JPS59161982A - Image processing device - Google Patents

Abstract

PURPOSE:To attain the multi-gradation processing at a high speed and also to obtain reproduced color pictures of high quality by binary coding the color picture data through the dither processing and converting the binary coded data into the multi-value coded data in time division within the transfer time of a picture element. CONSTITUTION:The eight-bit digital signal underwent the photoelectric conversion by a photodetecting unit and the conversion into the complementary color is applied to a multi-gradation processing means after the shading compensation, the gamma compensation and the UCR compensation. The multi-gradation processing is performed by the dither processing and the multi-value coding processing. The six-bit picture data is simultaneously compared with plural threshold value matrices by a dither processing means first and binary coded. Then the video signal underwent the dither processing in response to the picture transfer clock is stored in line memories 180-9-180-14 respectively by a multi-value coding circuit 180 and then delivered in the form of a multi-value coded signal via a data selector 180-15.

Description

[Detailed description of the invention]

TECHNICAL FIELD The present invention relates to an image processing apparatus that performs image processing using digital processing. Conventional technology Conventionally, an original image is separated into three colors using a color separation filter.
The original image is scanned for each color separation, a latent image is formed on a photoreceptor by the color-separated light image, and the latent image is developed using a developer of a complementary color.
There are color copying machines that reproduce color images by superimposing multiple colors. This type of color copying machine uses the analog characteristics of electrophotography for color balance, halftone expression, etc. necessary for color image reproduction, making adjustments to image exposure, photoreceptor charging conditions, etc. complicated. Because the corona charge, photoreceptor, etc., are directly affected by temperature and humidity, not by force/reduction, the image quality fluctuates widely due to environmental changes. Furthermore, since everything from reading the original image to forming the latent image is performed by a two-dimensional optical system, it is not possible to process each point of the image. Purpose The present invention was made in view of the above points, and an object of the present invention is to provide an image processing device capable of reproducing high-quality color images. A further object of the present invention is to provide an image processing apparatus capable of high-speed multi-gradation processing. Furthermore, it is an object of the present invention to provide an image processing device capable of multi-gradation with a simple configuration. =/-'-" = One Embodiment Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 is a sectional view of a copying apparatus to which the present invention is applied. The document is placed on the transparent plate 2, and the document is held down by the document cover 3 from above. The reflected light is reflected by the movable reflecting mirrors 9 and 10, and is reflected by the lens 11-
1, then enters infrared cut≠12, where three wavelengths of light, blue (B), green (G), and red (R)
It is spectrally separated into A blue filter 13 is applied to each of the separated B, G, and R lights. Green filter 15. The red filter 17 adjusts the intensity and corrects the spectral characteristics of the 93 colored lights, and solid-state imaging devices (CCD) 210, 220, 2
Receives light at 30. The reflected image from the original 3 is illuminated by a logen lamp 5 for illuminating the original.
With the movable reflective mirror 9 that moves together with the movable reflective mirror 9 and the moving speed j that is 1/2 of the movable reflective mirror 9, the optical path length is kept constant by the movable reflective mirror 10 in the same direction, and the lens 11- 1. Infrared cut filter 11
-2 and dichro≠≠ヰ≠412, the image is formed on solid-state image pickup devices (CCU) 210 and 220° 230 for each color as described above. Each CCD light receiving unit for solid-state imaging. The digital signal is converted into a digital signal at the image processing unit 1.
00 performs necessary image processing, and the laser modulation unit 30
0, a laser beam modulated by an image signal is emitted to the polygon mirror 22 and irradiates the photosensitive drum 24. The polygon mirror 22 is rotated at a constant speed by a scanner motor 23, and the aforementioned laser beam is scanned perpendicularly to the direction of rotational force of the photosensitive drum 24. - A photo sensor 64 is installed before the L position to start scanning the laser beam on the drum, and a laser horizontal synchronization signal BD is generated by the passage of the laser beam. After the photosensitive drum 24 is uniformly neutralized by the neutralizing electrode 63 and the neutralizing lamp 71, it is uniformly negatively charged by the negative charger 25 connected to the high voltage generator 77. When the laser beam modulated by the image signal is uniformly irradiated onto the negatively charged photosensitive drum 24, a photoconductive phenomenon occurs, and the photosensitive drum 24
The charge flows to the main body ground and disappears. Here, the laser is not turned on in areas where the document density is low, and the laser is turned on in areas where the document density is high. In this way, an electrostatic latent image is formed on the photosensitive drum 24 in a light area on the photosensitive drum surface corresponding to a high-density portion of the original. This electrostatic latent image is transferred to the yellow (Y) developer 36. Magenta (Incorporated) developer 37. Cyan (C) developer 38. A black (Bk) developer 39 develops the photosensitive drum 2.
4. Form a toner image on the surface. At this time, developing sleeves 85, 86, 87 in the developing devices of each color. In order to set the potentials of the photosensitive drums 88 to -300v to -400v, the toner in the developing device to which voltage is applied by the developing bias generator 84 is stirred and negatively charged, and the surface potential of the photosensitive drum 24 increases. The toner adheres to areas where the potential is higher than the development bias potential, and a toner image corresponding to the original is formed. Thereafter, unnecessary electric charge on the photosensitive drum 24 is removed by a negatively charged post electrode 41 by a capacitor 40 and a high voltage generator 77 for eliminating the potential on the surface of the photosensitive drum 24. make the surface potential uniform. On the other hand, the operation board 72 feeds the rolled paper stored in the selected cassette roller 43 or 44 by the paper feeding operation of the paper feeding roller 46 or 47, and skews it by the second registration roller 52 or 50. After correction, the conveyance roller 51. The second registration roller 52 transports the transfer paper at a predetermined timing of θ, the gripper 57 of the transfer drum 53 holds the leading edge of the transfer paper, and the transfer paper is electrostatically wound around the transfer drum 53 . The toner image formed on the photosensitive drum 24 is transferred onto the close-up paper 48 by the transfer-transfer electrode 54 at a position where it contacts the transfer drum 53. The toner image is transferred to the transfer paper a predetermined number of times depending on the selected copy color mode. After all the toner images are transferred, the transfer paper is Perform static electricity removal. After the transfer has been completed a predetermined number of times, the transfer paper is separated from the transfer drum 53 by a separating claw 90 and transferred to a conveyor belt 59 by an unloader 58 (for conveyance).
It is sucked upward and guided to the fixing section 6o. - The residual charge remaining on the photosensitive drum 24 is removed by the cleaning blade 89 in the seat 62. Furthermore, the charge on the photosensitive drum 24 is removed by the ACC static eliminator 63 and the static eliminator lamp, and the process proceeds to the next cycle. Incidentally, 19 and 20 are fans for cooling the optical system 11 and discharge the illumination system. Here, the main body operation sequence is read into four colors (Y s M + image processing unit) 100. Next, the document is scanned, and three colors (B, G, Iζ) are simultaneously scanned on the CCD 210. CCD220. CCD23
At 0, the image is captured and image processing is performed) At 100, Y, M, C, which are complementary colors of B, G, and H, and B of the black version
Calculate K's silkworm and perform processing such as color correction. The document is scanned rL 4 times, and the image processing J! is performed on the first scan. l
11 of the yellow (Y) component calculated in the unit 100. , a bow and the like by laser modulation to form a latent image on the photosensitive drum 24. This latent image is developed by a yellow 9i imager 36 and transferred onto a transfer paper wound around a transfer drum 53. Similarly, the second scan turned magenta (Ml-4 - the third scan turned cyan (C), and the fourth scan turned blue (Ml-4).
An image corresponding to BK) is transferred to a transfer paper and fixed by a fixing device 60, thereby completing image recording in the full caster mode. As shown in Figure 2-1, the spectral energy distribution of the halogen lamp used to expose the original is such that the light output is high at long wavelengths, that is, close to the red region, and the light output is low at short wavelengths, that is, close to the blue region. . In addition, the spectral sensitivity characteristics of CDD are 500~
It has high sensitivity in the edge region of 600 nm. Therefore, after the reflected light from the original is output from the dichroic mirror,
The spectral characteristics of the halogen lamp are as shown in Figure 2-2. In addition, as shown in Figure 2-3, the spectral characteristics of the dichroic mirror are not good, so by passing it through a multilayer interference filter with spectral transmittance as shown in Figure 2-4, it is possible to improve the spectral characteristics. As shown by the broken line in Figure 2-2,
A color-separated optical image without unnecessary wavelength components can be obtained. Furthermore, by stacking a plurality of filters for each color, it is also possible to change the spectral transmittance and correct the imbalance in output as shown by the broken line in FIG. 2-2. Figure 3-1 shows a block diagram of the main body control section. 422 and 421 are operation unit units used by the operator to operate the machine, 422 is a numerical key for setting the number of copies, and 72-1'6 and 72-17 are upper and lower cassettes. The cassette selection keys 72-2 to 72-p are color mode selection keys for selecting color copying modes (for example, 4F" selected by the 72-2 key). What is uLL mode?
K, do 4 exposure scans, 13, G for each scan. Corresponding to the original exposure image color-separated into R, Y
, M, and C toners, and in the fourth scan, it is preserved with BK toner to match the BK acid content of the original, and a full-color copy is obtained by combining all four color images. mode. Similarly, in 3FuLL mode, Y, M, C
In (BK+M) mode, BK and M, BK; Y, IVI correspond to two document exposure scans.
, In C mode, it corresponds to one original exposure scan,
A given copy is obtained with each monochromatic toner image. 72-23 is a 7 segment LED for displaying the number of copies. 72-18 is a 7-segment LED that displays the number of copies.
, 72-15 is a display device (not shown) that lights up when a detection device (not shown) detects that there is no replenishment toner in the hopper (not shown); An indicator 72-20 that indicates this when a paper jam is detected by the detection device is a display 72-20 that indicates this when a detection device (not shown) detects that there is no paper in the selected cassette. 1 is a heat pressure fixing device 6
0 is a weight indicator that lights up when the surface temperature of the fixing roller has not reached a predetermined value.Indicators 72-15 and 72
-14, 72-20, and 72-1 are lit, copying operations are prohibited. For 72-21 and 72-22, when the copy paper in the cassette selected by the paper size indicator is A3 size, 72-21; but when it is A4 size, it is 72-21;
2-22 lights up. Also 72-12
tfi is a copy density adjustment lever.If the lever is moved in the direction of 1, the lighting voltage of the document illumination rungs 5 and 6 will be lowered, and if it is moved in the direction of 8, the lighting voltage will be increased. Next, the sub control unit 421 will be explained according to FIG. 3-3. 421-14, -
15 and -16 are γ correction circuits 140 (described later) that correct the gradation of read data for 8-bit pixel data read by the CCD and quantized by the A/D converter.
It is a group of switches connected to a rotary digital code switch, each of which generates a digital code. noyr
Connect so as to select the data conversion memory element that achieves the desired γ characteristic! are doing. 42], -5 to -13 are along the masking processing switch, and are connected to the masking processing circuit 150, which will be described later.
Human-powered yellow image data Yi, magenta image data M
i, coefficients ai, bi, ci when performing the transformation of the following formula on cyan image data C1VC (i = 1°2.3
), these are the switch groups 421-14, -1
5. Like the -16, it is composed of a rotary digital coat switch that generates digital codes from 0 to 16. Note that data conversion for masking processing is as shown in the following equation. Also, 421-1. -2. -3. −4 is UC, R, which will be described later.
This is a loader-digital code switch that provides correction coefficients for each Y, M, C, and BK data in the processing circuit 160. Reference numerals 421-20, 21, 22, and 23 are controls respectively connected to the high-pressure generator 77, and the current flowing through the charger 25 for uniformly negatively charging the photosensitive drum is set to the open side 1. You can adjust the shading of each color in the image and change the color balance.
Reference numeral 11-65 is a sequence controller that controls the loads on all the internal components of the apparatus, and the loads shown in the timing chart of FIG. , and is driven at a fixed timing via the sequence controller → chopstick port 419 → load drive circuit 420 according to the 7-can control table in the ROM 423. In the figure, L+, L2...-LN correspond to individual loads, but the driving method of each load, such as the renoid, motor, rung, etc., and the method of sequence control according to the ROM are well-known. By the way, the explanation here will be omitted. Main control unit 422. The sub-control units 421 correspond to the operation section 1, respectively, and the loads to be driven are keys, lamps, and LEDs. These drives or inputs are performed by the key and display controller 412. Further, for example, the driving of LEDs and lamps, the scanning of keys, and the input method are performed by well-known methods, and detailed explanations thereof will be omitted. The sequence progresses according to the timing chart in Figure 3-3. This timing chart shows, as an example, the sequence of three colors Y, M, and C.
The sequence for obtaining a full color image is shown. In order to obtain full-color images of Y, M, and C with this device, it is necessary to rotate the photosensitive drum 5 times and the transfer drum 10 times, so the diameters of the photosensitive drum 24 and the transfer drum 53 are 2:1.
The ratio of Further, this sequence is executed based on the rotation of the photosensitive drum 24 and the transfer drum 53, and as shown in FIG.
A sequence clock generator consisting of a clock fm 24-7 driven by a gear 24-9 driven by a drive shaft 4 and a photointergeter 24-8 generates a clock fm as the photosensitive drum 24 rotates. The transfer drum 53A progresses according to the generated drum clock C, and with one rotation of the transfer drum 53A,
The drum clock counts 400 clocks. Therefore, the on/off control of the load is performed based on the count value from the reference point (hereinafter referred to as home position) of the transfer drum 53 (not shown). In the timing chart shown in Figure 3-4, the numbers indicating the operating timing and non-operating timing are as follows:
These are the clock count values when the transfer drum) IP is set to 0 clocks. For example, the exposure lamp 6 is ONL at 120 counts at the third rotation of the transfer drum, 120 counts at the fifth rotation, and 120 counts at the seventh rotation, respectively, 118 counts at the fourth rotation, and 118 at the sixth rotation.
At the count of 118 on the 8th rotation, it is controlled to turn off. Following this timing chart,
When detected by the device configuration 412 in FIG. 1, the sequence controller 411-65 starts a coherence sequence and drives the photosensitive drum 24, the transfer drum 53, the second registration roller 52, and the second registration roller 52. . During the first rotation of the B&light drum 24, the surface of the photosensitive drum is neutralized and standardized by the removing electric devices 61, 6.3, the removing lamp 71, etc. The original 1 is placed on the platen glass stand 2, and from the 120th clock of the third rotation of the transfer drum 53, the halogen 2 lamps 5 and 6 for exposing the original are turned on, and original exposure scanning is started. The reflected image from the original is reflected by mirror 9,
7 and is reflected by the lens 11 to the CCDs 13, 15 .
The light is condensed to form an image on the light-receiving surface of 17 and is incident on the gicroic mirror 12, and the reflected light image from the document, which has been color-separated into B, G, and H, is incident on each CCD 13, 15, and 17. Ru. The color-separated image corresponding to the document received by this COD is photoelectrically converted and then subjected to necessary real-time data processing in an image processing unit, which will be described later. As described above, in the order of M and C, the photosensitive drum is finally exposed with the laser source l modulated with the above image data, and a latent image corresponding to the original image is formed on the surface of the photosensitive drum. . In the timing chart of FIG. 3-'0, the latent image formed on the photosensitive drum 24 corresponds to the first exposure scan at the clock 254 of the third rotation of the transfer drum 53.
The Y (yellow) developer 36 starts operating at the 196th clock of the same rotation and stops at the clock 293 of the 4th rotation, and starts operating at the 196th clock of the next rotation of the transfer drum. When the transfer charger 54 stops operating, a yellow toner image that matches the yellow component of the original is transferred onto the transfer paper wound around the transfer drum 53.
In the same (・)f, there is a magenta toner image corresponding to the magenta component of the document at the 5th, 6th, and 7th rotation of the transfer drum 53,
After 7.8.9 rotations, a cyan toner image corresponding to the cyan component of the original is transferred to the transfer paper, and YlM, , C are multi-transferred to the same location. Note that the reflected light image from the original image is separated into three color components of B, G, and H by the dichroic mirror 12 and each enters the C0D13, 15, and 17 to form a yellow toner image. G when reading the image. itO information signal When reading an image to form a zenta toner image, the B and H signals are used, and when reading an image to form a cyan toner image, the 13 and G signals are used for color correction. Processing is performed sequentially in the order of Y, M, and C. On the other hand, at the 225th clock of the third rotation of the transfer drum, in which the first exposure scan is omitted, in order to feed transfer paper from the cassette 42 or 43 selected on the operation panel, in the case of the upper cassette. 47 is activated when the paper feed roller 46 is in one stage. The transfer paper fed from the cassette is conveyed by the conveyance loan 50 or 49, the skew is corrected by the second registration stroke 52, and the transfer paper is firmly held by the grits 57 of the transfer drum 53 by the second registration stroke 52. At a predetermined timing, the leading end is held firmly by grits (-57), and then wrapped around the transfer drum 53, and the multiple transfer of toner images as described above is performed. The sheet is peeled off by the drum 53, guided to the fixing device 60 by the conveyor belt 59, subjected to heat and pressure fixing, and then discharged.The operation timing of each of the above-mentioned loads is as shown in the timing chart of Fig. 3-4. 4 is a block 1/] showing a schematic configuration of the present invention with the image processing unit 100 as the middle part.In the image processing unit 100, the CCD light receiving unit 200 is
Yellow (y) r magenta M, cyan tc) required for printing based on the image signals of the three colors captured. This is a part that calculates the appropriate amount of each BK signal, and outputs it to the laser modulation unit 300 for each color. Therefore, in order to form a color image using this device, in the case of four-color printing (Y, M, C, BK), the original is scanned four times by the CCD light receiving unit 200, and in the case of three-color printing (Y, M, C), the document is scanned four times by the CCD light receiving unit 200. requires scanning the original three times. In the case of multi-color overlay printing, -41 (overlapping) documents are scanned. The image processing unit 100 is composed of the following circuit blocks. A shading correction circuit that corrects uneven illuminance in the image Y',
This is performed separately for each scan for the hl and C signals. 140 is a γ correction step fl'i'+, which corrects the gradation of each color by masking it and adding it to the LJCR cuffs. 150 is a masking processing circuit that prints the appropriate amount of Y, M, and C necessary for printing.
Calculate. Further, 160 is a UCR processing circuit which extracts an appropriate amount of K from Y, M, and C for creating a black plate. A dither processing circuit 170 binarizes a halftone image using a dither method. Reference numeral 180 denotes a multi-value processing circuit which further performs pulse width modulation on the image signal binarized by the dither processing circuit 170 to improve the gradation in intermediate tones. The image processing unit 100 is composed of these processing circuits and a synchronous control circuit 190 that synchronously controls them. COD9
The optical unit 200 passes the optical image through the dichroic filter 12,
[l' This is the part that separates the colors into three colors B, G, and R and converts them into electrical signals. The three color separated lights B, G, and R are CCDB210, CCDG220, and C, respectively.
Photoelectric conversion is performed by the CDR 230. Photoelectrically converted B
, G, and R signals are respectively CCD driver B240,
CCD driver G250 and CCD driver R260 were used to digitize 8 bits, making history B.
, G, and H into Y, M, and C signals, which are complementary colors. The digitized 8-bit Y, M, and C signals are VIDEOY, VIDEOM, and VID, respectively.
EOC, and Fl'-Fuko and Nisuru. VIDFX) 'Y
, VIDEOM, and VIDEX) are connected to the shading correction circuit 130 via No. 1 and No. 1 B mails 271, 272, and 273, respectively, and the shading correction circuit 130 performs the above-mentioned shading correction. The shading has been corrected. M,Ci? J issue VIDEOY, VIDEOM,
VIDEOC is route 48 105, 106, respectively.
The signal is supplied to the γ correction circuit 140 via 107. The γ correction circuit 140 converts the gradation into characteristics that are easy to color correct. Here, to simplify the following process, we will use VIDE
OY. VIDF, OM, VIDEX) CIti are each converted to 6-bit A symbol. γ corrected 6
Bint's VIDhXJ Y, VIDEXJ M,
V'IDEX) Ctj: Im live wire 108. It is sent to the masking processing circuit 150 via 109 and 110. VIDEX in the masking processing circuit 150)
Proper color correction for printing is performed from Y, VID addition M, VIDEOC, and the color corrected VIDEO'Y, V
IDEOM. VIDF and OC are sent to the processing circuit 160 (UC). [J
In the CR processing circuit 160, the colors are corrected 72M, C.
The under color removal amount is calculated using No. 11,j, and black BKi is obtained. Y, M, Y with BK subtracted from C4 neck color
, M, and C amounts become color-corrected Y, M, and C amounts. These four color image signals Y, M, C, and BK are supplied to the dither processing circuit 170 via each line 114 in the order of Y, M, C, and BK for each run. Here, halftone expression is performed digitally using dot density per unit area, and dither processing with three different threshold values is performed (described later). ,
It is output as a binary code at 05-3. Multivalue processing circuit 1
In 80, there are three binarized im numbers 115-1 and 115-2.
, 115-3, and is connected to a signal line 116 to supply a binary 1-value pulse width modulated signal to the laser modulation unit. A laser driver 310 and a laser unit 320 emit l/-the beam to form a latent image on the photoreceptor 24. The main body control unit 400 performs sequence control of this device, and
Controls each processing unit. Sequence controller 411- in the present control unit 400
65 (FIG. 3) instructs the image data processing unit 100 to start the original exposure scan i (/?X, yellow exposure C,
', the magenta exposure signal at the start of the document exposure scan for the second magenta toner image formation, the cyan signal at the third time, and the BK double signal at the fourth time. 403, 404, and 406 to the shading correction circuit 130 when the white history plate 4 is being irradiated. The shading correction circuit 130 sends out an exposure start signal>3 (shading start No. 48) 402, and upon receiving this, the shading correction circuit 130 starts the shading operation as described later. Read the 1ifti image data for correction corresponding to the front plate. FIG. 5-1 shows the configuration of the synchronous control circuit 190 shown in FIG. 4. Synchronous control circuit (crystal oscillator 190-1, C
It has an OD successive timing generator 190-2 and an address control unit 190-3, and drives the COD in synchronization with the beam detect signal BD321-1 for each line scanned from the laser scanner, and also outputs from the COD. Counts serial pixel data and performs scanning line address control. A clock 0LK 190-4 having a frequency four times that of image transfer locks 2φT 190-8 and 190-12 is supplied from a crystal oscillator 190-1 to a COD succession timing generator 190-2 and an address control unit 190-3. Image transfer/lock 2φT190-8 is CO
A clock that transfers serial image data output from COD via signal lines 102, 103, and 104.
Driver B'240. COD Drino<-G220.
It is supplied to COD driver 260. Further, the image transfer lock 190-12 is connected to each processing circuit in the image processing unit 100 by a signal line 101. ．． 119,120゜121
．． 118 and 117 (FIG. 4). In the address control section 19.0-3, in synchronization with the beam detect signal +31) 321-1, the horizontal synchronization signal,
IIs YNO], 90-5 and 190-11
emits. This signal causes the COD read timing generator 190-2 to control CC1)13210.001)G22.
0. Shift pulse S r-T 190-6 which is the signal to start reading the CCDB 230 All signals! 102,1
03,104 - COD driver 824'O,
COD driver ()250. OCD driver■mo2
60 to start outputting one line each. φ11
90-7゜φ2190-8, +(, S 190-1
0 is OC! I) Necessary I1 for driving (strong and COD)
From the read timing generator 190-2 to the signal line 102,
103i, to C via 04 I) Driver 8240
．． COD driver (]250, CCU driver 3
(, supplying 260?,). These Nos. 113 and 113 will be described later. Address: i'lJ! A I) Jyu 101-1 is 1
1il+il<41. -5°I)47
This is an address line that counts 52 bits. This signal is transmitted to the shading correction circuit 1 via the live wire 101.
30->Supplied. The shading start signal 5HDS, i" 401 is a signal generated when scanning the shading start signal 5HDS, i" (Fig.
l-2 to the shading correction circuit 1 via the signal line 101.
Output to 30. 0ODVIDEOEN117 is COD
This is a signal indicating the section in which 4752-bit data is output per line from the multilevel processing circuit 18.
0 via signal line 117. FIG. 5-2 is a timing chart showing the timing of each part of the synchronous control circuit 190. 2φT4 is an image transfer lock, and the image transfer lock 2φT4 transmits the beam detect signal B for each line emitted from the laser scanner.
, and generates a 1-clock horizontal synchronization signal H8YNO. This signal is also the COD read start shift pulse SH. φ14. Each signal .phi.2 is a signal having a period twice as long as the image transfer lock 2.phi.T+ and has a different phase, and is used to shift the analog shift registers of the odd and even parts of the COD, which will be described later. The image data signal VIDEODATA from the COD is such that the first &th image data D1 is read out from the output of the shift pulse 5t-f4, and the first &th image data D1 is read out from the output of the shift pulse 5t-f4, and the image data D2. D3. ...and 50
The 0θ bit is read out, but D1 to D4 are COD dummy pixels, and 4752 bits from D5 to D4.756 are one line of image data, and this section 00D
viogo EN@ becomes active. Signal box S →
is a pulse that resets the COD shift register for each shift and is generated at the trailing edge of image data. The shading start signal 5HDSTW is a signal input from the main body control section 400 as described above, and is a signal generated during the period of OOD VIDEOEN8 of the first line that becomes active. Next, details of the COD light receiving unit 200 shown in FIG. 4 will be explained. The COD light receiving unit includes a dichroic filter 12 for separating three colors, and a blue filter 1 for adjusting the amount and intensity of B, G, and R light obtained by the dichroic filter.
3. Green filter 15, red filter 17, CCDB 210 that receives blue light, 0ODG 220 that receives green light, 00D filter 2 that receives red light
30, and A/D converts these outputs and converts them into digital amounts of complementary colors yellow (1) and cyan (Q1 magenta). 0ODB 2,10, C0DG22
0, C0DR230 are each 00D driver 8
240, OCD Driver 0250.0 is installed in OD Driver R260. Figure 6-1 shows the structure of each 00D. In the figure, what is the original image that has passed through the infrared cut filter dichroic filter 12 and the spectral correction filter? D1 ~ Irradiated as an image. The photocurrent of the photodiode is accumulated in a charge storage part (not shown) in the form of a charge proportional to the irradiation time, and the MO8S
By applying a shift pulse of H, the analog shift register OCD 5bit 1%eg 1, and 2
The charge is transferred to Continuous pulses with opposite phases, Mssφ1 and MO8φ2, are applied to OOD 5hift Regl and 2, and the image '1 section charge transferred from the photodiode charge m type section is applied to this clock pulse MO.
C0Dshift1(, eg
It is transferred in series to the output transistor circuit Q1 along the '1-11 charge well formed in the channels 1 and 2. At the same time, a switching noise component due to the reset signal RS corresponding to the image charge is applied to the output transistor circuit Q2. This switching noise component is later used to cancel out the switching noise component mixed into the 11i1 image load. Clock pulse MO8φ1. The image charge transferred to the output transistor circuit Q1 by MO8φ2 is converted into an image voltage output VS here. In addition, the corresponding switching noise component is also output as a switching noise voltage VMS by the output transistor circuit Q2.
It is converted to '. In addition, a reset pulse MO8BS is applied to the output transistor circuits Ql and Q2 each time one image charge reaches the output transistor circuit Q1 and is converted into voltage, thereby preventing image charge accumulation in the output transistor circuit Q1. . FIG. 6-2 shows a block diagram of a COD driver that converts an original image into an electrical signal in an embodiment of the present invention. 201 is a dichroic filter 12, and a COD linear image sensor IMSENSI 202 is a COD linear image sensor that converts the image light that has passed through the light intensity adjustment filter into an electrical signal.
203 is a differential input video amplifier V-AMP that differentially amplifies VNS and creates a correct image output voltage VIDEO; 203 is a video A/D converter A/D-0 that converts the image output voltage VIDEO from an analog signal to a digital signal; 204
is a reference voltage source V-REF that supplies a conversion reference voltage to the A/D converter 203; 205 to 208 are pulse drive amplifiers for operating the IMSENS 201, 209 is a variable resistor VR2 for eliminating the DC voltage difference between the image voltage output VS, which is the IMSENS output, and the switching noise output VNS, and 210 is a V-AM
This is a variable resistor VR1 that sets the amplified output of P. Image output v from IM8ENS201 in the above circuit
S and the noise output VN8 are applied to the V-AMP 202 after the DC voltage levels at the time of no-light signal are equalized by VB2. The V-AMP 202 differentially amplifies the S and VNS, attenuates the noise component contained in the image output VS, and uses the VRI to create an image credit card ViDEO that is compatible with the human power of the A/D-0203. In this embodiment, as described above, the dichroic filter 12
However, due to the characteristics of the dichroic filter 12, the color sensitivity characteristics of the 00D linear image sensor in the 00D driver, and the characteristics of the light source, B
By using V-AMP202, the optical input versus electrical signal output characteristics of the three COD drivers for , G, and BK are adjusted so that they are accurately proportional from the no-light state without being saturated even when receiving the maximum amount of light, and have an appropriate dynamic range. Motsuyo B, G
, R, select the resistances of VRI and VB2, and Blu u
The gain is adjusted to decrease in the order of e * Green y Red. The VIDEO signal, which is an analog signal, is A/
D-0203 is converted into a digital signal. The timing of conversion is the timing according to the image transfer lock 2φT sent from the address control unit 190-3,
The VIDEO signal converted into a digital signal is transferred to the image data processing unit 100 and subjected to various image processing steps. In this way, by adjusting the gain of the amplifier so that B>G>R, the characteristics of the light source, etc. can be corrected. In this embodiment, high-speed A/D converter A/D-020
3 has an output resistance lower than that of the reference voltage source V-REF204.
1(, EF, 1/4 REF are applied, and the respective signals φ1, φ2R8, SH are converted into appropriate drive voltage waveforms M using pulse drive amplifiers 205 to 208.
O8φ1 , MO8φ2 +MO8I(, S , MO8
After setting it to 8L■, it is powered by human power. (Shading Correction) FIG. 7-1 shows a diagram of the principle of shading correction performed in this embodiment. In a device that reads an image by irradiating a light source onto a document and condensing the reflected light image with a lens, a non-uniform light image called shading is obtained due to optical problems with the light source, lens, etc. In FIG. 7-1, when the image data in the main scanning direction is 12...n·4756, the amount of light is attenuated at both ends. Therefore, in order to correct the shading, the shading correction circuit 130 performs the following processing. In Figure 7-1, M, A, and X are the maximum values of the image level, and Sn
is the image level of the η-th bit when the white correction plate 4 is read. If the image level when the image is subsequently read is Dn, the corrected image level D'r is D'H-Dn.4. Correction is performed for each bit so that (4t,) is obtained. Details of the shading correction circuit 130 are shown in FIG. 7-2. 130-2, 130-4, 130-6 are shading RAMs for reading one line of the soft white correction plate 4, 130-1, 1:30-3, 130-5 are stored in shading RAM when reading an image. This is a shading correction ROM that refers to the shading data and outputs 41i data. CCD driver B240. C.C.
D driver G250°CCI) The 8-bit image data read by driver RA is sent to signal lines 271 and 2, respectively.
72, 273 to the sonification correction circuit 130. First, image data obtained by reading one line of the white front panel 4 is stored in the shading RAM 130-2. ]
]30-4, 130-. At this time, the Sony write enable signal S is sent to the signal line 1.01-2 from the address control section ]-90-3 (Figure 5-1).
WE is input. -Also, image transfer lock 29 T is input to 4r3 J-Z-line 10 knee 1-3, and Nantes gate 130-
Gated by 20. Nando Kate 130-2
0 output is each shading RA M2B5-2.13
0-4. It is connected to the write enable terminal WE of 130-6, and when one line of the white correction board is read, the shape-in data is stored in the RAM of Nora and Mikoko. At this time, the address signal ADRIOI-1 is controlled by the address control section 19 (1-3), and the 475
Two-pixel image data is stored in each shading RAM. Signal lines 271°272. from the CCD light receiving unit 200.
The image signal VIDEOY output to 273. VIDEOIVI and VIDEOC are all 8-bit digital signals, and the valley bits of each signal are
EW)0~VIDEO7(LSB-MSBIl[
We will call it l'l). In this embodiment, there are seven grading RAMs 130-2 for shading data. ] ]
When storing to 30-4, 130-, signal lines 130-8,
130-10. 6-bit digital data VID via 130-12
EO1 to VIDEO6 (7) 0 is stored as shading data in each RAM, one pixel at a time.The reason why the 0 shading data is set to 6 bits is to reduce the storage capacity and to prevent the shading characteristics from changing sharply in the circuit. It is. When you start scanning the document after storing the shading data,
Image data VIDEOY, VIDEOM. Vll) Each 8-bit data VIDEO of EOC
O~VIDEO7 is signal line 130-7.130-9.1
Shading correction ROM 130-1 via 30-11
, 130-3130-5 are input to address terminals AO to A7. On the other hand, Shiniding RAM 130-2, 130-4゜1
The 4752-bit shading data stored in the shading correction ROM 130-L is controlled by the address signal ADHIOI-1, and is transmitted from the terminals %□ to %6 to the address terminals of the shading correction ROM 130-L 130-3 and 130-5, respectively. It is output to A13. At this time, the shading light enable signal 5WEIOI-2 does not become active and the shading light enable signal 5WEIOI-2 does not become active.
2,130-4. xjo-6 performs a read operation. Shading correction ROM1: Hl-1, 130-3, 1
In 30-5, ROM data is created so that the calculation shown in equation (4-1) is performed, and the 8-bit image signal VIDEOO~Vll)EO7 and 6-bit shading data are used as addresses. By accessing the shading correction ROM, the shading corrected output is output as an 8-bit image signal from terminals 01-08. In this case, it is performed every time the document is scanned. Further, this shading correction is performed on all image data. (γ Correction) Next, γ correction will be explained. FIG. 8-1 is a block circuit diagram showing details of the γ correction circuit 140. In this embodiment, γ correction is performed for each color using a reference ROM, and the γ characteristics can be arbitrarily selected. The 8-bit signal V I DEOY output from the shading correction circuit 130 is the synchronization signal 2φT output from the synchronization control circuit 90 to the signal line 119 at the launch 301.
Synchronization is achieved by The synchronized output is γ
It is input to the lower address 8 bits of the correction ROM 302. Further, the γ correction selection signal 403 output from the main body control section 400 is inputted to the upper 2 bits of the address, and the 1 area of the γ correction ROM 302 is selected in accordance with this signal. That is, the yellow switch 421-. 14 is a four-stage (selectable in the past, and is accessed by a high-speed digital signal input to the upper 2 bits and lower 8 bits of address of γ correction I (0M302) and stored in the ROM 302.
1: Sushi data is output. The data outputted from the ROM 7:l- is at a 6-bit level. This data is further transferred to the signal line 119 by the latch 303 to L4.1.
] The synchronizing force S is generated by the synchronizing No. 111 2φT, which is applied to the masking circuit 150.
The DEO'l' signal is output to the signal line 108. In this manner, the γ correction ROM 302 converts the yellow (Y) (H component) into data. Similar processing is also performed for the image signals VII) EOM and VIDEOC. That is, shading circuit 1
Image signal V output from 30 to signal lines 106 and 107
After the IDEOM and VIDEOC are synchronized with the latches 304 and 307, they are input to the γ correction ROM 3 (15 and 308).
-15,421-16 selection signal and image signal VI
According to DEOfvT, V T [) E (JC), the area of ROM for γ correction, 305 r 31 J 8 is accessed and γ corrected 6-bit data is output. After this γ correction, ■■DEOM, VfuEOCi11
After the same Jtlj is taken by the latch circuits 306 and 3 (19), it is output to the masking circuit 150 via the signal lines 10.9 and 110. Next, the sub control unit of the main body control unit 400 7
3 γ value control switches 421-14 to 421
-16 selection and γ correction ROM302, 305, 30
8. Address Input Data and Output Data Conversion Table 0 Here, as an example, the image signal VIt)
The EO1°C γ sleeve correction R correction 302 will be explained. First, gamma correction is a method to read a color original and calculate the density (abbreviated as OD) of the original when expressing it on transfer paper.
It is desirable to express the image on the transfer paper so that the density (abbreviated as CD) when expressed on the transfer paper is one-to-one. In this case, the characteristics of the COD ¥210 that reads the color original density, the characteristics of the image processing unit 100 that outputs the signal obtained from the CCD as a laser modulated signal, and the characteristics of the image processing unit 100 that outputs the laser modulated signal and expresses it on the transfer paper. Three characteristics of image density are problematic. Regarding this point, Section 8-2
This will be further explained with reference to the drawings. In the figure, the vertical axis of the fourth quadrant is 0. D, and the horizontal axis is
Represents a shading-corrected VIDEOY. Since the original density is O and D is expressed in logarithm, the image signal V I D
EOY is logarithmically related to the document density. This characteristic is fixed depending on the characteristics of the CCD'Y-210 and the CCD driver 240. Further, the second quadrant represents the relationship between the dither cumulative frequency and C and D. Here, the dither cumulative frequency is expressed as a ratio between a certain constant area (here, a dither matrix expressed by a dither processing circuit 170 described later) and a developed partial area within that area. So, dither!
'IY I,'j-When the uncle changes from 0% to 100% C1 ■) ・q: If you take the change, at 0% C and D are white level and when you gradually increase the teaser cumulative frequency from 0% , due to the characteristic that the CD starts suddenly from the middle, 10 (
'1% saturates at a certain concentration. This characteristic is fixed depending on the photosensitive drum 24 and the yellow image device 36. For this reason, if the image processing unit 100 cannot change the characteristics shown in the first quadrant, C in the third quadrant,
The relationship between D and OD is fixed. In the image processing unit 100, the γ correction circuit 140 and the dither processing circuit 1.70 are particularly capable of controlling the relationship between the COD output and the dither cumulative frequency. However, since the data handled by the teaser processing circuit is 6 bits, the quantum error is large when trying to correct the nonlinear part of the 2nd and 4th quadrants.
Even if the 15' to 1 coefficients of D, O, and D are linear, there is a drawback that they cannot be expressed faithfully. Moreover, the input data of the γ correction circuit 140 is 8 bits,
The output data is 6 bits X, b K ii positive or 1.1
15 of the signal output as the teaser cumulative frequency for the signal
If the −1 coefficient is a linear relationship, the characteristics of the first quadrant are γ correction R
It is determined by the data stored in OM:302. Therefore, if the relationship between the dither cumulative frequency and the output of the CCD in the first quadrant is made into the characteristic A by γ correction, the third
Quadrants C and D and ODO team 1 correspond to 1:] as in A', and a corresponding σgiru Hayabusa is formed. Next, 1. of the table. (As a side note, Table 1 shows the contents of the ROM 302 for the γ interpolation 11-. The 2 bits ``I'' to ``I'' on the address indicate its lh' property, and "00" indicates the A
. B for “old”, C for l” 1.01, D for 11,tj
represents. Image No. 44 with yellow in the lower 8 bits VI
When DEOY is input, 6-bit data as shown in Table 1 is output. In this way, the relationship between CD and OD can be achieved on a one-to-one basis. In addition, various copy density CDs such as B' in the third quadrant where the copy density CD becomes low, C' with high contrast characteristics, and characteristics with almost fogginess are adjusted by the γ correction of the sub-control unit 73. This is made possible by selecting switch 421 1'4. In this way, by using a gamma correction circuit for yellow signal characteristics, high-speed and faithful copying of originals (( It goes without saying that this is possible. Table 1 Furthermore, the relationship between the CCD output and the dither accumulation can be controlled by the γ correction circuit 140 and the dither processing circuit 170.As a specific example, the original Density OJ) and output signal after shading correction VII) EU Y
is not a linear relationship, J: is stored in the r correction ROM 302.
Then, the signal is converted so that the VIDEOY signal corrected by dVIDEoY signal 4 is proportional to the original density using the method described above.
A dither processing unit to which the EO signal is supplied from the signal line 114
M170 has VII) contrast density C and D for EO signal.
is proportional to the oil I by the fizer processing circuit VC described later.
It is also possible to do F. (Masking) The printing ink or toner blue coloring material has a Zi 1 spectral reflectance as shown in Figure 9-1. That is, the Y (yellow) coloring material is 400-'500 nm
t7) Absorbs light and reflects light from 500 nm. M(
Magenta) coloring material absorbs light between 500nm and 600nm and reflects the rest; C (cyan)
It absorbs light from 00 nm to 700 nm and reflects light other than that. On the other hand, when developing with Y color material, the reflected light from the original is
The spectrum shown in Figure 2-14 is Meru (IS
) filter for the latent image formed by the color separation ij'N beam image, and in the same way, for the color H of M:, green (U) filter, and the color material for C are red i (L filter). The latent image formed by the color-separated ill L)C light image is developed. Here, as can be seen from both figures, B, O
, l('s filters t are 500 nm and 60 nm, respectively.
0 nm, the color component resistance is relatively good, but the spectral reflectance of the coloring material is due to the separation property due to ripples, and the spectral reflectance of the coloring material is higher than K M (magenta) with I/'Io. t'', a considerable Y (yellow) component and CJ 戊 component makes it look great L% 17
This C (cyan) also contains some 84)z, i and y77 components, and simply corresponds to the color component Pi' (upper,
1 [With 4 color materials, when developed, it will be a unique color / t
rJ, tM color image threshold becomes cloudy. Therefore, in printing technology, masking processing is performed to correct this. This is a masking process (each color component manually applied to 41 threads, Yi, ~11°C
1, each output color component Yo, Mo,
C. is converted as shown below. Coefficients (ai, bi, ci) (i=1
If ゜2.3) is set appropriately, γ1f as shown above
A blur is created to correct the curvature. 10-i shows details of the masking processing circuit 150 and the V Ci (processing circuit 160, which will be described later). In the figure, 150-Y, 150-fVL, and 350-C are yellow (Y), This is a masking processing unit for the iI!ji image signals of magenta (M) and Nijian (C), and mass = v-/goku, (Ko Fi+; L5o-Y, 'l
it ”i monk 108 gold plate 1 C output luxury 6
Binoto's Y component video No. 1 IJ Month'io Y, I, J No., output via gland 109) L6 bits!
vl Itij minute Viteo count VIJ Togawa M's position 4 bits, 44 ';' i, '3I! 6-bit C component hito signal Vll) EO outputted through all J, IQ
All 4 bits of "C", respectively; tL Yi, Iv
As i, i, Ci, formula (3) is realized, tJ
, 6゜1111 color data, for example (3)
In the formula, i yohe4i, Ui, f4) Y' +
C', (j>) expression TeVJ, Yi, 1Vit, 1
, 1. compared to the bending stop data Yi, Mi, Ci.・4.・、I41i Ii month 1! , cheaply, and the coefficients (air') i) (j-1, 2, , i) are also set as 1/16.2/16. ...1d', 'C
'l 6 i'9: '1'+j, so the data to be corrected, Yl. For each 0 bit of IV〆Ii, L'i, cross 4 bits.') l, -C, mark. - Yet this allows for conversion! (・(,) Nl (ln (small)
The circuit shown in Figure 1 reduces the capacity to 1/4 of the massing process 1'1 in Figure K#.
Block times showing 1 unit +50-Y in detail::'?
i l':l, -J
50-JA. Since 150-C is also the same circuit, explanation of 150-4 to 4°150-C will be omitted. In the child 9-21, the signal source 150-1 is used for the masking processing unit.
0 through 6-bit Y data, I word live wire 15'0-
4 pins for correction via 12) M data, signal line 150-
14 through 4-bit C data and signal lines 150-11
．． ．． 150-13, 150-15, 4-bit code data SYY, SYM,
8YC is input. SYY + SYM + SY
C code data

[0]H~CF'3□ are respectively (
3) Give the coefficient (aI+"l+CI) in the equation and convert the digital code switch Δ21-5 to 421-13
If the set value is N, then the coefficient is given by N/16. l 50-1. L'50-2゜150-3 is a ROM for performing calculations, and 150-1 is a 6-bit Y
The signal and the 4-bit code data SYY form the address bus of the OM, and are designated by this address. In the R01V data, when the setting value of 4 bits is m, IJy "' Yr+bit Xm/l 6
(Y-=:011-3F H,m=(JH-10″l
) is stored in 6 bits. Same ship [Ni150-2
dl, 4-bit code data S y M for the set 41 value n, 1): 110M 41) + t -C4bi
t X i/ 16 is fully stored, and J) In and l) c are each 4
It's bit. Each j-ta Dy, I)r obtained here
I. 1). Kebil≦live wire 150-16. 150-
17. 150-18. And for these data, (3) j round (Koikeikan: ii-■nai, IJ218-')Ill J). Then, all of them are C-noted, I'j, A,' 1 Now, let's take the video data of Y as 7L illusion:, Regarding Y, we can prefix the expression fl). -1:/hi, M, CKeIshidemodo (thousand)
There is no sleeve IF. That is, the above 6-bit Y data and 4-bit each of the 11th I and C data are the rectangle Q-1.
are also connected to the address buses 01 to 4150-4, and (2) also obtains a predetermined calculated value using the OM table horizontal system. Reference numeral 150-5i is a latch element that latches the 6-bit φT data numerically calculated for masking processing in synchronization with video transfer lock 2←. (UCI is also processed) Figure 10-1 shows 1-iC1, (processing Iu path ri't-
+l'nr: is shown. Normally, when color reproduction is performed by subtractive color mixing method using the peak colors of coloring materials, for example, when Y, M, and C are arranged equidistantly, black (BK) is used because the hue absorbs all the spectral components. ) is written down. Therefore, the B f (part of the original is Y, M, C toner, etc.)
is equal to 1 for the door. However, Y! Vl, C
As shown in Figure 9-1, the spectroscopy and emission of the toner is as follows:
Poor color separation depending on wavelength, that is, there is some M in Y toner.
It has already been mentioned that the component M toner contains the Y component and the C component. Therefore, for the black component, color machining is performed using black toner, and the areas where black is used are the corresponding Y, M,
This is called undercolor removal (U CH,) and is performed at block 160 in Figure 10-1.
I'm doing it. Butsugo Hori 160 30, 160-3i, 1
60. s 13 through 32! J masking circuit 1
Y, M, C6 binoto image data output from 50≦Z
Comparators 160-1, 160-2, and 160-3 compare the relative softness of Y, M, , IVI, C9C, and Y, respectively. The comparators 160-1, 160-2, and 160-3 select the smallest value of the image data Y, M, and C from the latch circuit 16.
0-13.160-14.1 (to latch to io-15, size ratio / shark is performed, Y, M
, C, signals as shown in the table of FIG. 10-2 are sent to the signal lines 160-33. i 60-34
, 160-35. In other words, one song of Y, LIJ, C of 11IiII Sono (& 1 H for data comparison, Y
However, when it is minimum, fTh number 1G0-33 is "0"
"But, fg envy, 160-35 and 1" is the same, swimming is M
When the school was in elementary school, there was a signal line 1 on the Buddha Expressway 160-33.
When 60-34 has 0" and 1C is the minimum, signal line 160-
34"]-' % No. 44 + gland 160-35 "0"
is output. Further, when Y=M=C, Y:1 [1 is representative. the three comparators 160-1, 160-2,
At step 160-3, the minimum values of y, M, and C are determined and output from the latch circuits 160-13, 160-14, 160 = 15 as a signal aJ 16036, and this value becomes the basic data for inking. Another latch circuit 160-10,1
60-11.1 GO-12 transfers the image data of Y, +J, and C output from the masking circuit 150 and latches it at the rising edge of the lock 24-F to the subtraction operation ROM 160-16, 160 of the next stage. °-17
, 160-18. Also, signal, Su 160-3
In the basic data for inking (BK) outputted in step 6,
The coefficient value given in 4 bits to No. 46 and weight 6o-37 via selector 160-20 is
19, the value of the upper 4 bits of the 6 bits of the value k
Interchange L ”'ir subtraction approximately 0M160-16゜160-1
0M160 I for 0 reduction y- output to 7,160-18
-16, 160-17, 160-18, this value is subtracted from each image data and the result is applied to the signal line 160-3.
0 selector 160-21 K outputs to selector 160-21 via 9 I! 4.1 0M160-19 inputs 6-bit data for finalization via signal 160-38 These 1+fii image signals
Y, iv output via signal line 405 ff:
I, C, Bf (identification signal 5ELBK, S, I,
Safe image data is selected by Y, SELM, and SELC, and output as a 6-bit signal from the selector 160-21. In this case, the selection signal SELY is sent every time both images are scanned once. 8tL M, S, li; L C, Sli; L Bl (
is output, and Y. The color-corrected image data is selected in the order of iVI, C, BK. This is the coefficient selected by the switch flA2421-]~4 in the unit 73, and this is also the main body! II:J
, li: bII switch group selection signal 40 outputted from the first section! i-9, 405-10 and given to Raikan 11, 01'141 C50-19. As explained above, the OCR circuit 160 according to this embodiment
Now, for pixels having color components as shown in Figure 10-3, inking is performed using the value obtained by multiplying the minimum value, for example Y, by a certain coefficient k as BK, and Y is (Y=
BK), M is performing fγ with ([\1-BK), and C is (C-BK) as the final color component. (Multi-gradation) FIG. 11 shows a principle diagram of the multi-gradation processing of this embodiment. The multi-gradation process in this embodiment consists of a teaser process and a multi-value process. An example of teaser processing is shown in FIG. 11(a). In teaser processing, when a 6-bit 64-level (0 to 3F) digital image signal is binarized, the threshold value is changed within a certain area, and the floor area is determined by the area ratio of the number of dots within that area (hereinafter referred to as dither matrix). I'm in tune. 1st) A in Figure (a) is a 2×2 dither matrix in which the threshold value is changed to 8, 18, 28, and 38 degrees for each dot. For the values of 0 to 3F of the digital image direction signal Dn, if the white background in Figure 9 is set to ``0'' and the shaded area to ``1'' using the binarized signal, then (a) - (0) to (a) - (4
), five gradations can be obtained. The larger the dither matrix, the greater the number of gradations, but on the other hand, the resolution of the image deteriorates. Therefore, in the present invention, one pixel is further divided and the gradation is increased by Noculus width modulation. 1st→
Figure (b) shows an example in which 3-division pulse width modulation is performed and quaternary taser is performed. Divide one dot into three parts along dotted lines as shown. In other words, the area ratio of four gradations can be obtained in one dot. If three more threshold values are given in each dot of the 2×2 dither matrix as indicated by B in FIG. 3, an image with gradation characteristics can be obtained by emitting laser light only in the shaded area in Figure 1-0 (b) in a binary signal that has been converted into multiple gradations. ing. In addition, in the case of a ternary dither matrix, one dot is converted into two
FIG. 3 is a block circuit diagram showing details of a 31-invalidity processing circuit 180 that obtains a matrix by division. In the figure, a 2-bit signal YMCBK is sent from the main body control unit 4-00 through a signal line 406 (Figure 4).
O(A 10 ), YiVICBK 1(Al
1) to determine the color to be dithered. For example, if A+o "" I A11-1, then Y (yellow A+o= I A11= 0, then M (-r zenta) A, .= OA, if l= 1, then C (77 n) A, . = OA1□-0 Then, set it to BK (black). Also, switches SWI~3 are switches for selecting gradation, and have two contacts, a and b. By turning on switch SW1, one dot of the dither matrix is divided into three. By turning on the switch 5w2i, one dot of the teaser matrix can be divided into two.
W 1 on. S Vl/2'5]-7. Consider the case where SW3 is off. In this case, if the teaser ROMA-C is selected yellow and the video signal is 6 bits (64 level signal), oo is placed at address 00 of the dither ROMA, and 03.02 is placed at address oig.
Q6.03 for incense, 09 for incense, 12.21 for 20 incense
15. oi at address 00 of the dither ROMB,
04° in the oi search area, 07 in the 02 area...02 in the 00 area of the dither ROMC, 05,02 in the area 01 in the area? ,
Store dither patterns such as IR area 08... The circuit operation under the above conditions will be fully explained. If the video signal VIDEOO~5 is 04 in this state, the contents of dither ROM A at OO#1'r are OO
When compared with , video conversion 4 is larger, so the output Q of latch A is '1''. Also, at this time, dither RO
Since the video signal is larger than the content 01 of MB's 00 city, the output Q of Jituna B is 1. Also, since the video signal is larger than the content 02 at address 00 of this auxiliary dither ROMC, the output Q of Latte C is tt
1 becomes #. In synchronization with the next image transfer lock WCLK, the output Q of latch A becomes '1' compared to the content 03 of the 01 f + - ground of the dither ROMA. Also, at this time, the content of the 01 f + - ground of the dither ROMA becomes '1'. Since they are equal, the output Q of latch B becomes "0''. At this time, when compared with the content 05 of address 01 of the dither ROMC, the output Q of latch C becomes o". In this way WCL
Dither ROMA, B in synchronization with K. C each h tD 02 t”F address o3, oo address, 0
The contents of addresses 1, 02, 03, and oo are sequentially compared, and as a result, the outputs Q of latches A, B, and C become "θ" or 1. At this time, when the row siw'c signal is input, the address Counter B170-8 counts up by one and WCLK
20th, 21st, 22nd, 23rd,
The contents of address 20 are sequentially compared. In synchronization with the image transfer lock WCLK, address counter A17O-7 (lower address) (x0 address to x3 address) counts up, and every time H8YNC is input, address counter B17O-8 (upper address) (oX Address to 3×Address) counts up. At this time, latch A170-4. B170-5゜C170
-6 is output to line memory A180-9 by counting up the address of line address counter c180-7 in synchronization with image transfer lock WCLK.
．． Stored in B18O-10, C180-11. When the H8YNC signal is input at this time, the latch A170
-4. B170-5. C17O-617) Each output is a line address counter DI80 in synchronization with WCLK.
By counting up the address of -8, f7
Memo +7D180-12. B180-13, B180-
Stored at 14. Line memory D 180-12.
B180-13. During the sequential storage in Fl 80-14 in synchronization with WCLK, line memory A 1 is stored first.
80-9. The nft contents stored in B18O-10 and C180-11 are the signal RCLK from the oscillation circuit 180-3.
By counting up the addresses of the line address counter C180-7 and the read address counter 180-5 in synchronization with the data, the data is sequentially sent to the data selector 180-15. The read address counter 180-5 starts counting up because in order to form an image at a fixed position on the drum, it is necessary to delay the start of image formation by a certain period of time after H8YNC is input. The read address counter 180-5 is prohibited from counting up until the value of the left margin counter 180-6 reaches a predetermined value. After the pinch inhibition is lifted, the contents of the line memories A, B, C, or E, F are sent to the data selector 180-15. This data selector 180-15 is switched between the A side and the B side by the switching circuit 180-2 every time H8YNC is input, so the data selector 180-15
The output terminal of the line memory A18 is synchronized with RCLK.
0-9°B18O-10, C180-11 or line memory D180-12. B180-13. B180-
This means that the signal stored in either one of 14 is always output. The multilevel oscillation circuit 1so-16 has a contact 5WI-b (4oo
-6) is ON, the image transfer lock WCLK is activated by three signals φA as shown in FIG. Divide it into φB and φC and use it for anime games) A180-17,
AND gate B18O-18, AND gate C180-
Send to 19th. As a result, the outputs Yo, Y, Y2 synchronized with RCLK of the data selector 180-15 are gated by the analog signals A, B, and C, respectively. Next, the result is input into the OR gate 180-20, and this OR gate is. By turning on the laser with the output signal from -20, W
VID input to the comparator during one CLK wave
Depending on the signal size of EOO~5, the area irradiated with the laser can be changed into four types (■ No irradiation at all, d) Irradiation for 1/3 of RCLK, ■ 2 of RCLK
/3 time irradiation, ■3/3 time irradiation of RCLK). FIG. 13 shows a time chart of the signals explained above. To explain the signals again from above, they are B, D...
...H8YNC that occurs every time the laser beam scans the drum once...
After B and D become H, the first φ1 becomes H only while it is H. VIDEO ENBLE... Stores video 1 after de-processing into the line memory only while the signal is H. Laser output......This signal is H
Image transfer lock that irradiates a variable laser beam onto the drum only for a while... 3CJJ is applied to this signal and teases φ1-...・・・・・・
...Same side as this signal" from line memory 41:
Without taking out the numbers, φA, φB, φC・・・・・・・・・・・・φ,
The frequency of the A word taken out from the line memory is divided by 3 in synchronization with . Next, a case where the area irradiated with the laser during the image transfer lock WCLKI wave is changed to three types will be described. In this case, the switches SW1 to SW3 are S%' OFF,
SW20 N, SWs OFF. Other conditions are 5W10N, SW20FF, 5V
The explanation is the same as when Js is OFF. Under this condition, the dither ROM is D 170-12 E 170-1
3 is selected in the write address counter 180-i.
, read address counter 180-5, left margin counter 180-6. Switching circuit 180-2. Address counter C180-7, address counter D180-'8
The function of is exactly the same as the previous explanation, so it will be omitted. V
The result of comparing I DEOo-s with the contents of the dither ROMD 170-12 is sent to the terminal A of the data selector 180-15 via the latch A 170-4 line memory A 18O-9 (or line memory D 180-12). (or B
. ) is entered. Similarly VIDEOI~. The result of comparing with the contents of dither ROME170-13 is latch B 170-5 line memory B 180-10 (
The signal RLCK is input to the terminal AI (or Bl) of the data selector 180-15 via the line memory E180-13), while the multi-value excavation circuit 180-16 outputs the signal RLCK when 8w2° is ON. As shown in FIG. 13, at this time φ is divided into two signals φ and φ8. remains at 0. As a result, data selector 1
Output Y synchronized with RCLK of 80-15. , Yl are gated by the ant gate 180-17 and the seventh ant gate 180-18, respectively. Next, take the result with OR gate 180-20,
By turning on the laser with this signal, image transfer is possible.The VI input to the comparator during one wave of lock WCLK is
The area irradiated with the laser can be changed to three types depending on the signal size of DEO, ~, (■ No irradiation at all ■ Irradiation for a time of /2 of RCLK ■ 2/2 of RCLK
irradiation time). Next, a case will be described in which the area irradiated with laser during the image transfer lock WCLKI wave is changed into two types. (j
7) If Sw! ~3 is SW, OFF. SW20 F 'h'', SWs ON.Other conditions are 5WION, SW2 OFF, SWM
This is the same as when it is OFF. Under this condition, dither R
Dither ROM F 170-14 is selected as OM. Write address counter 180-1, read address counter 180-5, left margin counter 1
80-6° switching circuit 180-2, address counter C
180-7 Address Counter I) The function of 180-8 is exactly the same as the previous explanation, so a description thereof will be omitted. V I D E Oo ~5 and dither ROM F 1
70-14 (1) The result of comparing the contents is latch A17
0-4 line memory A18O-9 (or "; r i7me%
sl) 180-12) to the data selector 18
0-15 Tsuyoko A. (or B.). On the other hand, the multilevel oscillation circuit 180-16 has 5w3-b turned on.
t, when 瀉 is “1”, brass is 0”r Y2 is “0”
Since it does not change in the state of 180-17, it does not change in synchronization with RCLK, then outputs 180-17, then performs an OR at OR gate 116. By turning on the laser with this signal, one wave of WCLK is generated. Depending on the magnitude of the signal VIDEOO~ inputted to the comparator in between, the bright area (I) on which the laser was irradiated or not irradiated (with no change in content) is determined. The manuscripts to be copied can be broadly divided into six categories as follows. That is, those with only one picture, those with only two letters, and those with a mixture of six pictures and letters. Pictures can also be further divided into those with delicate hues, such as photographs, and those whose images are composed almost exclusively of primary colors, such as manga and pictures.As for photographic manuscripts, By converting to multi-value, the gradation is improved and subtle color changes can be faithfully reproduced.Also, it is possible to convert to binary for originals with almost only primary colors, such as -7 color and nushi pictures. It is possible to express clean colors without cloudiness.For text, it is also possible to express a sharp image in black and white without any intermediate or dark areas, so by switching switches SW1 to 5W11 depending on the type of document. Optimal image reproduction is possible and 42'7. ). The switches SWt to Sakai 3 are turned on and off by turning off the switch 421-24 in the sub-control unit.
When it is set, SWt is on, when it is set to the scale, switch sw2 is turned on.
is on, and when the scale is set to 2, the switch so-called 8 is turned on. Note that although this embodiment has a configuration in which images are recorded using a laser beam, the present invention is not limited to this. For example, it can be applied to inkjet printers and thermal printers. Further, the masking processing and the UCR processing circuit may be performed in either order first. Furthermore, the B, G', and R signals may be transmitted from the host computer's memory, etc. Also, each data of Y, M, C, and B may be stored in a single page memory, etc. The configuration may be such that the data is read out later.Also, the configuration may be such that it is not only recorded on transfer paper, etc. (or it may be filed on a disk.In addition, in this embodiment, multi-gradation is achieved by using time-division signals. However, it is also possible to use brightness modulation, etc. Effects As described above, according to the present invention, when performing dither processing, color image data and a plurality of threshold matrices ♂ are simultaneously compared, Since the gradations are expressed in a time-division manner, it is possible to reproduce images at high speed and with high quality.Furthermore, since the number of gradations or the number of threshold matrices is variable, it is possible to reproduce the number of gradations according to the number of gradations in the reproduced image. In addition, since the dither processing is transmitted using a different dither matrix for each color, moire etc. can be reduced.Also, since each color data is converted into multiple gradations using the same multi-value processing means, , the configuration is simple.

[Brief explanation of drawings]

Figure 1 is a sectional view of a color copying apparatus to which the present invention is applied, Figure 2-1 is a diagram showing the spectral characteristics of a halogen lamp and the spectral sensitivity characteristics of a CCD, and Figure 2-2 is a diagram showing the spectral characteristics of a halogen lamp and the spectral sensitivity characteristics of a CCD. Figure 2-6 is a diagram showing the spectral sensitivity characteristics of CC/D when
Figure-1 is a block circuit diagram showing the main control unit, Figure 6-2 is a diagram showing the operation unit of the main control unit, and Figure 6-2 is a block circuit diagram showing the main control unit.
-Figure 5 is a diagram showing the operation section of the sub-control unit;
6-4 is a timing chart showing the operation timing of each part of the color copying machine shown in FIG. Figure 5-1 is a block diagram showing the schematic configuration of the synchronous control circuit, Figure 5-2 is a timing chart of signals in the synchronous control circuit, and Figure 6-1 is the structure of the CCD. Figure 6 showing
Figure-2 is a block diagram of the COD driver, Figure 7-1 is the C
Diagram for explaining the light amount distribution on the CD surface, No. 7-
Figure 2 is a block circuit diagram showing the shading correction circuit.
Figure 8-1 is a diagram that does not show the relationship between γ-corrected image density;
Figure 9-1 is a diagram showing the spectral reflection characteristics of toner, Figure 9-2
The figure is a block circuit diagram showing the masking processing circuit, No. 10.
-01 is a block circuit diagram showing the masking processing circuit and the OCR processing circuit, 1st O-2ffi is a diagram showing the state of the signal output from the digital circuit according to the size of image data, and Fig. 10-6 is the U Ci (A diagram for explaining the process, FIG. 11 is a diagram for explaining the principle of multi-gradation processing,
FIG. 12-1 is a block circuit diagram showing a dither processing circuit;
FIG. 12-2 is a block circuit diagram showing a multi-value processing circuit;
FIG. 16 is a timing chart of signals in the circuits shown in FIGS. 12 to 1.12-2. In the figure, 100 is an image processing unit, 130 is a shading correction circuit, 140 is a γ correction circuit, 150 is a masking processing circuit, 160 is a UCR processing port j'ij,
170 is a dither processing circuit, 180 is a multi-value processing circuit, 1
90 is a synchronous control circuit, 200 is a CCD light receiving unit, 6
00 is a laser modified M, Il unit. Applicant Gyanon Co., Ltd. Oni; Father, Nori Punch 7mi 1st] One = Fi long sieve? -30 - Kanenagami? -2 Saekaka? -4 Figure A (0)
(f) B (0)
(P)0≦Island≦2 2
(071≦8 (5) ('?) 21XD 鴫吋2 (2) (3)
and 4 no (2) (3) (
4) g<On≦DD<Dn≦12 12<C)
F1≦184D<DnQ! 2 22<Dn≦
Al? 28<Dn≦2D branch<)≦38
8Dyt≦:jD MXDn≦9F 3-30-2, Shimomaruko, Ota-ku, Tokyo Canon Co., Ltd. Internal procedure amendment (method) % formula % ■ Display of the case 1982 Patent application No. 56516 2, Title of invention Image processing Device 3 Relationship with the person making the amendment Patent applicant Location 3 Shimomaruko, Ota-ku, Tokyo = 30-2 Name (+00) Canon Co., Ltd., Canon Co., Ltd. (telephone: 758-2111) 5, Date of amendment order Showa June 28, 1958 (shipment date) 6. Engraving of pages 1 to 3 and pages 71 to 76 of the specification subject to amendment 7 and the specification of contents of the amendment (no change in content) -(

Claims (5)

[Claims] (1) A dither processing means that binarizes the color image data by simultaneously comparing the color image data and a plurality of threshold matrices, and time-divides the binarized data within the transfer time of one pixel. 1. An image processing device comprising: a multi-value processing means for expressing multi-gradation. (2) If the image data is binarized by comparing the image data and the threshold matrix, and the binarized data is expressed in multiple gradations using a predetermined number of gradations, both the number of gradations An image processing device characterized in that the image processing device is made variable. (3) image data and a plurality of threshold values) processing means for binarizing the image data by comparing IJ lux;
An image processing apparatus comprising: selection means for selecting the number of threshold matrices. (4) An image processing device characterized in that the color image data is binarized (by comparing the color image data and a threshold value matrix), and the threshold values given by the threshold value matrix are differentiated for each color. . (5) An image processing apparatus characterized in that each color data is converted into multiple gradations by the same multi-value processing means.