opening frame. Thus a picture (b) where all picture elements are buried is defined as an estimated halftone picture and then binary coded again with the dither matrix corresponding to the size of the opening G. Thus a binary picture (d) is obtained. Here the original binary picture (a) is not coincident with the picture (d) and therefore the next opening F is selected. The binary picture (e) set at the initial position frame by the opening F has 14 white picture elements with the gain (area ratio) of the opening F set at '2'. Thus the estimated value '28' is obtained by multiplying '14' by said gain. The picture (e) is coincident with a rebinary coded picture (h) and therefore the halftone picture obtained here is defined as the estimated value.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to an image processing apparatus, and more specifically, to an image processing apparatus capable of determining a multi-tone image of a cabinet. Regarding equipment.

(Prior Art) Currently, many of the output devices in practical use, such as display devices and printing devices, can only represent binary values of white and black.

The 1131 Maro pattern method (luminance pattern method) is a method for expressing halftones in a pseudo manner using such an output device.
, dither method, etc. are known. The density pattern method and dither method are also types of area gradation methods, which change the number of dots recorded within a fixed area (matrix).

The density pattern method is a method of recording a portion corresponding to one pixel of a document with multiple dots using a threshold value of 7 trixes, as shown in Figure 22 (B), and the density pattern method is a method of recording multiple dots on a portion corresponding to one pixel of the original, as shown in Figure 22 (B). In this method, one dot is recorded in a part corresponding to one pixel of the original, as shown in the figure.A two-fold output arc is created as shown in the figure.This output data is pseudo-white,
A halftone image is expressed using black binary values.

(Problem to be solved by the invention) By the way, if it is possible to restore the original halftone image WA (corresponding to the input data in Fig. 22) from such a binarized pseudo halftone image, it will be a miracle of the gods. Since data processing can be performed, image conversion can have six different degrees of freedom, which is convenient. In the case of a density pattern image, once the pattern level arrangement is known, it is possible to immediately restore the original halftone image.

However, the resolution is low compared to the information j. On the other hand, although the di-IJ' image has a higher resolving power in relation to the information m than the density pattern image, it is difficult to restore it to the original halftone image. Therefore, it is not possible to perform various image conversions using just a digital image.

The present invention has been made in view of the above points, and its purpose is to provide an image processing device that can satisfactorily estimate an original halftone image from a multivalued image (including a 2fiα image). The aim is to realize this.

(Means for Solving the Problems) The present invention solves the above-mentioned problems.16 The present invention solves the above-mentioned problems. , the density pattern created using the threshold matrix is compared with the multivalued image inside the aperture from values obtained based on the ratio of areas of each density within the aperture while moving multiple types of apertures pixel by pixel. , an image processing device 43 having means for creating a halftone image is used to match the position of the threshold (direct matrix) used to obtain the multivalued image with the position of the threshold value matrix used to obtain the density pattern. It is characterized by having means.

(Function) The present invention prepares a plurality of types of apertures for each pixel, performs predetermined processing on the multilevel image in each aperture, and the multilevel image, and then re-multilevels the multilevel image. The most suitable aperture was selected by comparing the two.

(Example) Embodiments of the present invention will be described in detail below with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of the present invention. In the figure, reference numeral 1 denotes an image reading device that reads an original image and converts it into multivalued data. The image reading device reads the original image using a photoelectric converter such as a CCD and converts it into electricity.Then, the converted electrical signal is converted into digital data by Δ/D conversion, This digital data is subjected to shording correction (uniformization processing of COD output) and then converted to multi-value data. 2 is the digital multi-value data (direct data (21 direct-j This is a halftone image source circuit that receives a timing signal and performs predetermined processing in multiple shifts to restore a halftone image signal.

3 is an image processing unit that receives a halftone image signal and a timing signal from the halftone image fiI restoration circuit 2, and performs processing such as scaling and filtering according to a processing mode set from a host computer (not shown); It is a circuit. 4 is a multi-value conversion circuit that converts the halftone signal and timing signal from the image processing circuit 3 and performs multi-value conversion processing using a VA value selected by a threshold selection signal set from a host computer, keyboard, etc.; It is. 5 is a recording device that receives multi-value data output from the multi-value conversion circuit 4 and reproduces it as an image; 6 is a binary data output from the image reading 1A position 1 and/or multi-value data from the multi-value conversion circuit 4; This is an image memory 1- that stores the output. As the recording device 5, for example, a λ laser printer, an LED printer, or the like is used. The operation of the device configured as described above will be explained as follows.

The image recorded on the original is scanned by the image reading device 1 using the CCD
It is read using photoelectric conversion water such as and converted into an electric signal. The image signal converted into an electric signal is also read by the image reader i.
! ! It is converted into gage digital data by the △2/r) 1φ transformer of the i device. The converted f digital data undergoes shading correction for each pixel data, and is converted into digital multi-value data by a multi-value converting circuit in the image reading device 1 and output. The output multivalued data is sent to the halftone image restoration circuit 2. Both are stored in the image memory unit 6 at the same time. The halftone image restoration circuit 2 restores a halftone image from input multivalued data.

The operation of the halftone image restoration circuit 2 will be described below, but before explaining the hardware, the halftone image lI! used in the line halftone image restoration circuit 2! The restoration method will be explained.

FIG. 2 shows an embodiment of the method of the present invention in dimensions [1-chip-p-
It is h. The method of the present invention will be explained with reference to this flowchart.

Step (2) Set and cover multiple expansion apertures for each pixel in a multivalued image consisting of different 1fi1 degree areas.

Figures 3 (a) to (g) each represent two multivalued images.
If: An image stage is used and binary images [ ] are shown superimposed. 8 is 2 rows x 2 columns (2 x 2), (
(b) shows 1' B 4 2 rows x 4 columns (2X4), (c) shows 4 rows x 2 columns IX2>, (d) D shows 4 rows x 4 columns (4 ), (e) shows 113 (
4 rows x 8 columns (4 x 8), (to) (shown 4F is 8 rows x 4
41 (8x4), G shown in <1-) is 8 rows x 8 das 1 (
8x8) each shows an opening. Here, the black circles shown in J in each interval in the figure are the centers of movement when moving on the two-layered image, and the halftone image at this point is estimated. The following will explain the case of 2(la both images) as an example.

The present invention selects the optimal aperture among these multiple types of apertures.
However, in selecting the most optimal period, it is necessary to consider the following heat. In other words, human IT is in the low spatial frequency region (region 1 with few pixel level changes).
In 1i), it has a high gradation discrimination OL ability, but has only a low gradation discrimination ability in a high spatial frequency region (region where there are many changes in pixel level). Therefore, a, (a large noise in the spatial frequency domain [
If J5 is used in the high spatial frequency region to express high gradation, and pixels with high resolution are reproduced using a small aperture in the high spatial frequency region, a halftone image of 8 quality can be obtained as a whole.

Step ■ First, select the maximum aperture G.

As explained in step (3), the basic idea of the present invention is to select the interval [] as large as possible unless a change in concentration is observed within the aperture.

Therefore, here, the order of aperture selection is returned to O→[-→l]-→C→B-)A as shown in FIG. 4.

Step ■ The ratio of the white area to the black area in the selected aperture is i, : I; The constant value of if with t is set to II'l, and this estimated value is multivalued again using the di1F matrix by subtracting 1b from the size of the aperture in question. do.

Place the hand at the initial position of the skip as shown in the opening G @ Figure 3 (G), and take out the two octopus images in the area I surrounded by the opening frame, as shown in Figure 5 (A). L, I'm screaming. Now, there are 26 white pictures in this opening frame. Therefore, assuming that this 2G is estimated (self) and is the average level of both prefectures for all pixels existing within the aperture frame, Figure 5 (
2) Fill all pixels with 26 as shown in 1. The image shown in FIG. 5(b) is an estimated halftone image.

In this way ■ [Constant halftone both elephants are 1! P is 1. = et al. Next, this halftone image is re-binarized using a dynamometer 1j corresponding to the depth of the aperture G as shown in FIG. 5(c). For example, (b) 1f of the 1st row and 1st column of a halftone image showing snoring.
Feedback 2G of fl (1, 1) and Δ゛ dizama 1 shown in (c)
- Comparing (45 nights) with the same risk <<1.1> (
Since 1) of [1] is small, the pixel at (1°1) is set as black. Comparing l1fi 26 of ([1) of next km (1, 2>) and value 5 of (c), the right side of ([1) is larger, so (1,
2) pixel 4 white IJ4. In this way, by converting one halftone image into ilT 211I'l according to C (mouth), Figure 5 (
d) is shown in t, k 5 to 2 (``One image is 1!゛1.

Step (4) Atom (1fi image and 1 multivalued image 41- are defeated 1.:
Taking the case of FIG. 5 as an example, compare the original binary image shown in (i) with the direct image shown in (d).

In this case, as is clear from the figure, there is a mismatch.

A mismatch means that there has been a change in pixel level within this aperture G.

If they do not match, it means that the aperture G is not appropriate, so the next aperture is selected (step 2).

The selection order of the apertures is as shown in FIG. Therefore, the next aperture to be selected is F. Once aperture F is selected,
Internuclear 11 Step (Ji returns the operation of the conflict) to F.

FIG. 5(E) is a binary image of the initial position framed by the aperture F. If we count the number of white pixels within this frame, there are 14.

Since the gain of the aperture F is 2, the gain is 28 which is 14 multiplied by the gain.
is the if constant value here.

Here, gain refers to the aperture used! 5 The area of the largest opening is divided by the area of the opening in question. For example, if we calculate the gain of 4 to the aperture, we get the following. The area of the maximum opening G is 8 x 8 64, and the area to the gap 10 is 2 x 2
4, so the gain to the aperture is 64. /4=16. The numbers written under each aperture in FIG. 3 indicate the gain of that aperture. Such gain correction is performed to match the gradation characteristics of each aperture.

Assuming that the obtained estimate 1lIII28 is the average pixel level of the binary image shown in (e), all pixels are compensated with 28 as shown in FIG. (f) becomes the estimated halftone image in this case. Once the estimated halftone image is obtained, this halftone image is re-binarized using a dither matrix corresponding to the size of the aperture F as shown in FIG.
A binary image as shown in FIG. 5(H) is obtained.

Next, the original binary image (e) and the re-21lfItiji image (chi)
Compare. From the diagram, it is clear that the two ancients match. This shows that there is no change in pixel level within the aperture F.

Therefore, open [1F is appropriate.

Step ■ When the original 2 ffi image and the re-binary image match, the aperture F used at that time is selected on the same day, and the estimated value (28 in this case) obtained using the aperture is used as the halftone image l of the center point pixel.
It is set as a fixed value. The value 28 shown in FIG. 5(f) is the estimated value that should be obtained as is.

In this way, the optimal aperture is selected for all pixels,
The operation of estimating a halftone image based on the optimum aperture is
1, high-quality image estimation is performed for all pixels. Therefore, if an image is reproduced η on a recording device based on the estimation 1 obtained in this way, high-quality pixels will be (J).

In the comparison between the original binary image and the re-211α image shown in step (4), there may be a case where the two images do not match for all the apertures prepared in advance. In this case, if you select the closest opening (here, t and △), you can exit the loop shown in FIG. 2.

Next, we will explain the specifics of the second halftone image restoration circuit (1iS2) to realize such a halftone image restoration method. Fig. 6 shows the halftone image restoration circuit. 2 is a diagram illustrating an example of a shell configuration of 2. In the figure, J'3 indicates the image (2 fi (+) of the area capture device 1); The select circuit 21 is a line memory section that receives binary data sent from the select circuit 20 and stores data for each line.

As shown in the figure, the line memory section 21 is composed of nine lines L+ to L9. Therefore,
The circuit shown in the figure can simultaneously store 21 direct data for 9 lines. Here, the reason why nine lines of line memory are prepared is because the number of lines of the maximum aperture G (see FIG. 3) is eight lines, and one additional line is required for real-time processing.

22 is the current one of the 9 lines in the line memory section 21! I
A second select circuit 23 selects the eight lines of data necessary for the process, and 23 receives the data output from the select circuit 22 and selects the estimated halftone image value and original binary value for each aperture. A halftone lit unit that outputs a comparison of the image and the two field images. Reference numeral 24 denotes a selection circuit which receives comparison result information of the constant value and both binary images for each aperture output from the halftone HE setting section 3, selects the optimum constant value 11, and outputs it as a halftone signal. .

Reference numeral 25 indicates various timing signals (synchronous clock, +-+-VΔLID) output from the image gt reading device 1.

Receive V-VAL [D, H-8YNC) (edge first and second select circuits 20, 22, line memory section 21
．． Timing (f
This is a timing generation circuit that outputs the f1 number (address in the case of the line memory section 21).

Here, the synchronization clock is a clock (main scanning IfJfa number) that is output to 1 data 1H of binary data.
3YNC is a clock (sub-scanning amount lit] signal @) output for each line. 1-VALID is an enable signal indicating the valid data width in the main scanning direction.

V-VALID is an enable signal indicating the effective data width (original reading width) in the sub-scanning direction. The timing table showing the mutual relationship of these timing signals is shown in the seventh column.
As shown in Fig. and Fig. 8. FIG. 7 shows a timing chart in the main scanning direction, and FIG. 8 shows a timing chart in the sub-scanning direction.

7th and 8th, a timing chart not shown in FIG. 8 will be explained. In Figure 7, (a) is the H-3YNC signal Mi
(D > is H, , -V Δ I-
ID (, i number, <A) is a synchronous clock,
(D) is the image information 61" - One scanning time (COD exposure time) from the rise of the -8YNC pulse to the rise of the next pulse, and 11-V A L [S of the D pulse
From the falling edge of γ to the rising edge of the next pulse is image data (there is an interval between signals). Image information is established on the bus every 1 pulse of the synchronization clock. In FIG. 8, (a) is the document reading start pulse. , (mouth) is H-8YNC signal, (c)
is the V-VALID signal. The document reading width is from the fall to the rise of the v-vΔLID signal.

In other words, the timing generation circuit 25 receives such various timing signals and controls the operations of the internal circuits. The operation of the circuit configured as described above will be explained as follows.

Upon receiving the 21 direct data for 8 lines sent from the image reading station 1 and the timing signal from the timing generation circuit 255, the select circuit 20 sequentially distributes the binary data and inputs it to the line meters L1 to L9. . For example, when inputting to L2 memory, and when L2 memory is full, next L3 memory is input.
The 21st shift data is inputted by sequentially switching to the memory. The selection circuit 22 selects eight lines of data necessary for the current processing from the line memory of the line memory section 21 and sends the selected data to the subsequent halftone estimating section 23 .

The halftone estimation unit 23 generates a halftone image @1 (see FIG. 9).
1 [The constant circuit is the number of apertures (] (This circuit is composed of 77 pieces. Figure 9 shows the halftone image estimation circuit for the aperture G. The halftone image estimation circuit for the remaining apertures The image estimation circuit is as shown in FIGS. 10 to 15.
The figure shows the aperture ', Fig. 11 shows the aperture E, Fig. 12 shows the aperture, Fig. 13 shows the aperture D C(r),
L, i fin D B i7), FIG. 15 shows the halftone image □□□It constant circuit for each aperture △ as a child circuit (). Here, FIG. 9 will be explained in detail. still,
The numbers in the figure indicate the number of bits of the signal line.

The shift register 30 consisting of latches LA1 to L88 selects the 8 pins 1 to 21 shifts selected by the turn 1'l'122 to the right side of the diagram in response to the timing signal from the timing generation circuit 25. is shifted to the left. Here, the shift register 30 consisting of latches LA+ to l-As is common to the halftone image estimation circuits shown in FIGS. 10 to 15. Note that the circle (not shown) on the data line in the figure represents image data of one round (21st shift data). Since 141 cases of interval DG have a size of 8 lines x 8 games J, t is shifted.

In this case, the number of white pixels in the shift register 30 can be reduced to 1, but using the h method takes a long time and requires only one circuit. It becomes 1tl. Next, this example is 2
Using the property that the value data is shifted to the left from the A side of the figure and that only the data in the first column at the end (in this case, the contents of latch L8) is replaced, the count of the number of white pixels is simplified. It became.

I will explain in detail. If you shift the i data in one column,
Latch-1-A1 newly latches C21 direct de-no2. The number of white pixels for this one column is counted as 51 by the counter 31. Also, due to this shift operation, the number of white pixels for one column is increased from the shift register 30. It is latched at Δ9. The latched number of white threads for one row is counted by a counter Jl2. On the other hand, since the number of white pixels in the IIG is held in the latch 33 during the shift 1) to 0, the RI reducer 34 subtracts the number of white pixels for one column that protrudes from this number of white pixels, and decreases the number of white pixels. The number of white pixels in the aperture G after the shift can be determined by adding the number of pixels corresponding to the current pixel value of /j to the newly input white image ia of IV1 using the adder 35. The determined number of white pixels, 9, is newly latched into the latch 3ζ3.

The output of the Wrap-33 is multiplied by a gain (8 (in this case, x1)) and output as a halftone image II constant value, and sent to the subsequent selection circuit 24.

Above is the halftone image estimation circuit for aperture G! Although we have explained the FIJ operation, the same applies to the other openings shown in FIGS.
is changed to increase the number of self-pixels to 91 and output an estimated halftone image value. For example, in the case of the opening shown in FIG. 10, the size of the space [] is 8 lines x 4 @J, and the inside of the shift register 30 is also set to 8 x 4. The same goes for the pond circuit. It should be noted that the multiplier provided at the final stage of these circuits can be easily constructed by shifting to the left (by the number of the magnification using a no register).

Next, the operation of the pattern comparison circuit for the original binary image and the 100 binary image will be explained. As before, FIG. 9 will be explained. Assuming that a binarization threshold pattern as shown in FIG. 5(C) is prepared, l1ll
The density pattern corresponding to the self-pixel count value within 1 is as shown in FIG. Figure 16 (a) shows when the number of self-pixels is 63, and (b) shows when the number of white pixels is 62.
(c) is when the number of self-pixels is 61, (d) is when the number of self-pixels is 3, (e) is when the number of white pixels is 2, and (k) is when the number of white pixels is 1. The respective density patterns are shown.

The figure shows 6ff! Although only similar patterns are shown, in reality, 64 (+D) patterns are prepared and stored in the m pattern ROM 37.
In this embodiment, the OM37 needs to simultaneously output a pattern of 64 dots (shown in parentheses on the signal line in the figure), so one dot per line as shown in Figure 16 (a). In the 8th edition of the ROM, OM is made up of (14). Ml in Fig. 16 (a) indicates one ROM. And, the 'a degree pattern R0M37 has a large number of white pixels. 11, 2=J The position information obtained by moving the interval t-1 is received as the lower 17 address, and a pattern (corresponding to the above-mentioned re-binary image) is output. ) When the number of self-pixels is 1, the contents of ROMM3 of the third screen IJ "1 are as follows: [10-=20→40→80→01→02→0
It changes like 4 → 08 (hexadecimal).

Figure 17 shows thirst 11. c [pattern f? , OM 36 is a diagram showing the relationship between addresses and data. The data in the ROM is 1 where white is '1°' and black is 'O', where 10 is the initial position f-ta.

In this way, the density pattern (re-2111'1 image) output from the U, fJ degree pattern ROM 37 is output to the judgment circuit 3.
8, the output from the shift register 30 is compared to see if it is the same pattern as the right image, and if the pattern is the same, +1' level is output, and if the stomach is output from.

The pattern comparison circuit for the aperture G has been explained above, but even for other apertures, the number of dots compared is only 5.
The operation is exactly the same.

'' When the halftone estimator outputs the halftone image estimation value for each aperture and the judgment result of [1] in two series, the selection circuit 24 receives these signals and selects the optimum aperture. Halftone image @ lit is output.

By the way, halftone painting that refers to the threshold matrix like this! In the restoring circuit, if the positional correspondence of the threshold matrix for a binary image does not match, the restored result will be ambiguous and a good quality image will not be obtained. , in this R light, when the position of the dark 1 odor pattern is 1 + ri image from both image reading device 1, the threshold 1ft ``7'' for performing 21 normalization of the image is set.
The threshold address for determining the -data is taken out from the image reading device 1, and the value -no' address signal is stored in the ilQ pattern ROM shown in FIGS. 9 to 15 as position information! I tried to make myself happy.

FIG. 18 is a diagram illustrating the flow of signals between the image reading device 1 and the media processing unit 23. Components that are the same as those in FIGS. 1, 6, and 9 are designated by the same number 61. In the figure,
41 is a threshold pattern ROM in which threshold Iit data for 2hn conversion is stored, and 42 is image data and @value pattern RO that have been read by COD etc. and converted into digital data.
This is a binarization circuit that compares the W41 data output from M41 and outputs binary data (binary signal). 43
is a main scanning counter that counts the synchronization clock, and 44 is h1 that counts the 1-1-SYNC clock.
In one scan counter, the outputs of these two counters 43 and 44 are used as an address l? It is applied to the i-la pattern ROM 41.

The threshold pattern ROM 41 outputs threshold data that is 1.6 times smaller than the threshold value and enters the binarization circuit 42 . This I
The second shift is determined, for example, as shown in FIG. 5 (c).

The 2 tff converting circuit 42 sequentially converts the image data inputted in synchronization with the threshold fa'F-tff to the halftone estimator 23.
Send to η. At the same time (in parallel with this, the main scanning counter 43
The output of the 0j running counter 44 is given to the position information counter 39 within the threshold address signal. In addition to this, the location information counter 39 also includes a restoration star [.

Synchronized clock and! -1sYNc clock is provided.

When a restoration start signal is input to the circuit configured in this way, the outputs of the main scanning counter 43 and sub-scanning counter 44 (combined 1 = 6 bits), which correspond to the first data to be restored, reach the threshold. It is given as an address to the pattern ROM 41 and set in the position information counter 39. Here, the restoration start signal is the 2nd line signal f! I
J. If it is from a reading device, it is generated from the document reading start signal shown in FIG. 8(A). Thereafter, the position information counter 39 counts the synchronization clock and H-8YNC clock from the preset value, and provides position information (part of the address) to the 1lIf pattern 1≧0M37. This position information is shown in the @ value pattern R in Fig. 9.
It is the same as that included in OM37.

In this manner, in the present invention, the density pattern ROM3
Fi to give 7km! Information first 11111tn d14
Since the address given to 11fIROMi1 can be set, the binary image and l1iIK pattern ROM
It is possible to always match the positional correspondence between the output pattern (re-binary image) and the output pattern (re-binary image).

When the binary value (t;) sent to the halftone image II restoring circuit 2 is from the image resetting unit 6, the following will occur.

When the 241 signal from the image reading device 1 is stored in the image memory 1, the threshold address signal shown in FIG. 18 is stored as a header in a specific old location (for example, address O). At the time of restoration, the header portion is first read out and set in the position information counter 39 as the cutoff j1Δ. In this way, by starting from the 1st point based on the threshold 1st point at 1st signal, there are no errors in restoration, and both images of good quality are now (q). 19 will be described as an example in which copying information is transferred from the image reading device 1.The main scanning counter 43 is
The synchronization clock, which is the main scanning direction signal, is 7Junl.
- Then, the sub-scanning counter 44 counts l-1-3YNC and sends the address signal to the threshold pattern ROM 41.
give to Ialla pattern RO, M41 +i, outputs the data, and compares the magnitude of both 1τ2 data in the straightening circuit 42 and outputs the 2nd straight signal.Threshold pattern R
The contents of OM41 are shown in FIG. Here, the initial value is 1
．． The address is 1φ. FIG. 21 shows the timing when the original m>1 is taken. In the figure, (, A) is [Ne manuscript reading star 1-signal, (11) is), 1-8 Y N C
Pulse, (ha)! ,! V-VAI-ID credit ra, (d)
i, tQ light clock, Kuho) is H-V A t, -I
D signal) is the threshold ROM address 'f:11
I accept that. In response to the original reading start signal, the sub-scanning counter 44 displays ABC l+lli as the first n! ] M
As i, I get kicked in the ass. Here it is φ. Sub-scanning counter 44 (7) E N Ogt 7 KaV-
V A I-I I) di-tangential Di i:5 h T ,
l: + When the level is “o” (that is, the original reading l
ii) Run the l-1-8YNC signal from 1 to 1. On the other hand, ten scan counter 43I! , l-i-S Y N
The values of Δ, B, and C are set as initial values by the C signal. The sub-scanning counter 44 has the same method as φ. [N is l-
1-VALID allows the synchronous clock to be counted only during the valid period of the image data. As can be seen from FIG. 21, the threshold tin R.

The M address starts from the φ address (opening).

Since restoration of halftone 1 [Ube 23T: starts in response to the original reading start signal, the positional relationship between the threshold (1α matrix and the density pattern) completely matches.

Returning to the explanation of FIG. 1 again. The halftone signal output from the halftone image restoration circuit 2 is sent to the subsequent image processing circuit 3,
The image processing circuit 3 performs both processing according to the processing mode set in advance. For example, if the enlargement/reduction mode is set, enlargement/reduction processing is performed, and if the filtering mode is set and a3, filtering processing is performed. The halftone signal image-processed in the image processing circuit 3 in this manner is re-binarized in the subsequent binarization step '84 and converted into binary data. This binary data is immediately stored in the recording device 5.
The images can be reproduced one by one or stored in the image memory unit 6 as binary data. The 2 fil) image data stored in the image memory unit 6 is
If necessary, it can be read out and reproduced as a one-sided image in the recording device 5, or it can be returned to the σ1 and halftone image restoration circuit 2 and restored to a halftone image again.

As described above, according to the present invention, it is possible to perform image processing such as 2 (enlarging/reducing because a halftone image can be estimated from a white image. Image processing such as filtering can be performed at the halftone level, and a high-quality image can be obtained. Even when various types of image processing are performed, it is possible to store the data in the memory as binary data, thereby saving memory space.

In the above description, the case where the number of white pixels within the aperture is counted is taken as an example in the middle IN adjusted image estimation 1-.

However, the present invention is not limited to this, and any method may be used as long as it estimates a halftone image based on the ratio of the white area to the black area within the aperture. In the above description, halftones are obtained by scanning one pixel at a time, but the present invention is not limited to this, and it may be possible to scan two or more pixels at a time. Further, in the above description, the 1q opening of the 7 second class was taken as an example of the plurality of types of openings, but the present invention is not limited to this, and any type may be used. Further, the size of the opening need not be limited to the size shown in example B, and any size can be used.

Furthermore, image restoration is not limited to the case of b211fI images, and restoration processing from multivalued images such as 31111 conversion and 41 conversion can be performed.

(Effects of the Invention) As explained in detail above, according to the misfire Ill, a plurality of apertures are used for one pixel, and the constant value of the halftone image 111 is set to 17 in order from the largest aperture to each of the halftone images. By selecting one of the 111 constant values that satisfies a predetermined determination condition as the halftone image 111 constant value of the pixel, it is possible to realize an image processing circuit with good reproducibility.

[Brief explanation of drawings]

FIG. 1 is a block diagram of an embodiment of the method of the present invention;
FIG. 2 is a flowchart 1 showing the halftone image estimation method used in the present invention, FIG. 3 is a diagram showing 2-8 images and the types of apertures, FIG. 4 is a diagram showing the selection order of apertures, and FIG. FIG. 5 is an explanatory diagram of ``1 selection,'' FIG. 6 is a diagram showing a specific configuration example of a halftone image restoration circuit, and FIGS. 7 and 8 are timing charts showing the operation of the apparatus of the present invention. 9 to 15 are diagrams showing specific configuration examples of the halftone image @ estimation circuit, FIG. 16 is a diagram showing the density pattern, and FIG. 17 is a diagram showing the relationship between the address and data of the density pattern ROM, FIG. 18 is a dimensional diagram showing the signal flow between the image reading device and the halftone estimating section, FIG. 19 is a diagram showing an example of the configuration of the image reading device, FIG. 20 is a diagram showing the contents of the threshold ROM, Fig. 21 is a diagram showing the timing of document reading, and Fig. 22 is a diagram showing the conventional binarization method. 1... Image reading device 2... Halftone self-cleaning restoration circuit 3... Image Processing circuit 4...2 Briberization circuit 5...
-Recording device 6...Image memory unit 20.22...Select circuit 21...Line measurement unit 23...Halftone estimation unit 2
4... Selection I scale circuit 25... Timing generator 1 circuit 30... Shift register 31. 32... Counter 33... Latch 34... Numbness device 35...
- Adding sieve 36... Multiplier 37... m-degree pattern ROM 38... Judgment circuit 39... Position information counter 41... Threshold pattern ROM 42... Binarization circuit 43... Main scanning counter 44...Sub-scanning counter L+~L9...Line memory LA+~I-A s...Latch patent applicant Roku Konishi Photo Industry Co., Ltd. Agent
Patent attorney #4' Fuji Shima 1 person Fig. 2 paralyzed 7 Fig. (d) Image information = (ko (kom-Xtsu QC Fig. 8 (a)

Claims (3)

[Claims] (1) Within a multivalued image consisting of regions of different densities created by performing multivalue conversion using a predetermined threshold matrix, multiple types of apertures are moved pixel by pixel, and the ratio of the regions of each density within the apertures. In an image processing device having means for creating a halftone image by comparing the density pattern created using the threshold value matrix with the multivalued image within the aperture, the multivalued image is obtained from the values obtained based on the values. An image processing apparatus comprising: a matching unit that matches the position of the threshold matrix at the time of determining the density pattern with the position of the threshold matrix at the time of obtaining the density pattern. (2) When the matching means creates a halftone image from the multivalued image, it starts creation based on a threshold address signal for performing the multivalued image. The image processing device described in Section 1. (3) When creating a halftone image from a multivalued image, the matching means sets a threshold address signal for performing multivalued image conversion to an initial value. The image processing device described.