JPH03212065A - Picture processor - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

Detailed Description of the Invention [Industrial Application Field] The present invention is a digital copier, facsimile, etc. that optically reads document image information using a one-dimensional or two-dimensional image sensor such as a CCD, and converts it into an electrical signal. The present invention relates to an image processing apparatus that converts original image information into a document image and finally reproduces the original image information on a recording device such as a printer.

[Conventional technology]

Conventionally, in copying machines or fax machines that perform digital image processing, methods such as dithering and smoothing processing have been used to remove or reduce moiré fringes that occur when inputting halftone originals (for example, Publication number 64-78384)
There are methods of vibrating the COD itself, or of optically focusing it with a lens or the like.

[Problem to be solved by the invention]

However, the above method has the disadvantage that the input signal is distorted or distorted, resulting in a signal that is quite different from the original signal, resulting in deterioration of image quality.

Also, since moiré only occurs on halftone originals,
There is no need to take measures to remove moiré from text and solid photographic areas, and resolution is particularly important in text areas, so image deterioration due to blurring and the like must be avoided as much as possible.

An object of the present invention is to provide an image processing device that removes moiré only from halftone dot areas without degrading the resolution of one image 'IW'. ] The image processing device of the present invention includes a document reading unit (101) that scans a document in the main scanning direction and the sub-scanning direction and outputs an image signal for each read pixel, and an image signal output by the document reading unit (101). An image processing apparatus comprising: an image processing means (102) that performs received image processing; and an image recording means (103) that records on a recording medium an image represented by an image signal output by the one image processing means (+02); Discrimination means (104) for discriminating the halftone dot area of the document from the image signal output by the reading means (101);
detection means (104) for detecting the position of the true extreme value of the periodic waveform of the halftone dot area determined by 04); A sampling point adjustment means (104) is provided that recognizes the deviation ff1(r) and shifts the sampling position by the deviation.

In a preferred embodiment of the present invention, the sampling point adjustment means (1, 04) determines the halftone dot area from the document using a black dot detection pattern and a white dot detection pattern, and determines the maximum value of the periodic waveform of the determined halftone dot area. The points and minimum values are obtained by comparing the sizes of adjacent pixels from the maximum and minimum values read using a normal sampling pinch, and the maximum and minimum values at this sampling position are determined as the maximum and minimum values in the halftone area. Shift the local maximum point to match the local minimum point,
Pixels around the minimum value point are also shifted by the same amount (r).

Note that the symbols in parentheses above indicate corresponding elements in the embodiments shown in the drawings and described later.

[Effect]

In other words, the halftone dot area is automatically detected and the sampling points match the halftone dot distribution in the halftone dot area, so that the image of the halftone dot area does not produce moiré. therefore,
Moire-free image recording is performed for both the halftone dot area and the images outside the halftone dot area.

In a preferred embodiment of the present invention, the sampling position is shifted to match the maximum value point and minimum value point of the halftone dot area, and the pixels around the maximum value point and the minimum value point are also shifted by the same amount (r). Therefore, the sampling pinch is uniform even within the halftone dot area, and no distortion occurs in the halftone dot area image.

Other objects and features of the present invention will become apparent from the following description of embodiments with reference to the drawings.

〔Example〕

FIG. 1 shows an overview of the configuration of a document reading section 101 of a digital copying machine incorporating an embodiment of the present invention, FIG. 2 shows an overview of the configuration of a laser printer section 103 of the digital copier, and FIG. Figure 2 shows an overview of the configuration of the electrical component of a digital copying machine.

Please refer to FIG. The contact glass l on which the original to be read is placed is illuminated by light sources 2a and 2b, and the reflected light from the image surface of the original to be read is reflected by mirrors 3, 4, 5, 6, .
7 and a lens 8 to form an image on the light receiving surface of a CCD image sensor 9.

In addition, the light sources 2a, 2b and the mirror 3 touch the lower surface of the contact lens 1 in parallel with the contact lens 1 in the sub-scanning direction (
The mirrors 4 and 5 are mounted on a traveling body 10 that moves in a direction perpendicular to the CCD image sensor 9.
It is mounted on a traveling body 11 that moves in the sub-scanning direction at a speed of 1/2 in conjunction with zero, and as the optical system moves, the image surface of the read document is scanned in the sub-scanning direction. Main scanning direction (
The direction in which the CCD image sensors 9 are arranged) is scanned by the CCD image sensor 9.
This is performed by solid-state scanning of the image sensor 9. In this way, the document image is read one-dimensionally by the CCD image sensor 9, and the entire surface of the document is scanned by moving the optical system.

In addition, in this embodiment, the reading density is 40 in both main and sub-scanning.
It is set to 01-Soto/1nch and can read up to A3 size originals.

Referring to FIG. 3, the reflected light from the original is converted into an electric signal by the CCD image sensor 9, and is subjected to necessary processing by the original reading section 101 and the image processing section 102.
t, r of the laser printer unit 103 (to the laser diode 1) are input to the drive circuit 43.

The l-4D drive circuit 43 energizes I, []44.

A modulated laser beam is emitted from the LD/14.

Next, reference is made to FIG. The document reading section +01 and the laser printer section 10; 3 may have an integral structure, or may have separate lateral movement and are only electrically connected.

The laser printer section 103 includes a laser writing system.

It is equipped with an image II feed system, paper feed system, etc. The laser writing system is a laser writing system 12. Equipped with an imaging lens 13 and a mirror 14, inside the laser output unit 12 is a polygonal mirror (rotated at a constant speed at high speed by a laser diode LD44 as a laser light source and an electric motor).
Polygon mirror).

The laser beam emitted from the LD/+ 4 is reflected by the polygon mirror, and then passes through the imaging lens 13. The light passes through the mirror 14 and is irradiated onto the photosensitive drum 15 provided in the image reproduction system. Around the photosensitive drum 15, a charger 16. Eraser 1F, development unit 1B.

Transfer charger 191 Separation charger 20. Separation claw 21
．． It is equipped with a cleaning unit 22 and the like. In addition,
A beam sensor (not shown) that generates main scanning synchronization signals (MS, YNC) is arranged near one end of the photosensitive drum 15 at a position where the laser beam is irradiated.

The surface of the photoreceptor tram 15 is uniformly and highly charged by the charger 1G. When that surface is irradiated with laser light, the potential of the irradiated area decreases.

Since the laser light is controlled on/off depending on whether the recorded pixel is black or white, a potential distribution corresponding to the recorded image, that is, an electrostatic latent image, is formed on the surface of the photoreceptor by irradiation with the laser light. When the portion on which the electrostatic latent image is formed passes through the developing unit 18, toner is attached depending on the level of the potential, and a toner image that visualizes the electrostatic latent image is formed. A recording sheet is fed into the area where the toner image is formed at a predetermined timing and overlaps with the toner image. This toner image is transferred to a recording sheet by a transfer charger 19, and a separation charger 20
The six recording sheets separated from the photoreceptor drum 15 are conveyed by a conveyor belt 23, thermally fixed by a fixing roller 24 having a built-in heater, and then discharged to a paper discharge tray 25.

The paper feed system has two R units, one of which is equipped with a paper feed cassette 26, the recording sheets in the paper feed cassette 26 are fed by a paper feed roller 28, and the other is equipped with a paper feed cassette 26. The paper feeding system is equipped with a paper feeding cassette 27, and recording sheets in the paper feeding cassette 27 are fed by paper feeding rollers 29. The fed recording sheet stops once in contact with the registration rollers 30, and is sent to the photosensitive drum 15 at a timing synchronized with the progress of the recording process.

Although not shown, each paper feeding system is equipped with a size sensor that detects the sheet size of the cassette.

Referring again to FIG. The electrical components of the digital copying machine mainly include a document reading section 101 that reads recording sheets and outputs image data signals. An image processing unit 1029 that processes the image data signal; a laser print unit 103 that performs recording based on the image data signal; A synchronization control section 105 that generates a synchronization signal and matches the timing of signal transmission and reception between each section and between each element within the section. It is composed of an operation display section 106 for inputting and displaying various processing modes, and a system control section 104 for controlling each section.

In the document reading section 101, the image signal read by the CCD image sensor 9 at a sampling density of 400 dots/1 nch is first amplified to a predetermined voltage amplitude by the amplifiers 31 and 33, and then amplified by the A/D converter 34. The data is converted into digital data with one pixel number of gradations (64 gradations in this embodiment) (the number of gradations is 2 to the n power and treated as an n-bit binary signal). In this embodiment, the A/D converted signal becomes a 6-bit digital signal and is input to the shading correction circuit 35. The shading correction circuit 35 corrects illuminance unevenness of the light source, CCD image sensor 9
This circuit performs corrections for uneven sensitivity of the internal light-receiving element, dark', g, current, etc.

The image data No. 1] output from the shape ink correction circuit 35 is generally processed by the MT in the 1 image processing section 102.
Spatial filter processing such as F correction and flattening.

No change? N processing, writing processing, etc. are performed, but in the present invention, after the shading correction, halftone dot detection 36, which will be described later, is performed, and sampling point adjustment 37 is performed only for the detected halftone dots. Further, signals for text areas other than halftone areas and solid photographic areas are input to the MTF correction circuit 38, where they are filtered to remove noise and enhance the sharpness of the image. The scaling processing circuit 39 enlarges/reduces the manually inputted image data signal in the main scanning direction, and enlarges/reduces it in the sub-scanning direction by adjusting the rotational speed of a motor (not shown) that drives the traveling body 10. This is done through control.

Thereafter, γ correction 40 is performed to change the gradation characteristics, and after indentation modulation processing 411 and editing 42, the data is manually input to the LD drive circuit 43 of the laser printer section 103.

! -D drive circuit 43 changes the output of the laser beam due to temperature etc. in response to the image data signal → [-04
4 and modulated laser light r-1) 1 = 1
(in) ■.

In addition, system system unit IO,1 is [10 port 107
゜R, M・r1108 r ROM i09 and CP
It is a microcomputer system equipped with U1i0, etc., and performs (the whole) installation.

FIG. 1 shows the processing operation of the system control unit 104 shown in FIG.

The power is turned on.
The system control unit 10.1 sets the processing mode etc. to the initial state (M2).

Next, 51'! It waits for the copy number to be inputted to the display section 106, and when the copy number is input, so, h, is read into the register (~13). Then perform “halftone dot detection” 41
). The threshold point extraction method will be explained. Various dot richness can be considered, but in this example, it is 100 dots/1 "1 ci".
Moiré occurs when reading 'I' 1001/1nch
~2001/1nch is targeted for detection. As the black dots become rounder and the density increases, the halftone dots become larger, the peaks gradually become smaller, and finally it becomes completely black. Furthermore, the arrangement of black circles is a regular, repeating pattern. Therefore, the image data is binarized in a two-dimensional space using a certain threshold value, and the binarized data is shown in FIG. 7a.

Check whether it matches the patterns shown in Figures 7h, 7c, 8a, and 8b, and if there is one or more matches in the 8x8 matrix shown in Figure 9, the 8x
8 matrices as halftone dot candidates. Since halftone dots do not exist locally but repeat the same pattern over a wide range, when several candidate 71-linoxes are in a row, they are determined to be a halftone dot area. , it should be noted that Fig. 7a, Section-t
b 'I'l and the seventh (.) are patterns for detecting black dots.

Figures 8a and 8b show the white throat detection button, and which pattern it matches varies depending on the density. Also, the density of 1w4 points is 1gl'4 of this pattern.
', it takes a lot of talent.

The halftone dot area where the halftone dot has been detected is subjected to the [sampling one point adjustment] described later (M6), and the areas other than the halftone dot area are
After "Si TF genuine" (~17) was performed, "change 1'Δ
"Processing" (~18) is carried out, where "Rescaling processing J
(M8), the 5 scaling processing circuit is shown in FIG. The speed converter is composed of one speed converter and 29 circuits υ control unit 110, and performs scaling operation by combining an external line memory. FIG. 6 shows a block diagram of the interpolation calculation unit 11i. The image data is converted to the same size as the image data, and the image data is scaled by 1%, and the data after the scale change is output using the ChronoCrate as the input.The circuit is stored in the internal RAM 108. il + your data
The image data at the same magnification, which was set by two people who entered the circuit, was created by three people working as software for this circuit control button.
This is image data of 6 bits (64%?2) sampled at a certain sampling pitch P. Now, as shown in Figure 12, let us consider a one-dimensional locust moxibustion with P].
The sunblink pitch (virtual) Pr at a small time is Pr>
1, and the sunblink pitch Pe when there are five players is Pe(
It becomes l.

If the unchanged rate of 11 is 9%, etc. 47'! , α pieces of data are extracted from 100 pieces of data (distance a + 00 ), and by repeating this process, the magnification is changed. Here, the virtual sampling pitch P at the time of reduction and expansion
Since r and Pe are 100/α, the coordinates of the α sampling positions are: X+t=lOO/αX n (n = O, I, 2−,
-1) ...(1). Circuit control data means this coordinate value. In addition, the circuit control data is internal 5]2
It is routed to lj A\'110111 of X4bit.

The flow of internal processing is different between reduction and enlargement 11+i; during reduction, data is processed manually → interpolation calculation → speed conversion → data output, while during enlargement, it is processed as follows: take → speed conversion → interpolation calculation. →It will be processed as data output. this is,
This is due to the difference in speed conversion method, which will be described later.

The circuit control data is designed to make it easier to convey Xn in equation (1) to the circuit; it is a sequence of 100 to 400 numerical values, but the meaning of the numerical values can be understood by explaining the actual calculation method of these numerical values. The basic principles of the speed conversion section 112 and the interpolation calculation section 111 will be explained as follows.

First, the principle of the speed converter 112 will be explained.

Speed conversion is necessary to make the scaled image data the same chrono rate as the input. Letting the frequency of input data be f, the virtual frequency f′ at the scaling factor α% is f”: α/100X f... (2
). When outputting this f' data, it is necessary to convert it to f, so when reducing, the speed is increased,
When enlarging, slow down the speed. Actual speed conversion is performed by line memory input and output.

Speed conversion during reduction will be explained.

When the sampling points 0 to 100 at the same magnification in Fig. 12 are the coordinates X, the sampling point at the time of reduction is Pr) 1, so X
Either there is one between and x+1, or there is none at all. And when human power is applied to the line memory,
It determines between the data obtained by interpolating the data of the virtual sampling point between When outputting from memory,
Speed conversion is achieved by sequentially reading all data.

Speed conversion during enlargement will be explained.

The sunblink points during expansion are X and x+ since Pe<1]
There is one or more between. When inputting all the equalized sampling data into the line memory manually, when outputting it from the line memory and sending it to the interpolation calculation unit +11, the read operation is stopped for several sampling points between X and x+1. Speed conversion is realized by adjusting the knob and holding peripheral data as is.

Next, the principle of the interpolation calculation section 111 will be explained.

In Figure 12, if the virtual sampling point when changing the magnification matches any of the sampling positions when the magnification is the same, you can take that data as is, but if the virtual sampling point is between the sampling points when the magnification is the same. , the data of the virtual point must be predicted from the surrounding equal fΔ data. Although there are various conventional prediction methods, the present invention uses a cubic function convolution method (approximating a sine function by a cubic function), which is known as a relatively sophisticated interpolation method.

In this method, as shown in FIG. 13, when there is a virtual sampling point between the 18-time data S'' and 81+1, the distance r between the virtual point and the 18-time data is calculated.

From the distance A'I r, the interpolation coefficient h as shown in FIG.
(r) is obtained from the following formula % formula % and the data of the virtual point σ is calculated using the following formula h(1+r)+h(r)+h(1-r)+h(2-r)
... Decide from (3).

Here, in this embodiment, the accuracy required for r is 0.
．． 25 increments. That is, r==o, rO,25,
r"=0.5. r=0.75, whichever is the closest point. As a result, the distance data "can be treated as 2-bit binary data.

Table 1 shows the values of h(r) for these four types.

Virtual sampling data at the time of scaling can be obtained.

Furthermore, since the distance fir data has an accuracy of 0.25 increments, for convenience of calculation, equation (1) is corrected by σ as shown below to perform scaling processing.

Xn=100/αXn+1/8 (n=o, ], ]2
-, α-1... (4) In this formula, if the integer part is Pn and the decimal part is Qn, then Xn=
Pn+Qn ・ ・ ・ (
5), Pn represents the coordinate value of Sl, and Qn represents ".

Referring again to FIG. 4. Here, r halftone detection" (xq
Fine adjustment of sampling points is performed on the halftone dot area detected in step 4) [Sampling point adjustment J (M6) will be explained.

'As explained in ``Scaling process J (M8)'' mentioned above, in this embodiment, the same sampling pitch [
) is set to 1, the sampling point is changed in steps of 0.25. The smaller the increments are, the J1 heavy viscosity of the variable j1 will also increase, but here we will explain in increments of 0.25.

Normally, processing for shifting the sampling position is performed only in the unchanged Z1 processing. In other words, in the 5-character section and the solid photo section,
Variation of A'7 Processing 9 Same-magnification processing is performed. In the present invention, when a halftone dot area is detected, the sampling point is shifted to that part regardless of the scaling process.

That is, the sampling point may be shifted even when the magnification is the same, and when changing the magnification, a position different from the obtained virtual sampling point may be selected.

However, it is assumed that the sampling position does not deviate by more than one position from the conventional sampling position.

The sampling position is set so that the peak point, that is, the maximum value point and the minimum value point of the halftone dot document coincide with the sampling position as much as possible over the entire halftone dot area.

First, the local maximum and minimum values are extracted from data read at a normal sampling pitch. In the matrix shown in Figure 1O, the pixel of interest Xo is XI+X2 T xa J x
If it is louder than all pixels of 4, it is the local maximum value, and if it is smaller, it is the local minimum value, and X I + X2! X, (+1 out of x4
If only one value is equal but larger than all the other three, it is considered a local maximum value, and when it is smaller than all three, it is also considered a local minimum value.

After detecting the local maximum value and local minimum value, processing is performed to shift the sampling point in order to find the true extreme value point from pixels adjacent to the local maximum value. This is shown in Figures 11a, 11b, 11c and 11d.
When T1 reaches an extreme value at pixel 2 +, T 3, adjacent pixels To , T2 . Compare the size of T3. Chapter 11a
When it becomes as shown in the figure, it is determined that T1 and the true extreme value match. In the case of Figure 11b, the true extremum is T. It is determined that there is an extreme value at a position between and T1 and shifted by 0.25 from 1゛1 to the 'ro side, and the sampling position of pixels around this value is set to r=0.75. In the case of FIG. 11c, it is determined that the extreme values are T and '1゛2, and that there is an extreme value between them, so that r=0.5. Similarly, in the case of Figure 11d, shift r to 0.25. These criteria are whether 1゛1 and ``r2 are both extreme values, and in that case, whether ``ro'' and ``1゛3 are close values. If only +TI is an extreme value, T
r is determined by checking the level difference between o and 1'2.

Next, the range of pixels in which the sampling position is rotated from one extreme value is determined. There are several ways to do this, as shown in 1ζ below.

A method of determining a fixed value of i and making it a fixed 71-linox range centered on the extreme value for both main and sub-scanning.

2j The density of halftone dots is calculated from the number of extreme values in a certain area by a +
q How to divide Ut (same part J) according to L density.

J) A method in which the range is defined as half the number of pixels between an extreme value and the next extreme value.

In this example, using θλ in the case of ■, the halftone density is selected stepwise from the degree of dispersion of the extreme value. I and thereby determine the 71~Linox range. 1001/jnch.

1331/1nch, 1501/1nch, 175
Select the detection range from 5 levels: 1/1nch, 2001/1nch, 7X7, 5X5, 4X4, 3X3 respectively.
, 2×2 matrix, the rear side of the raster scan is given priority, and the sampling point of the image J that is not included in the matrix is not changed.

How to change the sampling point? This is "variable magnification processing" (M8)
The interpolation calculation explained in is also used. In other words, the coefficient bar of the multiplication):; of the interpolation calculation circuit Ill shown in FIG.
Change h+ + h2+ hj. This coefficient is an approximation h (r) of the 5jnc function of the subtraction theorem, and in the example, r is defined in 0.25 increments, δ is written into the RAM 108 as variable magnification circuit control data, and this is sequentially read out. This determines r, but the data is changed while being read from this RAM and supplied to the circuit.

Note that the interpolation calculation unit I in the scaling process J (M8)
ll is ``sampling point adjustment J (M6) interpolation operation t
If a configuration is adopted in which each image is made independent without being combined with 'l'', moiré will occur due to scaling if the sampling position is shifted again due to scaling.

This method is not practical except for scaling, which selects the actual sampler data closest to the virtual sampling point on which interpolation is performed. There is also a method of performing "sampling point adjustment" after "variable G'j processing", but this method has the disadvantage that the range of halftone dot density is expanded by scaling, making control complicated.

111 degrees, referring to Figure 4, after "sampling point adjustment J (M6), rMTI'-" correction J (M7), "magnification processing J (M8) is performed, and then system I and control unit +04 are set. The copy operation is controlled under the specified conditions (M9), and when this is completed, fΔ) C input reading (~13
), wait for the user to set the magnification, and repeat this process.

〔Effect of the invention〕

As described above, according to the present invention, the halftone dot area is automatically detected and the sampling points match the halftone dot distribution in the halftone dot area, so that the image of the halftone dot area does not produce moiré.

Therefore, images are recorded without moiré in both the halftone dot area and the images outside the halftone dot area.

[Brief explanation of drawings]

FIG. 1 is a sectional view showing an outline of the configuration of a document reading section 101 of a digital copying machine incorporating an embodiment of the present invention. FIG. 2 is a sectional view showing an outline of the configuration of a laser printer section 103 of a digital copying machine incorporating an embodiment of the present invention. FIG. 3 is a block diagram showing an outline of the configuration of the electrical components of the copying machine shown in FIGS. 1 and 2. FIG. FIG. 4 is a flowchart showing the processing operation of the system control unit 1011 shown in FIG. FIG. 5 is a block diagram showing an outline of the configuration of the scaling processing circuit of the image processing section 102 shown in FIG. 3. As shown in FIG. FIG. 6 shows the interpolation calculation section 11 of the scaling processing circuit shown in FIG.
1 is a block diagram showing the configuration of FIG. Figure 7a, Figure 7b, Figure 7c, Figure 8a, Figure 8b,
FIG. 10, FIG. 11a, FIG. 11b, FIG. 11c, and FIG. 11d are plan views two-dimensionally showing the distribution of one image data. FIG. 9 is a plan view showing an 8×8 matrix for setting the distribution of image data. FIG. 12 and FIG. 13 are plan views showing the relationship between the sampling data position at the time of equal magnification and the sampling data position at the time of variable magnification. FIG. 14 is a graph showing the values of the interpolation function used in the cubic function convolution method. 1: Contact glass 2a, 2b: Light source 3.4.5
．． 6, 7, 14: Mirror 8, 13: Lens 9: CC
D image sensor 10, 11: Traveling body 12: Laser output crab unit 15: Photosensitive drum 16: Charger 17: Eraser 18: Developing unit 19: Transfer charger 2
0: Separation charger 21: Separation claw 22
:Cleaning unit 23:Transport belt 2
4: Fixing rollers 26, 27: Paper feed cassette 30, registration roller 25: Paper ejection tray 101: Original reading unit (original reading means) 102: Image processing section (image processing means) 103: Laser printer section (image recording means ) 104 System control section (discrimination means. 105: Synchronization control section 107: I10 port 109: ROM l1l: 1lilil11 calculation section detection means, sampling, lily adjustment means) 106: Operation display section 108: RAM +10: CPU 112 :Speed conversion department Riko Co., Ltd.

Claims (2)

[Claims] (1) a document reading device that scans the document in the main scanning direction and the sub-scanning direction and outputs an image signal for each read pixel; an image processing device that receives the image signal output from the document reading device and performs image processing; An image processing apparatus comprising an image recording means for recording an image represented by an image signal outputted by the image processing means on a recording medium, and a determining means for determining a halftone dot area of a document from the image signal outputted by the document reading means. , a detection means for detecting the position of the true extreme value of the periodic waveform of the halftone dot area discriminated by the discrimination means; and a detection means for detecting the position of the true extreme value of the periodic waveform of the halftone dot area discriminated by the discrimination means, and a deviation amount between the position of the true extreme value and the sampling position read by the document reading means. An image processing device comprising: a sampling point adjustment unit that recognizes the deviation and shifts the sampling position by the amount of the deviation. (2) The sampling point adjustment means shifts sampling positions of a plurality of pixels around the position of the true extreme value by the amount of deviation between the position of the true extreme value and the sampling position read by the original reading means. characterized by
An image processing device according to claim (1).