JPH0363888A - Image processing device - 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] [Field of Industrial Application] The present invention relates to an image reading device that outputs a digital image signal by reading reflected light obtained by exposing an original image. This relates to a device that reads characters while separating them.

(Prior Art) Conventionally, this type of device has become widely used as a reader of a digital copying device or as an input device of an electronic file device.

In addition, with recent advances in electronic technology, a method has been developed that separates a read image into an edge area containing characters, etc. and a halftone area containing photographs, etc., and applies appropriate processing to each area. No, 4
25.1987. I) It is proposed in p100-101 etc.

[Problem to be solved by the invention]

However, with conventional technology, halftone areas represented by halftone dots in some manuscripts may be judged as edges, or very thin characters may be judged as halftone dots. However, there have been problems such as undesirable processing being applied to a part of the read image due to so-called "misjudgment", resulting in a significant loss of image quality.

This invention was made to solve the above problem, and by evaluating the density level in a read image using a judgment level that differs depending on the original mode, it converts and outputs an optimal image density signal for the original mode. It is an object of the present invention to provide an image reading device that can accurately separate a character image area with gradation image changes from read document image information and generate and output digital image information that is faithful to the document.

(Means for Solving the Problems) An image reading apparatus according to the present invention includes a mode setting means for setting and inputting a plurality of document reading modes corresponding to the type of document, and a density of peripheral pixels including a pixel of interest in an image signal. a conversion means for comparing the level with a judgment density level and converting it into a predetermined density level signal; a judgment level variable means for varying the judgment density level based on the document reading mode set by the mode setting means; and character area separation processing means for separating a character edge region from an image signal based on a predetermined density level signal output from the conversion means in accordance with a judgment density level variably set by the means.

Further, the determination level variable means is configured to variably set the determination density level so that the detection ratio of high-density level pixels to low-density pixels is given priority to low-density pixels.

Further, the converting means is configured to convert the density levels of surrounding pixels including the pixel of interest of the image signal into a predetermined density level signal based on a conversion lookup table.

Further, the conversion means is configured to convert the density levels of surrounding pixels including the pixel of interest of the color image signal into a predetermined density level signal based on a conversion lookup table.

(Operation) In the present invention, when the mode setting means sets an individual document reading mode corresponding to the type of document, the image processing unit starts processing the image signal output from the image sensor. At this time, the determination level variable means sets different determination density levels for the conversion means based on the document reading mode. In this way, the image signal is converted into a density level coefficient with different determination density levels set in the conversion means. At this time, the character area separation processing means makes it possible to faithfully separate the character area from the evaluated and output density level signal.

Further, the determination level variable means variably sets the determination density level so that the detection ratio of high density level pixels to low density pixels is given priority to low density pixels, and the determination density level when the density level is high is set to The density change is set to be more difficult to detect than the determination density level in the case of lightness, and it is possible to faithfully detect a character area in an intermediate image.

Furthermore, the conversion means converts the density levels of surrounding pixels including the pixel of interest of the image signal into a predetermined density level signal based on a conversion lookup table, and quickly converts the density level of the input image signal into a predetermined density level signal using the address as an address. to a corresponding concentration level signal.

Further, the conversion means converts the density levels of neighboring pixels including the pixel of interest of the color image signal into a predetermined density level signal based on a conversion lookup table, and converts the character area from the intermediate chroma image area in the color document. This enables faithful determination and separation.

(Embodiment) FIG. 1 is a sectional view showing an example of an image reading apparatus according to an embodiment of the present invention, and shows, for example, a full-color digital copying machine.

In the figure, an image scanner unit 201 reads a document and performs digital signal processing. A printer unit 202 prints out an image corresponding to the original image read by the image scanner unit 201 in full color on paper.

In the image scanner unit 201, 200 is a mirror pressure plate, 203 is an original table glass (hereinafter referred to as a platen), and an original 204 containing a simple binary image, a halftone image including a halftone image, a color image, etc. is placed. A lamp 205 exposes the original 204 placed on the platen 203. 206
208 is a scanning mirror, which images the light reflected from the original 204 through an imaging lens 209 onto a three-line sensor 210 composed of a charge-coupled device such as a COD. The 3-line sensor 210 has full color information red (R
), green (G), and blue (B) components to the signal processing unit 211. In addition,
The scanning mirrors 205, 206 have a velocity V, and the scanning mirrors 20
7, 208 scans the entire document by mechanically moving at v/2 in a direction perpendicular to the electrical scanning direction of the line sensor.

The signal processing unit (image processing unit) 211 electrically processes the read color analog image signal (image signal) and converts it into magenta (M), cyan (C), yellow (Y), and black (BK) components. and sends it to the printer section 202. Furthermore, for each document scan in the image scanner section 201, one component among M, C, Y, and BK is sent to the printer section 202, and one printout is completed by scanning the document four times in total. .

M, C, Y sent from the image scanner unit 201
, BK image signals are sent to the laser driver 212.

The laser driver 212 drives the semiconductor laser 2 according to the image signal.
13 is modulated and driven. The laser beam is a polygon mirror 214
．． A photosensitive drum 217 is scanned through an fθ lens 215 and a mirror 216.

Reference numeral 218 denotes a rotary developing device, which includes a magenta developing section 219, a cyan developing section 220 . It is composed of a yellow developing section 221 and a black developing section 222, and the four developing devices alternately contact the photosensitive drum 217 and develop the electrostatic latent image formed on the photosensitive drum 217 with toner (developing material).

A transfer drum 223 winds the paper fed from the paper cassette 224 or the paper cassette 225 around the transfer drum 223, and transfers the image developed on the photosensitive drum 217 onto the paper.

After the four-color images of magenta (M), cyan (C), yellow (Y), and black (BKI) are sequentially transferred in this manner, the paper passes through the fixing unit 226 and is discharged.

FIG. 2 is a plan view illustrating the configuration of the operation unit in the image reading apparatus according to the present invention, and 4201 is a numeric keypad, which is pressed to input a numerical value from 0 to 9 degrees for inputting the number of copies, etc.

4202 is a liquid crystal display unit that notifies the operator of the current system mode status and the like.

4203 is a reset key, which is a key for initializing the currently set mode.

4204 is a clear/stop key, which stops processing when the system is in operation, and when the system is not in operation.
Press to clear numerical values set using the numeric keypad, such as the number of sheets. 4205 is a copy key.

Reference numeral 4206 is a paper size selection key, and when the paper size selection key 4206 is pressed, the paper size, such as "A4", is displayed on the liquid crystal display section 4202.

Reference numeral 4207 denotes a density key, which is pressed to adjust the copy density from light to dark. When the density key 4207 is pressed, the density level indicator 4, which is formed of, for example, an LED, is activated.
The density levels set at 208 are displayed in stages.

Reference numeral 42o9 denotes a document type mode selection key constituting the mode setting means according to the present invention, which is pressed to select different document types such as text mode, photo mode, text/photo mode, map mode, etc. depending on the document type. The selected mode indicators 4210 to 4213 light up. Note that this corresponds to a state where the current map mode is selected.

Next, each of the above document type modes will be explained.

First, the text mode is a mode for clearly copying text originals, and corresponds to a mode that is optimal for text originals output by a word processor, for example.

Furthermore, the photo mode is a mode that emphasizes the color and gradation of an original to realistically reproduce a photo (including halftone dots), and corresponds to a mode that is most suitable for color photo originals.

Furthermore, the text/photo mode is a mode that separates the text and photos in originals that contain both text and photos (including halftone dots), making the text clear and reproducing the photos realistically.
For example, it corresponds to the optimal mode for documents such as newspapers that contain text and photographs.

The map mode is a mode that clearly expresses characters in halftone dots like a map, and is suitable for color printing or black and white characters, lines, symbols, etc.

For normal originals, if the text/photo mode is selected using the original type mode selection key 4209, the photo portions will be copied realistically and the text portions will be copied clearly.

However, if the above text/photo mode is selected for a document containing fine and intricate text, for example, the edges of the text will be difficult to capture and will be recognized as part of the photo (halftone dots), resulting in a sharp image. It may not be reproduced.

Furthermore, if there are sharp edges in a photographic document, the photographic image may be recognized as text, and the edges may be unnaturally emphasized, resulting in an unsightly appearance.

Furthermore, if the background is a solid image such as a map in which characters are drawn in halftone dots, it may be detected as a halftone image, making it difficult to copy the text clearly.

Therefore, in response to such a situation, the document type mode selection key 4209 is configured to allow the user to select text mode, photo mode, or map mode, thereby making it possible to form an optimal image. There is.

FIG. 3 is a flowchart illustrating an example of the document type mode selection processing procedure according to the present invention.

Note that (L) to (5) indicate each step.

When the power is turned on, first, it is determined whether or not the desired document type mode has been input using the document type mode selection key 4209 on the operation unit 1). If YES, the process proceeds to step (3) and thereafter, and the key Switch the specified document type mode (text, photo, 1 character/photo, map) in sequence (2
) Waits for the start key (copy key 4205) to be pressed (3).

Next, set the original type mode selection signal MOD0.1 corresponding to the set original type mode (4), output it to the signal processing section 211, and start the copying operation (5). 2 is a block diagram illustrating the configuration of the image scanner unit 201 shown in FIG. 1, where 101 is a counter that outputs a main scanning address 102 that specifies the main scanning position for line sensors 301 to 303 forming the 3-line sensor 210. . That is, the horizontal synchronization signal H
When 5YNC is "1", it is set to a predetermined value by a CPU (not shown) and incremented by a pixel clock signal CLK.

The image formed on the 3-line sensor 210 is photoelectrically converted by the three line sensors 301-303, and is sent to amplifiers 304-306. Sample hold circuit 307-30
9 and A/D converters 310 to 312 to output an 8-bit digital image signal 313 (corresponding to R), a digital image signal 314 (corresponding to G), and a digital image signal 31 for each color through A/D converters 310 to 312.
5 (corresponding to B).

Next, the configuration of the signal processing section (image processing unit) 211 shown in FIG. 1 will be described in detail with reference to FIGS. 5 and 6.

FIG. 5 is a block diagram illustrating the configuration of the signal processing section (image processing unit) 211 shown in FIG. 1, and the same components as in FIG. 1 are given the same reference numerals.

In the figure, CLK is a clock signal for transferring pixels, HSYNC is a horizontal synchronization signal, and is a synchronization signal for starting main scanning. CLK4 is a clock signal that generates a 400-line screen, which will be described later, and is output at the timing shown in FIG. Signal processing unit 211. It is sent to the printer section 202.

The image scanner unit 201 reads a document 204 and receives image signals (digital image signals) 313 to 31 as electrical signals.
5 is sent to the color signal processing unit 402 and feature extraction unit 403. The feature extraction unit 403 sends the color processing control signal generation unit 404 a black image analysis signal BL indicating that the currently processed pixel is a black image, and a tint analysis signal indicating that the image has a tint. COL, mixed analysis signal UNK indicating whether the currently processed pixel is a black image or a tinted image; UNK; the black image analysis signal BL;
An edge signal EDGE indicating that it is a CAN1 character edge is sent out.

The color processing control signal generation section 404 receives the above-mentioned signals from the feature extraction section 403 and generates various color processing control signals (multiplying coefficient signals GAiN1, GAiN2, filter switching signal F I L, etc.) to the color signal processing section 402 .

A density characteristic switching signal CAM) is generated. Note that the multiplication coefficient signal GAINI,2 is the digital image signal 313~
Weighting 315 is a coefficient for calculation (described later).

A 2-bit phase signal PHASE is sent from the control section 401 to each processing block. This phase signal PHASE corresponds to the developed color of the printer unit 202,
00 (magenta), 01 (cyan), io (yellow)
, 11 (black), respectively.

The color signal processing section 402 generates a recording image signal VIDEO for the printer section 202 based on the phase signal PHASE and the color processing control signal.

The printer unit 202 pulse-width modulates the laser emission time based on the recorded image signal VIDEO, and outputs a copy output 406 with gradation.

Here, a screen control signal SCR is input to the printer section 202 from a color processing control signal generation section 404 . The printer unit 202 receives this screen control signal SCR.
switches between multiple pulse width modulation basic clocks (screen blocks) to express the optimum density for the original.

Note that in this embodiment, the screen control signal SCR is
'', pulse width modulation is performed in units of one pixel, and when the SCR signal is 1, pulse width modulation is performed in units of two pixels.

FIG. 7 is a detailed block diagram of main parts of the signal processing section (image processing unit) 211 shown in FIG. 5, and the same parts as in FIGS. 1 and 5 are given the same reference numerals.

In the figure, 103 is a light amount signal-density signal converter;
~255 range image signal (digital image signal) 313~3
15 to the calculation process based on the conversion formula described later.
Converts to 5-range print signals C, M, and Y signals. 1
04 is a black extraction unit that determines a block signal BK from the minimum value of the C, M, and Y signals, and then sends it to the subsequent masking processing unit 10.
In step 5, arithmetic processing is performed to remove color turbidity of the developing material, and the developed color signal Vl selected by the inputted face signal PHASE is output to the line delay memory 112. The line delay memories 112 and 113 store print signals C, M,
It functions to delay Y and BK by 3 lines and 4 clocks for character edge determination processing. 114 and 115 are multipliers, the multiplier 114 multiplies the multiplication coefficient signal GA I N 1 and the color recording signal 2 (details will be described later) and outputs a multiplication output v3, and the multiplier 115 multiplies the multiplication coefficient signal GAI
N2 is multiplied by the density signal M2 (details will be described later), and a multiplication output M3 is output to the adder 116. Adder 116
Now, the above multiplication output M3. An image signal V4 is generated by performing addition processing with the multiplication output V3, and an edge emphasis filter is configured with a 3×3 pixel Laplacian filter in the filter circuit 117 at the subsequent stage, and the Laplacian multiplier is set to 1/2 and 1/2. After switching the type and performing smoothing processing, for example, the gamma conversion unit 118 converts the 8-bit image signal v5 by referring to a conversion table in the brain and converts the image signal VIDEO to the printer unit 202.
The signal is output to the PWM modulation section 119 of.

On the other hand, in the feature extraction unit 403, reference numeral 106 is a color determination unit, which extracts from the color analog image signal (image signal) a black image analysis signal BL, a color analysis signal COL indicating that the image has a color tone, and a current color analysis signal COL. Whether the processed pixel is a black image or a tinted image, a mixed analysis signal UNK and a cancellation signal CAN for canceling the black image analysis signal BL are output to the delay memory 120.

Reference numeral 107 denotes a character edge determination unit, which detects from the green component of the color analog image signal the presence or absence of a steep density change and whether sharp density change points are continuous in a specific direction through arithmetic processing described later, and determines an edge area. At the same time, the halftone dot area is determined.

That is, when an individual original 8% reading mode corresponding to the type of document is set by the document type mode selection key 4209 serving as a mode setting means, the image processing unit 211 outputs an output from the image sensor (3 line sensor 210). Processing of the analog image signal, the image signal 314, is started.

At this time, the CPU, which will be described later, also serves as the judgment level variable means.
1871 outputs a setting switching signal MD for setting a different determination density level to a conversion means (in this embodiment, a density conversion ROM to be described later) based on the document reading mode. In this way, the analog image signal, in this embodiment the image signal 314 (G signal), is converted into a density level signal with different determination density levels set in the density conversion ROM (conversion means). At this time, the feature extraction unit 403 forming the character area separation processor can faithfully separate the character area from the evaluated and output density level signal.

Further, the CPU variably sets the determination density level for the conversion means so that the detection ratio of high-density level pixels to low-density pixels is given priority to low-density pixels. The density change is set to be more difficult to detect than the judgment density level when the level is light, and the character area in the intermediate image is faithfully detected.

Furthermore, the density conversion ROM (conversion means) converts the density levels of surrounding pixels including the pixel of interest in the analog image signal into a predetermined density level signal based on a conversion lookup table, and converts the density level of the input image signal into a predetermined density level signal based on a conversion lookup table. is converted into a predetermined corresponding density level signal at high speed as an address.

Further, the density conversion ROM (conversion means) converts the density levels of surrounding pixels including the pixel of interest of the color analog image signal into a predetermined density level signal based on a conversion lookup table (described later), and This makes it possible to faithfully determine and separate a character region from an intermediate chroma image region.

Furthermore, when an individual document reading mode corresponding to the type of document is set using the document type mode selection key 4209,
The image processing unit 202 processes the color analog image signals (image signals 313 to 31
5) starts the process. At this time, the color determination section 10 constituting the third detection means detects the achromatic color portion while analyzing based on the third determination condition, and based on this detection result, the second character area separation processing means (Edge determination unit to be described later) Achromatic character areas are determined from the color analog image signal by the above-mentioned first to
3. Based on the detection result of the detection means of step 3, the character edge area is separated from the color analog image signal.
It is possible to faithfully separate character edges.

The color processing control signal generation section 404 also includes the filter control signal generation section 109. Gamma switching signal generation section 110. It consists of a screen switching signal generation section 111, and a filter control signal generation section 109 generates the color analog image signal (
image signal), a black image analysis signal BL, a color analysis signal COL indicating that the image has a tint, and whether the currently processed pixel is a black image or an image with a tint. For example, a 2-bit filter switching signal FIL is stored in the delay memory 12 from a possible intermediate saturation signal UNK, a cancellation signal CAM for canceling the black image analysis signal BL, etc.
1 to the spatial filter section (filter circuit 117)
Outputs a bit delay filter switching signal DFIL.

Further, the gamma switching signal generation unit 110 generates a black image analysis signal BL, a tint analysis signal COL indicating that the image has tint, from the color analog image signal (image signal),
An intermediate saturation signal UNK, in which case the current processing pixel is a black image or a tinted image;
Cancellation signal CA that cancels the above black image analysis signal BL
A selection control signal GAM for selecting, for example, four types of conversion tables is output from the edge signal EDGE and the like to the gamma conversion section 118 as a delay selection control signal DGAM via the delay memory 121.

Further, the screen switching signal generating section 111 generates a black image analysis signal BL from the color analog image signal (picture signal).
, a color analysis signal CO indicating that the image has color.
L, whether the currently processed pixel is a black image or a tinted image, an intermediate saturation signal UN that may be in either case;
K. From the cancel signal CAN that cancels the black image analysis signal BL, the edge signal EDGE, etc., the screen control signal SCR is outputted as a delayed screen control signal DSCR to the PWM modulation section 119 of the printer section 202 via the delay memory 121, and is processed as described below. the pulse width modulation signal pw or pulse width modulation signal PW4 to be selected.

FIG. 8 is a block diagram illustrating the configuration of the color determining section 106 shown in FIG. 7, and the same components as in FIG. 7 are given the same reference numerals.

In this figure, 1101 is a pixel color determination unit, which generates a black pixel signal BLP indicating that the pixel is black, a tint pixel signal C0LP indicating that the pixel is black, and an intermediate saturation signal UNKP indicating that the pixel is an unknown pixel. , is sent to the area processing unit 1102. The area processing unit 1102 generates a black pixel signal BLP, a tint pixel signal C0, within a 5×5 area, for example.
LP.

The area of the intermediate saturation signal UNKP is determined to remove errors, and the true black image analysis signal BL, the tint analysis signal COL indicating that the image has tint, and whether the currently processed pixel is a black image or not are determined. An intermediate saturation signal UNK and a cancellation signal CAN are generated, which are likely to be images with flavor or in both cases.

Note that the character edge determination unit 107 uses the wavelength/
As is clear from the relative sensitivity characteristics, character edges in white/black images are detected using a brane component that is close to the viewing sensitivity characteristics, and the edge signal EDGE, which is a character edge detection signal, is generated.
is output to the color processing control signal generation section 404.

Next, details of the color determination section 106 will be explained with reference to FIG. 10.

FIG. 10 is a circuit block diagram illustrating the detailed configuration of the color determination section 106 shown in FIG. 7.

In the figure, 1301 is a MAX/MIN detector which executes maximum value/minimum value detection IA processing from the image signals 313 to 315 to generate determination signals soo and s.

1. Selector 13 for SO2, SIO, Sit, S12, etc.
Output as shown in 02-1309. 1310
〜1315 is a subtracter, and A- for input A and input B.
Output B. 1316 to 1323 are comparators,
Comparators 1316 and 131 for input A and input B
9 is active (H level) when 2A>B holds.
, that is, outputs "1", and otherwise outputs "o".

Also, for input A and input B, comparator 1317.
1320, 1322, and 1323 are active (H level), that is, output "1" when A>B holds, and otherwise output "O".

Additionally, comparators 1318 .
1321 is active (H
output level), i.e. rlj, otherwise output "O"
Output.

1329 is a NOR gate, AND gate 1325.1
330 and the AND gate 13
Output to 27. AND gate 1327 generates and outputs the above-mentioned intermediate saturation signal UNK.

1326.1328 is an AND gate, AND gate 1
326 generates and outputs a black image analysis signal BL, and an AND gate 1328 generates a tint analysis signal c.

Generate and output L.

FIG. 11 shows the MAX/MIN detector 1 shown in FIG.
301 is a circuit block diagram illustrating the detailed configuration of 1.
350 to 1352 are comparators, each of which receives the R component image signal 313. G component image signal 314. When the density level of the B component image signal 315 is RAG, and the relationships of G>B and B>R hold, "1" is output. ■1 to I3 are inverters, ANDI to AND7 are AND gates, OR
1, OR2 is an OR gate, and the above judgment signal S00,
SOl, SO2, SIO, Sll, S12, etc. are output as shown.

FIG. 12 is a correlation diagram illustrating the output conditions of the input image signal and each determination signal shown in FIG. 11.

As can be seen from these figures, when the density level of the R component image signal 313 is MAX or when the R component image signal 313
．． G component image signal 314. If the density levels of the B component image signals 315 are all equal, the determination signal SOO is "1".
”, the determination signals Sol and SO2 become “0”, and B
When the density level of the component image signal 315 is MAX,
The judgment signal SO2 becomes rl, and the judgment signal soo, s○
1 becomes rQJ. Further, when the density level of the R component image signal 313 is MIN, or when the R component image signal 313. G
Component image signal 314. When the density levels of the B component image signals 315 are all equal, the determination signal 310 is rIJ,
Sit, S12 becomes rO. Further, when the density level of the G component image signal 314 is MIN, the determination signal Sl
l becomes "1" and determination signals SIO and S12 become "0".

Further, when the density level of the B component image signal 315 is MIH, the determination signal S12 becomes "1", and the determination signal SI
O, Sll becomes "0".

Therefore, for example, the density level of the R component image signal 313 is M
In the case of AX, since the relationships R>G and R≧B hold true, the comparator 1350 outputs "1" and the comparator 1352 outputs r□. And gate A
NDI outputs "1" and OR gate OR1 outputs "1". At this time, the AND gate AND2. AND3 outputs "0". That is, the determination signal SOO is "1"
Therefore, the determination signal So1. SO2 becomes “O” (first
(See Figure 2).

On the other hand, the judgment signal SOO of the MAX/MIN detector 1301
, Sol, SO2 are input to the selector 1302, and the determination signals SIO, Sit, S12 are input to the selectors 1303 to 13.
09 is entered.

The selectors 1302 to 1309 are composed of AND gates ANDII to AND13 and OR gates OR11, etc., as shown in FIG.
According to the state of 2, as shown in FIG. 14, one of the inputs A to C is selectively output.

Therefore, when the determination signal SO is "1" for inputs A-C and the determination signals Sl and S2 are "0", input A is selected and output, and the determination signals So and Sl are rQJ, and the determination When the signal S2 is "1", the input card is selectively output, and in this embodiment, each R component image signal 313 . G component image signal 314. The B component image signal 315 input corresponds to this.

[Pixel color determination processing]

The R component image signal 313. G component image signal 314, B
The maximum density value of the component image signal 315 is MA
Let X be the value of the minimum concentration level, and further,
This is done by dividing the area as shown in FIG.

FIG. 15 is a schematic diagram illustrating the pixel color determination classification of the image reading device according to the present invention, in which the horizontal axis represents the maximum density level and the vertical axis represents the minimum density level.

In other words, in the achromatic color area, the difference between MAX and MIN is small, and the closer it is to a chromatic color, the smaller the difference between MAX and M
Taking advantage of the fact that the difference in IN increases, MAX, MIN
By linear simultaneous inequalities with parameters, MAX-
Divide the MIN plane. Specifically, the constants ka, kb
, kc, ia.

i b, i c, WMX, and WMN are divided into predetermined areas A1 to D1.

In FIG. 15, area A1 is a dark achromatic (black) area, and in order for (MAX, MIN) to be included in this area, MIN≦WMN or MAX≦WMX, and the following (1st ) holds true.

In addition, section B1 is an area between dark achromatic colors and chromatic colors, and in order for (MAX, MIN) to be included in this area, MIN≦WMN or MAX≦WMX, and the following (2nd This is a case where either one of the equations ( ) and the equation (3) all hold true.

Furthermore, the section C1 is a chromatic color area, and (MAX, MIN
) is included in this region if MIN≦WMN or MAX≦WMX and the following formula (4) holds true.

Moreover, the division Dl is a bright achromatic (white) area, and (
In order for MAX, MIN) to be included in this region, one of the following equations (5) holds true.

Figure 16 shows the area classification and MAX/M shown in Figure 15.
15 is a correlation diagram illustrating the relationship with the output signal of the IN detector 1301, and the same parts as in FIG. 15 are given the same reference numerals.

As can be seen from this figure, in order for each pixel to be included in the classification A1 by the maximum/minimum processing of the image signals 313 to 315, the black image analysis signal BL must be "1", the color analysis signal COL, the intermediate This is the case where both the chroma signals UNK and "OJ" are present.

In addition, by maximum/minimum processing of image signals 313 to 315,
In order for each pixel to be included in the section B1, the intermediate chroma signal U
When NK is "1", black image analysis signal BL and color analysis signal C
This is the case where both OLs are "0".

Furthermore, by maximum/minimum processing of image signals 313 to 315,
In order for each pixel to be included in the classification C1, the color analysis signal C
When OL is "1", black image analysis signal BL, intermediate saturation signal U
This is the case where both NK are "O".

In addition, by maximum/minimum processing of image signals 313 to 315,
In order for each pixel to be included in the division Dl, the black image analysis signal BL must be "1", the intermediate saturation signal UNK, and the tint analysis signal C
This is the case where both OLs are "0".

The above-mentioned area determination is performed by the circuit shown in FIG.
The selectors 1304 to 1309 are also selected from constants ka, kb, and k in conjunction with the selector 1303.
c.

Select values for ia, ib, and ic. For example, MAX is an R signal (image signal 313), MIN is a G signal (image signal 314)
In this case, selector 1304 selects KAG and selector 1
306 selects KCG, selector 1307 selects iAG, selector 1308 selects iBG, selector 1309 selects iCG, and constants ka, kb, kc, ia, ib, respectively.
The reason for changing the value of ic is as follows.

Generally, in the case of a full-color sensor, there is a color balance unique to the sensor, so if all colors are judged as chromatic/achromatic based on the same criteria, it may cause erroneous judgment. Therefore, as shown in FIG.
-H's three-dimensional space is divided into three parts to accommodate the color balance characteristics of the sensor.

FIG. 17 shows the 3-line sensors 301 to 3 shown in FIG.
FIG. 3 is a spatial characteristic diagram showing the color balance characteristics of No. 03.

As you can see from this figure, the R-G-B three-dimensional space is M
It is divided into a region 5702 where IN is R, a region 5703 where MIN is G, and a region 5704 where MIN is B, and the vertical constants ka, kb, kc, ia, ib,
Use the value of ic.

For example, for a sensor with a low R component signal, the constants KAR, KBR, KCR, iAR, and
By setting the values of iBR and icR a little larger, when MIN is R, it is possible to narrow the classification A1 and the classification C1 in the region shown in Fig. 15, allowing fine support for various sensors C. can do.

Subtractors 1310-1312 and comparators 1316-1
318 determines the magnitude relationship between MAX-ka and 2MIN MAX-kb and MIN MAX-kc and 1/2MIN.

In addition, subtracters 1313 to 1315 and a comparator 131
9 to 1321 determine the magnitude relationship between MAX-ia and 2MIN MAX-ib and MIN MAX-ic and 1/2MIN.

Further, the comparator 1322 and the comparator 1323 respectively determine the magnitude relationship between MAX and WMX and MIN and WMN.

The above region determination is performed as described above, and the results are output as determination signals of the intermediate saturation signal UNK, the tint analysis signal COL, and the black image analysis signal BL.

FIG. 18 is a circuit block diagram illustrating the configuration of area processing section 1102 shown in FIG. 8, and the configuration and operation will be explained below.

Black pixel signal B determined by pixel color determination unit 1101
Line memory 17 stores LP, tint pixel signal C0LP, and intermediate saturation signal UNKP indicating that the pixel is an unknown pixel.
01 to 1704, synchronized by the horizontal synchronizing signal HSYNC and clock signal CLK, and five lines are output simultaneously. Here, black pixel signal BLP1 tint pixel signal C0LP, intermediate saturation signal LINK
P delayed by one line are BLP2 and C0, respectively.
LP2. UNKP2, which is further delayed by one line, is BLP3. C0LP3. UNKP3, which is further delayed by one line, is BLP4°C0LP4. UNKP4, which is further delayed by one line, is BLP5. C0LP5.
When setting UNKP5, the counting means 1705 delays each signal by 5 pixels to obtain a count value NB, which is the number of BL black pixels within a 5×5 area (see FIG. The count value NL is obtained by counting the number of chromatic pixels. Furthermore, a comparator 7007 compares the number of black pixels (count value NB) and the number of chromatic pixels (count value NL) within the 5×5 block.

Furthermore, through gate circuits 1708 to 1715, 5×5
BLP3. which is the output of the pixel color determination unit 1101 for the center pixel of the area. A black pixel signal BL, which is calculated together with the result of C0LP3゜UNKP3 and indicates that the center pixel is black, a tint pixel signal COL that indicates that the center pixel is a chromatic color, and a color tone pixel signal COL that indicates that the center pixel has intermediate saturation. An intermediate chroma signal UNK indicating 1711, 1712, .
1715.

At this time, the determination criterion does not overturn the determination for those whose determination result according to the first determination criterion is a black pixel or a chromatic pixel. In other words, both BLP3°C0LP3 are “1”.
”, the black pixel signal BL is “1” or the tint pixel signal COL is “IJ”. Also, the judgment result of the first judgment criterion is between the chromatic pixel and the achromatic pixel. For those, a comparator 1716 determines whether the number of black pixels is greater than or equal to a predetermined value (NBC), and a comparator 1717 determines whether or not the number of chromatic pixels is greater than or equal to a predetermined value. At con 1 < rate 17o7,
It is determined whether the number of black pixels or the number of chromatic pixels is greater.

Then, if the number of black pixels is greater than a predetermined value and the relationship NB>NC is satisfied, the UNKP is outputted at the gate 1708.
3 is output as the black pixel signal BL.

Further, the number of chromatic pixels is greater than or equal to a predetermined value, and NB≦NC
In this case, the gate 1709 outputs UNK3 as the tint pixel signal COL.

This is to eliminate color blurring at points where the color of the document changes due to scanning unevenness of the scanning mirrors 206 to 208 shown in FIG. 1 constituting the scanning optical system and magnification error of the imaging lens 209.

Gates 1713 to 17 determine that there are no more than a predetermined number of black pixels or chromatic pixels in the surrounding pixels of UNKP3.
Intermediate chroma signal UNK which is detected by 15 and becomes an intermediate chroma signal
Output.

Next, with reference to FIG. 20, the color determination unit 1 shown in FIG.
The output processing of the cancel signal CAN sent from 06 will be explained.

FIG. 20 is a cancel signal generation circuit diagram of the color determination section 106 shown in FIG. 7, and the configuration and operation will be described below.

In the circuit shown in FIG. 18, if the pixel of interest is a black pixel, a black image analysis signal BL is output regardless of surrounding pixels. However, if there is the aforementioned scanning speed unevenness or imaging magnification error, a black signal may occur around the color signal due to color fringing, as shown in FIG. 21. Since the black signal due to this color fringing occurs around the color signal, the light amount value becomes larger than the color signal, as shown in FIG.

FIG. 21 is a schematic diagram illustrating the state of occurrence of color fringing, and FIG. 22 is a characteristic diagram illustrating the light amount characteristics at a specific position shown in FIG. 21, where the vertical axis shows the light amount and the horizontal axis indicates a specific location.

As mentioned above, when the black signal due to color fringing occurs around the color signal, the light amount value becomes larger than the color signal as shown in Fig. 22, so there is a color tone around the pixel of interest that has a smaller light amount value than the pixel of interest. It detects whether the pixel signal COL is present and generates a cancel signal CAN.

Here, as the light amount signal, an image signal 314 which has the closest to the visibility characteristic shown in FIG. 12 is used, and this image signal 314
14 is delayed by one line of FIFO memories 1718 to 1720, and the line signal G3 of interest and the line signals G2 . G4 is input to the calculation unit 1722. At the same time, the tint pixel signals C0LP2 to C0LP4 outputted from the circuit shown in FIG. 18 are input, and are arithmetic processed by the circuit shown in FIG. 23.

FIG. 23 is a circuit block diagram illustrating the configuration of the arithmetic unit 1722 shown in FIG. Or delay by 3 pixels. Note that the pixel signal G32 and the pixel signal C0L3
2 is the pixel of interest. Pixel signal G32 is comparator 1
Peripheral pixel signals G22.737-1740. G31
, G33. G42, etc., and the output of this comparator becomes H level when the light amount value of the peripheral pixel is lower than that of the pixel of interest. Then, the AND gates 1741 to 1744 perform an AND with the color determination signal of the surrounding pixels, and the OR gate 174
At step 5, a cancel signal CAN is output.

(Character Edge Determination Process) Next, the configuration and operation of the character edge determination section 107 shown in FIG. 7 will be described with reference to FIG. 24.

FIG. 24 is a block diagram illustrating the configuration of the character edge determination unit 107 shown in FIG.
The density change processing unit 1801a performs density change detection processing on the input image signal 312 (G signal) using a logic circuit, which will be described later, to obtain density change information AK.
I to AK8 are output to the concentration change analysis section 1802. The density change analysis unit 1802 analyzes and captures the continuity of density changes and density changes in specific directions from the density change information AKI to AK8, and outputs the edge signal EDGEO to the edge determination unit 18.
Outputs 022.

Further, the point processing unit 1801b of the density change analysis unit 1802 determines the presence or absence of a halftone dot pixel from the density change information AKI to AK8, and sends the halftone pixel detection signal DOTO to the halftone signal area processing unit 1802.
Output to 8021. The halftone signal area processing unit 18021 receives the mode signal MODO, which is output from the CPU 1871.
A halftone dot area signal DOTI is output to an edge determination section 18022 based on MODI.

FIG. 25 is a correlation diagram illustrating the corresponding mode types of the mode signals MODO and MODI output from the CPU 1871 shown in FIG.
I(7) settings are roIJ, rlOJ, rti, roOJ
``Text mode'', ``Photo mode'', ``Text/Photo mode'', and ``Map mode'' are respectively supported.

Next, the principle of character edge determination processing will be explained with reference to FIGS. 26 and 27.

FIG. 26 is a principle explanatory diagram illustrating the principle of character edge determination processing according to the present invention.

A document 1901 is an image having shading, and includes a character edge area 1902 and a halftone area 19o expressed by halftone dots.
Contains 3. Pixel blocks 1905 to 1912 are composed of 10 pixels including 9 neighboring pixels of the pixel of interest XIJ, as shown in FIG. 27, for example.

First, the density change processing unit 18028 determines whether or not there is a steep density change using nine neighboring pixels surrounding the pixel of interest Xlj of the pixel block 1904 shown in FIG. 27 as one unit. Determine whether or not concentration change points exist continuously in a specific direction.

Specifically, as shown in equation (6) below, the company's featured pixel XI
Neighboring pixel X for J I-1+J-I +neighboring pixel X
l-1. j+neighborhood pixel X'l-1+J+l+neighborhood pixel XI+,-1,neighborhood pixel XI,, 1+1. Neighboring pixel XI+I+J-1+neighboring pixel x1+1+J+neighboring pixel X
The difference values J1 to J8 of I+l+J are taken, and it is determined from the magnitude relationship whether or not there is a steep concentration change.

Specifically, in the vertical edge detection process where there is a high density on the right side as shown in the pixel block 1905 shown in FIG. Concentration change information AKI (second
The shaded pixels of pixel blocks 2101 and 2102 shown in FIG.

In addition, in the horizontal edge detection process where the lower right side has a high density as shown in the pixel block 1906 shown in FIG. Continuous concentration change information AK2 (second
8) are output.

Furthermore, in the edge detection process in the diagonal right direction where the density is high in the lower right direction, as shown in pixel block 1907 shown in FIG. Continuous density change information AK3 (hatched pixels in pixel plots 2105 and 2106 shown in FIG. 28) is output.

In addition, in the edge detection process in the diagonal left direction where there is a high density at the lower left as shown in the pixel block 190B shown in FIG. Continuous concentration change information AK4 (
The shaded pixels of pixel blocks 2107 and 2108 shown in FIG. 28 are output.

Furthermore, as shown in pixel block 1909 shown in FIG. 26, in the vertical edge detection process where the high density is on the left side, the points where the difference value J5 in equation (6) is large are continuous in the vertical direction. Density change information AK5 (shaded pixels in pixel blocks 2109 and 2110 shown in FIG. 28) is output.

In addition, in the horizontal edge detection process where there is a high density on the upper side as shown in the pixel block 1910 shown in FIG. The density change information AK6 (hatched pixels in pixel blocks 2111 and 2112 shown in FIG. 28) is output.

Furthermore, in the edge detection process in the diagonal right direction where there is a high density on the upper left side as shown in the pixel block 1911 shown in FIG. 26, the point where the difference value J7 in equation (6) is large is Continuous concentration change information AK7 (
The shaded pixels of pixel blocks 2113 and 2114 shown in FIG. 28 are output.

Furthermore, in the edge detection process in the left diagonal direction with high density in the upper right direction as shown in pixel block 1912 shown in FIG. 26, the point where the difference value J8 in equation (6) is large is Continuous concentration change information AK8
(hatched pixels of pixel blocks 2115 and 2116 shown in FIG. 28) are output.

On the other hand, pixel block 190 in halftone area 19o3
9 to 1912 as well, the above-mentioned difference values J1 to J8 become large, but the continuity in the specific direction is low. Also,
The halftone area (halftone area) 1903 exhibits a characteristic pattern due to a specific combination of density changes in a specific direction, and is therefore detected as being distinguished from edges.

FIG. 29 is a circuit block diagram illustrating the detailed configuration of the character edge determining section 107 shown in FIG. 7, and the same components as in FIG. 24 are given the same reference numerals.

In the figure, line memories 1803 and 1804 sequentially delay the image signal 314 (G signal) in synchronization with the horizontal synchronization signal H3YNC. 1805 is a delay circuit, for example, the third
As shown in Figure 0, it is composed of flip-flops, comparators, registers, etc.
The density change information AKI to AK8 defined by the above is output from the above difference processing.

1806 to 1809 are line memories, and detector 1805
The density change information AKI to AK8 outputted from the horizontal synchronization signal (ISYNC) is delayed in synchronization with the horizontal synchronization signal (ISYNC), and the density change information AKI to AK8° BKI to BK8. CKI~
CK8. DKI~DK8. AKI to AK8 are 1 to 1 stages of flip-flops 1881a to 188 corresponding to each line.
1e flip-flops 1882a-1882e, flip-flops 1883a-1883e, flip-flops 1884a-1884e, flip-flops 1885
a to 1885e, the pixel distribution direction characteristics in each direction are checked, and the flip-flops 1881a to 1885e check the pixel distribution direction characteristics in each direction.
From 81e, directional characteristic factor data f A U R1,
i A BL 1 , i A UP2, 1ABT2
,1AUP3,1ABT3,1ALF3,1ART3,
1AUP4, 1ABT4.1AUL5, 1BR5 are outputted to the subsequent Nantes gates 1809a to 1824a or the halftone dot processing section 1802b shown in FIG. 33 as shown. In addition, flip-flops 1882a to 1882e
From pixel 1BLFl, tB, which becomes directional characteristic factor data
RTl, 1BUR2, 1BBL2, 1BUP3, 1BB
T3,1BLF3,1BRT3.1BUL4,1CBT
4, 1BLF5, 1BRT5 is Nantes Gate 18 in the latter stage
09-1824 or halftone processing section 180 shown in FIG.
2b as shown in the figure.

Furthermore, from flip-flops 1883a to 18838, pixels 1CUPI and 1cBT1, which serve as directional characteristic factor data, are
, 1cLF1, 1cRT1, 1CUP2, 1CBT2,
1cLF2, 1CRT2.1cUP3, 1CBT3. t
cLF3, 1cRT3, 1CUL3, 1cBR3, 1C
UR3゜1CBL3,1cUP4,1CBT4,1CL
F4.1CRT4, 1cUP5, 1CBT5.1CLF
5, 1CRT5 is the rear Nantes gate 1809a-18
24a or the halftone dot processing section 1802b shown in FIG. 33 as shown.

Furthermore, from the flip-flops 1884a to 1884e, pixels 1DLFl and 1DRT1, which serve as directional characteristic factor data.
,1DUL2,1DBR2,1DUP3,1DBT3,
1DFL3. 1DRT3.1DUR4, 1DBL4,
1DLF5, 1DRT5 is Nantes gate 1809 in the latter stage
a to 1824a or halftone processing unit 180 shown in FIG.
2b as shown in the figure.

Furthermore, from flip-flops 1885a to 1885e, pixels 1EUL1 and 1EBR1, which serve as directional characteristic factor data, are
,1EUP2,1EBT2,1EUP3,1EBT3,
1ELF3, 1ERT3.1EUP4,1EBT4,1
EUR51EBL5 is the latter stage Nantes Gate 1809a~
1824a or halftone processing section 1802b shown in FIG.
is output as shown in the figure.

1825 is Noah Gate, Nantes Gate 18098-1
When the center pixel constitutes a continuous edge, the edge detection signal EDGEO is output to the edge determination unit 18022.

(Bottom right) (Of J3 ≦ 3 8, in the above formula, T and ~T3 indicate preset slice levels, and each density change information AKI ~ AK8
``rl'' corresponds to the case where there is a steep concentration change,
The case where each of the density change information AKI to AK8 is "O" corresponds to the case where there is no steep density increase. In addition, the above T, ~T3
The value is variably set depending on the document mode set on the operation unit.

Specifically, the concentration change point detection unit 1801 detects the concentration change point as shown in FIG.
a) It is configured as shown in e.

FIG. 30(a) is a circuit block diagram illustrating the detailed configuration of the detector 1805 shown in FIG. 29α, and the same components as in FIG. 19 are designated by the same reference numerals.

The line memories 1803 and 1804 output the delayed image signal 314 to the density conversion ROMs 2023a to 2025a. The density conversion ROMs 2023a to 2025a are
Mode signal MODO,M output from CPU1871
ODi stores any judgment density level data (lookup table) selected by the switching control signal MD output from the NOR gate 2026.
Any judgment density level data addressed by the bit image signal 314 is sent to the flip-flops 2001 to 20.
06.

Flip-flops 2001 to 2o06 are neighboring pixels xi-l for the pixel of interest XIJ, which has undergone density conversion.

J"'I+neighboring pixel X I-1+J+neighboring pixel xi-
1. J''' Neighboring pixel XI, j-1, Neighboring pixel X l+
J and 1. Neighboring pixel x141 +J-1+neighboring pixel Xl+
Pixel signal for I+J+neighboring pixel Xl*1゜Jul (
concentration level signal) to the subsequent difference calculators 2007 to 20.
Output to 14.

Difference value J1 from these difference calculators 2007 to 2014
-J8 are led to comparators 2015-2022, and each density change information AK1-AK8 is output. 2023～
A register 2025 holds slice levels lower 1 to T output from the cpu 1871 and the like. By variably setting this slice level T I"" T3, the determination conditions can be varied, and the degree to which character edges are captured can be set in various ways.

On the other hand, whether or not the steep density change continues in a direction having an angle of, for example, 90'' with respect to the direction of the density change can be determined by using the density change information AKI to AK8 further downstream. At 1802, continuity determination processing in a specific direction is performed assuming, for example, 5×5 pixel blocks 2201 to 2208 shown in FIG.

FIG. 30(b) shows the density conversion R shown in FIG. 30(a).
It is a schematic diagram explaining the conversion table data of OM2023a-2025a, the horizontal axis shows the input address, and the vertical axis shows the output data.

As can be seen from this figure, the switching control signal MD is "1"
In the case (input A8 is "1"), when the output is y and the input is X, the conversion characteristic (through mode) where y=x is selected, and the input density level and output density level are the same. .

On the other hand, when the switching control signal MD is "o" (input A8 is "0") and the input is X, y = (x/255)
The conversion characteristic ■, which is 2×255, is selected.

This means that the characteristic relationship of a quadratic curve holds true, and the total differential of this y is Δy = 2 X (X / 255) ・Δ
X, the amplitude disappears in dark (high density) parts, and the amplitude is amplified in bright (light density) parts.

Specifically, the determination density level is variably set so that the detection ratio of high-density level pixels to low-density pixels gives priority to low-density pixels.

Furthermore, the switching control signal MD is connected to the mode signals MODO, M
Since it is configured so that r l and J are set only when ODI is both r With the characteristics corresponding to the following curve, the density level is converted and the subsequent flip-flops 2001 to 2006
Outputs the density pixel signal to. This makes it easier to detect changes in the density of characters in a bright background, and makes it difficult to detect changes in density in halftone dots, except when in the map mode.

On the other hand, when the map mode is set, characters in halftone dots can be detected faithfully.

Figures 30(C) and (d) show the detector 1 shown in Figure 29.
19A and 19B are a circuit blank diagram and a main part circuit diagram illustrating other detailed configurations of the 805, in which the same reference numerals as in FIG. 19 are given.

In the figure, 4401 to 4403 are concentration converters, each with 4
401 to 4403 are AND gate AD and selector SEL
The selector SEL selects input A or input B based on the switching control signal MD.

Specifically, when the switching control signal MD is "o",
If the input image signal 314 is converted (through mode) based on the conversion characteristic shown in FIG. The AND output (7 bits) outputs a density pixel signal to the subsequent flip-flops 2001 to 2006 as a conversion signal. That is, the range below 128 is in the undesirable range for the concentration. As a result, as in the above-mentioned embodiment, the amplitude of the high-density portion is suppressed, and as a result, it is possible to prevent erroneous detection of characters in halftone dots in modes other than the map mode.

FIG. 30(f) is a circuit block diagram illustrating still another detailed configuration of the detector 1805 shown in FIG.
Components that are the same as those in FIG. 19 are given the same reference numerals.

Reference numerals 4501 to 45o3 are look-up tables, and the comparator 20 uses the upper 3 bits of the center pixel signal Xll latched in the flip-flops 2001 to 2006 and the lower slice levels 1 to T output from the CPU 1871.
By varying the comparison values TTl'' to TTS set to 15 to 2022, it is possible to evaluate the density level using different criteria as described above.

For example, comparison values TTI~TT3 and slice levels lower 1~T
The relationship with 3 is as follows.

However, α corresponds to the upper 3 bits of the pixel signal X, and α
takes a value from 0 to 7.

FIG. 31 shows the density change processing section 1802 shown in FIG.
FIG. 3 is a schematic diagram illustrating a specific pixel pattern detected by a, and 2201 to 2208 are pixel blocks, which are composed of 5×5 perpendicular blocks centered on the pixel of interest.

In this figure, pixel blocks 2201.2201 are
Indicates the reference pixel when detecting a succession of edges in the vertical direction, and the characteristics of the density change of the surrounding pixels described above are the density change information A.
This corresponds to the case where three consecutive pixels of KI and AK5 are detected.

Similarly, the pixel blocks 2203 and 2204 have density change information AK2.
This corresponds to the case where three consecutive pixels of AK6 are detected.

Further, for pixel blocks 2205 and 2206, the density change characteristics of the peripheral pixels described above are similarly determined by the density change information AK4.
This corresponds to a case where three consecutive pixels of AK8 are detected.

Similarly, the pixel blocks 2207 and 2208 have density change information AK3.
This corresponds to the case where three consecutive pixels of AK7 are detected.

In this example, when extracting a series of density changes,
The reason why the pixel of interest is not at the center of continuity judgment is that
This is because pixels forming the end portion 2301 of a character as shown in FIG. 2 are also determined as pixels included in a continuous edge.

In addition, the Nant Gate 1809a has density change information AK6
It is detected that the features are continuous in the form of a pixel block 2203.

In addition, the Nant gate 1810a has concentration change information AK6
It is detected that the features are continuous in the form of a pixel block 2204.

Furthermore, the Nant gate 1811a receives concentration change information AK2.
It is detected that the features are continuous in the form of a pixel block 2203.

In addition, the Nant gate 1812a has concentration change information AK6
It is detected that the features are continuous in the form of a pixel block 2204.

Furthermore, the Nant gate 1813a has concentration change information AK5.
It is detected that the features are continuous in the form of a pixel block 2201.

In addition, the Nant gate 1814a has concentration change information AK5.
It is detected that the features are continuous in the form of a pixel block 2202.

Furthermore, Nando Game) 1815a has concentration change information AKI.
It is detected that the features are continuous in the form of a pixel block 2201.

In addition, the Nant gate 1816a has concentration change information AKI.
It is detected that the features are continuous in the form of a pixel block 2202.

Furthermore, the Nant gate 1817a has concentration change information AK7.
It is detected that the features are continuous in the form of a pixel block 2208.

In addition, Nando Game)-1818a has concentration change information AK
It is detected that the features of No. 7 are continuous in the form of a pixel block 2207.

Furthermore, the Nant gate 1819a has concentration change information AK3.
It is detected that the features are continuous in the form of a pixel block 2208.

In addition, the Nant gate 1820a has concentration change information AK3.
It is detected that the features are continuous in the form of a pixel block 2207.

Furthermore, the Nant Gate 1B21a has concentration change information AK8.
It is detected that the features are continuous in the form of a pixel block 22o5.

In addition, Nando Game) 1B22a has density change information AK8
It is detected that the features are continuous in the form of a pixel block 2206.

Furthermore, the Nant gate 1823a has concentration change information AK4.
It is detected that the features are continuous in the form of a pixel block 2205.

In addition, the Nant gate 1824a has concentration change information AK4.
It is detected that the features are continuous in the form of a pixel block 2206.

In this way, Nante Gates 1809a-1824a
Then, the center pixel CUP3. CBT3. CLF3. CRT3
．． CUL3. The continuity of three pixels having CBR3, CUR3, and CBL3 as ends is detected, and an edge detection signal EDGEO indicating that the center pixel constitutes a continuous edge is output to the edge determination unit 18022.

FIG. 33 is a circuit block diagram illustrating the configuration of the halftone dot processing section 1802b shown in FIG.
58, 1863 and 1864 are Noah gates, and Noah gates 1851 to 1858 are the concentration change point processing section 180 described above.
In this section, each pixel signal processed in step 2a is input, and a specific combination of density changes in a specific direction of the input image signal 314 is extracted as a halftone image as shown in FIG.

1859 to 1862.1865 are gate circuits.

In the circuit configured in this way, the NOR gate 185
Pixel group 22 consisting of 4 pixels such as 1 to 1858
A halftone detection signal DOT indicating that the pixel of interest 2250 is a pixel in a halftone image by detecting whether or not there is a change in density of pixels corresponding to pixels 51 to 2254.
O is output to the halftone signal area processing section 18021.

Specifically, the output of the NOR gate 1851 is the output of the pixel group 22
54 indicates that there is a downward density change of at least one pixel, and the output of the NOR gate 1852 indicates that there is an upward density change of at least one pixel in the pixel group 2253. 2254 indicates that there is an upward density change of at least one pixel, and the NOR gate 1
The output of 854 is at least 1 in pixel group 2254.
The output of the NOR gate 1855 indicates that there is a downward density change of at least one pixel in the pixel group 2252, and the output of the NOR gate 1856 indicates that there is a density change of at least one pixel or more to the right in the pixel group 2252. This indicates that there is a density change to the left above the pixel, and the output of the NOR gate 1857 is
52 indicates that there is a leftward density change of at least one pixel, and the output of the NOR gate 1858 indicates that there is a leftward density change of at least one pixel in the pixel group 2251.

These outputs are sent to subsequent gate circuits 1859-1862.
Logical operations are performed by 1865 and NOR gates 1863 and 1864, and as a result, the halftone determination shown in FIGS. 35(a) to 35(d) becomes possible.

FIGS. 35(a) to 35(d) are schematic diagrams illustrating the halftone determination pattern of the pixel of interest 2250 shown in FIG.
=>, e=, fJ, [1 in the figure is pixel group 22
51 to 2254, the concentration change is 1 in the direction indicated by the arrow.
Indicates that there are more than pixels.

As a result, if a direction as shown in FIG.
4 or the presence of the halftone dot pattern 2215 is detected, and the halftone dot detection signal DOTO is set to "1".

Next, the configuration and operation of the halftone signal area processing section 18021 that processes the halftone detection signal DOTO will be described.

FIG. 36 shows the halftone signal area processing unit 18 shown in FIG.
021 is a circuit block diagram illustrating the configuration of 183.
1 is a determination unit that determines whether or not there is one or more pixels whose halftone detection signal DOTO is “1” in a 4×3 window including the pixel of interest, and if so, halftone detection is performed. Internal signal.

TOI, otherwise halftone detection internal signal DOTO
Let I be "0".

18311 and 18312 are line memories, each with 1
A line delay is applied, and halftone detection signals DOTO for three lines are simultaneously input to flip-flops 1831 and 3.
OR gate 183'14. frib flop 181315
, 18316, and 18317, and their outputs are input to an OR gate 18318, which outputs the halftone detection internal signal DOTO1 to the area processing section 1832 at the subsequent stage.

In the area processing unit 1832, 18321°18322
are line memories, each of which delays one line. 18323.18324 is arithmetic unit, 18325
．． 18326 is an adder, which calculates the sampling sum SUML, S
Output UMR. 18327 and 18328 are comparators, which are compared with the slice level set in the register 1830, and when the relationship between input A and input B is A>B, the OR of the comparator output is the gate circuit 18329. A halftone dot area signal DOTl is outputted via the dot area signal DOTl.

The operation of the circuit configured in this way will be explained with reference to FIG. 37.

37 and 38 are schematic diagrams for explaining the operation of FIG. 36, and the same parts as in FIG. 36 are given the same reference numerals.

Note that in FIG. 37, the shaded area indicates that the pixel is rlj.

For example, halftone detection signals for three consecutive lines.

3 for the pixel of interest 1851 when TO is output.
A logical OR is performed within the ×4 window, and the halftone detection internal signal DOTO1 is calculated by the determination unit 1831.

Through this processing, the halftone detection signal DOTO which is sparsely scattered in the halftone image is converted into a relatively continuous halftone detection internal signal DOTO.
Converted to TO1.

The halftone detection internal signal obtained in this way.

TOl is line memory 18321.1 shown in FIG.
8322, and the main scanning 4
Halftone dot detection internal signal DO every other sub-scan line every pixel
Sample the TOI. Then, one line before (i-
1st line), the halftone detection internal signal DOTOI is "1" in each main scanning pixel j, j-4, j-a, . . . , j-4th, where N is an appropriate integer. The main scan j, j+4 . j+8. . . , for each pixel at the j+4th pixel, the arithmetic unit 18323 outputs the sum SUMRl of the halftone detection internal signal DOTO1 which is rIJ.

Furthermore, for the pixel of interest 1861 (5!! I-th line of scan, first pixel in main scan), the halftone detection internal signal DOTOI is sampled every four pixels in the main scan and every other line in the sub-scan. Then, 1 line later (i-1st line)
In, main scanning j, j-4 .
For each pixel of j-8, . . . , j-4, the dot detection internal signal DOTOl is "1".
With L2 and N being appropriate integers, main scan j, j+4 .
For each pixel of j+8, . . .
The arithmetic unit 18324 outputs R2.

In response to this, the adder 18325 calculates the addition of the sum sum SUML1 and the sum sum SUML2, and the pixel of interest 186
The sampling sum SUML of the halftone detection internal signal DOTOl on the left side of 1 is output to the comparator 18327.

On the other hand, the adder 18326 outputs the sum SUMRI and the sum 30M.
The addition with R2 is calculated, and the sampling sum SUMR of the halftone detection internal signal DOTol on the left side of the pixel of interest 1861 is calculated and output to the comparator 18328.

In this way, when the sampling sum SUMR and the sampling sum SUML are obtained, the comparator 18327 compares them with the slice level T4 set in the register 1830, and if at least one of SUML>T4 or SUMR>74 holds true, That is, in the halftone dot area, the halftone dot detection signal DOTl becomes "1".

Next, the operation shown in FIG. 24 will be explained with reference to FIG. 39.

FIG. 39 shows the edge determination unit 18022 shown in FIG.
1841.1 is a circuit block diagram explaining the configuration of 1841.1.
842 is a line memory that receives the edge detection signal EDGE.
O is delayed in synchronization with the horizontal synchronizing signal HSYNC, synchronizing sub-scanning with the halftone dot detection signal DOTI, and each flip-flop 1843a to 1843 of the subsequent stage synchronizing circuit 1843
43d etc., the main scanning synchronization of the edge detection signal EDGEO and the halftone dot detection signal DOTl is performed.

1844.1850.1851 is inverter, 1848
is Nantes Gate, 1845.1849 is And Gate,
1846 is an or gate, and 1847 is an or gate.

Note that MODO and MODI are mode signals output from the CPU 1871.

The edge determination unit 18022 shown in FIG. 24 is configured as described above, and the detected edge detection signal EDGE is
Edge control signals EGI to EG3 defining the relationship between O and the halftone detection signal DOTl are determined as described later, and a final area signal (edge signal) EDGE is output.

Specifically, in the character mode, the mode signal MO
Since DO, MODI are set and controlled as "r□J,ri", the edge control signal EGI is always "1", and the output of the OR gate 1846 is always "1", so the area signal (
edge signal) EDGE becomes "1".

In addition, in the photo mode, the mode signals MODO,
Since MODI is set and controlled as "rl" and "r□", edge control signals EGI and EG2 are both "0", and the area signal (edge signal) EDGE is always "O".

Furthermore, in the map mode, the mode signals MODO,
Since MODI is set to "0" and "O", the edge control signal EG3. Since EG2 becomes "1" and the edge control signal EGI becomes "0", the edge detection signal E
DGEO1 itself is output as an edge signal EDGE. Therefore, the edge detection signal EDGEO1 is output at the density change point regardless of the presence or absence of a halftone dot area, so that it is possible to clearly copy blanks in halftone dots, such as on a map.

In addition, in the text/photo mode, the mode signal MO
Since the settings of DO and MODI are controlled as "rl" and "rl", the edge control signals EGI and EG3 are both "O", and the edge control signal EG2 is "1", so the edge signal EDGE is set to be equal to the edge detection signal EDGEO. The logical product of the dot detection signal DOTO and the date is taken.

Therefore, at the density change point, the edge detection signal EDGEOI is "1" and the halftone detection internal signal DOT is not in the halftone region.
Edge signal EDGE is “1” only when OI is “O”
Then, the character edge area other than the halftone area is separated.

In other words, in the pixel of interest, the edge signal ED
When GE is "1", that is, a continuous density change image exists, and the halftone dot detection internal signal DOTOI, that is, it is not a halftone dot area, the edge signal EDGE, which is regarded as a character edge and is faithful to the original, is "1". Become.

Next, signal processing by the color signal processing section 402 shown in FIG. 7 will be explained.

The light amount signal-density signal conversion unit 103 converts the image signals 313 to 315 in the range of 0 to 255 into 0 to 315 according to the following equation (8).
The signal is converted into concentration signals C, M, and Y in the 255 range.

(8) Note that the density signal that is the black component included in the density signals C, M, and Y is determined by the black extraction unit 104 using the following equation (9).

K=min (C, M, Y) =...(9)
Then, the density signals C and M of the above four colors.

In Y, the undercolor is removed in the masking processing unit 105, and the calculation is performed according to the following equation (1o) in order to remove the color clouding of the developer material in the printer unit 202.

......(10) In addition, in the above formula, aI l~a 14+ a21~a2
4゜a31~a341 a4□~a44 are preset masking coefficients for removing color turbidity, and U1~U3 are UCR coefficients for removing the color component from each color component of M, C, and Y. . Also, Ml.

One of C1, Yl, and Kl is selected by a face signal PHASE (2-bit signal) outputted from the control section 401, and outputted as a developed color signal vl.

The line delay memory 112, 113 is a feature extraction unit 403
The development color signal Vl and the density signal M, which are used to delay the generation of the character edge determination signal from the character edge judgment signal from the image data, are similarly delayed by three lines and four clocks.

On the other hand, the color determination unit 106 delays by two lines and two clocks by the line delay memory 120 until determination outputs such as the black image analysis signal BL and the mixed analysis signal UNK are generated. Mixed analysis signal UNK
I, delayed color analysis signal C0LI, delayed cancellation signal C
Generate AN 1.

Next, with reference to FIG. 40, a description will be given of color determination processing and character edge determination processing for characters rAJ included in various types of originals.

FIG. 40 is a characteristic diagram illustrating the characteristics of the character determination signal output from the feature extraction unit 403 shown in FIG. 7, and various determination signal outputs corresponding to the cross section a of the character shown in FIG. The states are shown in (a) to (f) of the same figure.

Particularly, FIG. 4A shows the case where the black character rA4 is read as black, and the achromatic density signal (density signal M2) is read duller than in FIG. In addition, the edge signal EGDE is the density change information AK.
3 and density change information AK7, it is formed in a shape that bulges from the end of the character, and as a color judgment signal,
Only the delayed black image analysis signal BLI is generated.

Note that the concentration signal M2. Since the edge signal E D GE uses a green color separation signal, except for the green characters, the density signal M2 from (b) to (f) in the same figure is the same as the density signal M2 shown in (a) in the same figure. A density signal M2 is output. Furthermore, in the case of green characters, the density signal M2 is not output.

In addition, (b) in the same figure corresponds to the judgment output when reading FAJ composed of colored characters, and shows a delayed color analysis signal C0L1 indicating that a color exists and a white pixel or more around a white pixel. This shows a state in which a delayed cancellation signal CANI indicating that a color pixel exists with a density of .

Furthermore, in (e) of the same figure, r
This shows a state in which a delayed mixed analysis signal UNKI indicating intermediate saturation is generated, corresponding to the judgment output when AJ is read.

In addition, (d) in the same figure corresponds to the judgment output when r A J consisting of black characters is read with color shift.
Compared to a), the delayed black image analysis signal BLI is thinner, and furthermore,
Delayed intermediate chroma signal UNKI due to color shift around that pixel
corresponds to the state determined and output by.

Furthermore, Figure (e) shows the case where 'AJ, which is composed of colored characters, is read with color shift, and compared to Figure (b), the delayed color analysis signal C0LI is thinner and the edge of the character is The delayed intermediate saturation signal UNKI is output for determination. Furthermore, since the portion of the delayed cancel signal CANI that is determined to be a color decreases, the portion corresponding to the outside of the character edge becomes thinner.

Further, FIG. 6(f) corresponds to a case where a color character with near intermediate saturation is read with a color shift and a black determination pixel occurs at the edge.

This case corresponds to the state in which the same determination signal as in FIG. 4(e) is output, except that the delayed black image analysis signal BLI is generated instead of the delayed intermediate saturation signal UNKI.

FIG. 41 is a characteristic diagram showing the signal output characteristics of each part of the color signal processing section 402 shown in FIG. 7, and corresponds to the judgment output of the main part of the section a of the letter 'A4' shown in FIG.

FIG. 7A shows a case where black characters are read, and the masking processing unit 105 shown in FIG. 7 performs UCR processing (undercolor removal processing) to generate density signals Y and M.

Although the color component of C has decreased to about 20%, since this character is a black character, it is desirable to record it with black toner as much as possible.

Furthermore, FIG. 40(b) shows that it is desirable for the intermediate saturation that occurs at the edges of black characters as shown in FIG. 40(d) to cover the color components of the density signals Y, M, and C as much as possible. ing.

Furthermore, FIG. 40(c) shows that it is desirable to reduce the intermediate saturation component that occurs at the edges of colored characters as shown in FIG. 40(e).

Because of these characteristics, it is necessary to distinguish the black component generated at the edge of the colored character shown in FIG. 40(f) from the black character shown in FIG. 40(a).

Therefore, as shown in FIG. 42, the edge signal EDGE,
Delayed black image analysis signal BLI, delayed mixed analysis signal UNKI
, delayed color analysis signal C0L1. Delay cancellation signal CA
A predetermined calculation (details will be described later) is performed on the color recording signal v2 from N1 and the phase signal PHASE, etc. with the density signal M2 to obtain an image signal v4.

FIG. 42 is an explanatory diagram showing an example of arithmetic processing of the adder 116 of the color signal IA embedding section 402 shown in FIG.
(h) shows an example of calculation.

In the case where the edge signal EDGE, delayed black image analysis signal BLI, delayed mixed analysis signal UNKI, delayed tint analysis signal C0LI, and delayed cancellation signal CANI as shown in calculation example (a) output the values shown in the figure, The developing color is magenta,
In the case of cyan and yellow, the image signal v4 is rO'', and when the developing color is black (BK), the image signal V4 is
4, the color recording signal V2 itself is calculated and output.

In addition, as shown in calculation examples (b), (d), and (g), when the delay cancel and cell signal CAN 1 is "1" and the developed color is magenta, cyan, or yellow, the image signal v
4 becomes the color recording signal V2J, and when the developed color is black (BK), the image signal 4 becomes the color recording signal v2.
That information is calculated and output.

Furthermore, in the calculation example (C), when the edge corresponds to an intermediate chroma edge as shown in FIG. In this case, 5 of the color recording signal v2 is used as the image signal V4.
Also, the developing color is black (BK) so that it outputs 0%.
), 50% of the density signal M2 is added to 50% of the image signal V2.
The result obtained by adding 0% is calculated and output from an adder 116, which will be described later.

Calculation example (f) corresponds to the calculation of the non-edge portion of black characters shown in FIG. When the color is magenta, cyan, and yellow, the image signal v4 is reduced to 3/4 of the color recording signal v2, and when the developing color is black (BK), the image signal v4 is reduced to 3/4 of the color recording signal v2. The adder 116 calculates and outputs an image signal V4 obtained by adding the density signal M2 reduced by 1/4 to the image signal M2.

Next, a more detailed explanation will be given with reference to FIGS. 25 and 26. In the following description, V2(M) corresponds to the density signal when the 7-phase signal PHASE=0 (corresponding to magenta). Similarly, V2 (C) corresponds to the density signal when the phase signal PHASE=1 (corresponding to cyan). Similarly, V2(Y) is the phase signal PH
This corresponds to the density signal when ASE=2 (corresponds to yellow). Similarly, V2 (BK) is the phase signal PHA
This corresponds to the density signal when SE=3 (corresponding to black).

■ in Fig. 41(a) is the black text part, and Fig. 42(a)
). Here, the color recording signal V2 becomes "O", and the density signal M2 is output as the image signal v4.

41(a) is a black non-edge area corresponding to the calculation example (f) in FIG. 42, and the image signals V4 (Y), V4 when the developed colors are magenta, cyan, and yellow
(M), V4 (C) are recorded image signals V2 (Y)
X3/4. V2 (M)X3/4, V2 (C)X3/
4, and when the developing color is black (BK), the recorded image signal V2(Y)X3/4. V2 (M)X3/4.
V2 (C)X3/4 and color recording signal V2 (BK)x3
An image signal v4 obtained by adding /4 is output.

Furthermore, ■ in FIG. 41(b) is an edge portion corresponding to the calculation example (c) shown in FIG.
(Y), V4 (M), and V4 (C) are recorded image signals V
2 (Y)Xi/2. V2 (M)X 1/2. V2
(C) V2
(M) Xi/2°V2 (C) xi/2 and color recording signal V
An image signal V4 obtained by adding 2 (BK)X1/2 is output.

In addition, ■ in FIG. 41(C) is a portion where intermediate saturation occurs in the edge portion of black characters, and is processed in the same manner as the edge portion ■ in FIG. 41(b), and non-edge portions are determined to be black. , is processed in the same way as the 0 part in FIG. 25(a), and the color signal at the edge of the black character is reduced.

The above operations are executed by the multiplier 114 and adder 116 shown in FIG. These multipliers 114.
The adder 116 receives each edge signal EDGE, delayed black image analysis signal BLI, delayed mixed analysis signal UNK1 . Delayed color analysis signal C0LI, delay cancellation signal CANI (each determination signal EDGE, BL
The above calculation is performed on the image recording signal V2 and the density signal M2 based on the multiplication coefficient signals GAINI and GAIN2 determined by IUNKI, C0LI, and CANI.

Figure N43 is a block diagram explaining the configuration of the multiplication coefficient generating section 10B shown in Figure 7, and 108a is a ROM;
5-bit judgment signal EDGE.

Two multiplication coefficient signals GAINI of 3 bits each are generated from the data table shown in FIG. 44 according to the address determined by BLl, UNKl, COLI, CANI and the 2-bit phase signal PHASE.

Output GAIN2. Note that the two multiplication coefficient signals GAINI and GAIN2 are obtained by multiplying the actual gain by 4, and are multiplied by 174 by multipliers 114 and 115 shown in FIG. 45 to obtain the desired calculation output.

FIG. 45 is a block diagram illustrating the detailed configuration of the multipliers 114 and 115 shown in FIG. 7, and the configuration and operation will be described below.

The 8-bit image signal is processed by a bit shift type multiplier 2901.
The gate 29 is controlled by the 3-bit gain signal GAIN (2), (1), and (0), which are multiplied by 4 and 2 by 2902, respectively.
Selected from 03 to 2905, adder 2906.290
7 is added. After this, the bit shift type divider 29
Multiplied by 1/4 in 08, 25 in 255 limiter 2909
All 9-bit data of 5 or more are combined into 8-bit data of 255 lenses, and an image signal V4 is calculated and output.

As described above, the color recording signal ■2 and the density signal M2 are weighted using the color determination signal and the character edge determination signal.
is input to a filter circuit 117 which serves as a spatial filter.

FIG. 46 is a circuit block diagram illustrating the detailed configuration of the filter circuit 117 shown in FIG. 1/2, 1 of 2
It is possible to switch depending on the type.

The configuration and operation will be explained below.

3001.3002 are image signals V4.3002 for three lines generated by each line delay memory. V42 and V45 are each delayed by one clock by flip-flops 3003-3006. Here, the pixel of interest is pixel signal V43, pixel signals V41, V42 . V44. V46 is "-1" in multipliers 3007 to 3010 to form a Laplacian.
They are multiplied and added by adders 3011, 3012, and 3013, respectively. Furthermore, the pixel signal V43, which is the pixel of interest, is multiplied by 4 in the multiplier 3014, and V43F is added to the output of the adder 3013 in the adder 3015, and the Laplacian L is generated.

This Laplacian signal is multiplied by 1/2 in a multiplier 3o16. In the adder 3017, the pixel of interest V43 and L/2 are added to generate a weak edge emphasis signal E1. Adder 3
At step 018, the pixel signal V43 and the Laplacian L are added to generate a weak edge emphasis signal E2. These two types of edge-enhanced signals and the pixel signal V43 of the pixel of interest are control signals DFIL (1) and DFIL (0).
is selected, and the image signal 5 is output. Note that when the control signal DFIL(1) is "0" and the control signal DFIL(0) is "1", the weak edge emphasis signal E1 is selected,
When the control signal DFIL (1) is "1", the control signal DFIL
When (0) is "1", the strong edge emphasis signal E2 is selected, and when the control signal DFIL (0) is rQJ, the pixel signal V43 with no edge emphasis is selected by the selector 301.
9.3020 and output as image signal V5.

In addition, the above control signals DFIL (1), DFIL (0)
is output from a filter control signal generating section 109, which will be described later.

In this embodiment, strong edge enhancement is applied to the edges of black characters so that the edges of the black characters become sharp.

Furthermore, processing is performed so that edge enhancement is not applied to non-character edge portions in order to prevent the color tone from changing due to edge enhancement. The edge portions of characters with intermediate saturation and colors are processed so that the edge portions are sharply recorded and a weak edge emphasis is applied so that the change in tone due to edge emphasis is not so noticeable.

Furthermore, when the determination signal CAN 1 is "1", the determination signal BLI, which is generated due to color shift at the edge of the color character,
Since there is UNKI, processing is performed so as not to emphasize the edges.

Next, the configuration and operation of the filter control signal generating section 109 will be explained with reference to FIGS. 47 and 48.

FIG. 47 shows the filter control signal generator 1 shown in FIG.
10 is a circuit block diagram illustrating the detailed configuration of 09;
9a to 109e are AND gates, 109f is Nantes gates, 109g to 109j are OR gates, 109 to 10
9m is an inverter.

FIG. 48 shows the control signals DFIL(1) and DFI output from the filter control signal generator 109 shown in FIG.
This is a correlation diagram illustrating the relationship between L(0) and filter processing, and when setting the text/photo mode as shown in FIG. ), DFI L (0) are alternately changed and filter processing is performed, and weak edge processing corresponding to each is performed.

In addition, in the character mode or photo mode as shown in FIG.
Both become "1", and strong edge emphasis processing is performed on the entire copy.

Furthermore, in the photo mode as shown in FIG. 2C, the control signals DFIL (1) and DFIL (0) become "O", and the entire copy is subjected to smoothing processing.

Further, in the text/photo mode in the above embodiment, the areas that are not judged as edges by the text edge judgment unit 107 also include the halftone area (halftone area') 1903 shown in FIG. When such a halftone original is read pixel by pixel by the 3-line sensor 210, moiré fringes occur due to the regularity of the charge-coupled device such as COD that constitutes the 3-line sensor 210 and the regularity of the halftone original. . In order to prevent this, in this embodiment, the filter circuit 117 uses a smoothing filter as shown in FIG. The pixel signal V43 of the remaining four adjacent pixels is multiplied by 1/2 to obtain the pixel signal V43 of the remaining four adjacent pixels.
41. V42°V44. A smoothing signal (smoothing filter signal) SMG is generated and output by a smoothing filter that converts each value by multiplying V48 by 1/8.

That is, in FIG. 46, adders 4201 to 420
4, the 4 pixel signals around the pixel of interest V41. V42
．． V44. V46 is subjected to addition processing, and the pixel signal V43, which is the pixel of interest, is multiplied by 4 to the added signal.
The smoothing signal (smoothing filter signal) SMG is obtained by adding 43F and further dividing the addition result by 1/8 using a bit shift time divider 4205.

Note that in the filter circuit 117, the pixel of interest is delayed by one line and one clock, so the control signals FIL(0) and (1) from the filter control circuit generator 109 are delayed by one line and one clock in the line memory 121. and is input to the filter circuit 117 as the delayed filter switching signal DFIL.

Next, the configuration and operation of the gamma conversion section 118 shown in FIG. 7 will be explained with reference to FIGS. 50 and 51.

FIG. 50 is a block diagram illustrating the configuration of gamma conversion section 118 shown in FIG. 7.

In the figure, 118a is a gamma conversion ROM, and the image signal V
For example, the 8-bit video signal VIDEO is sent to the printer unit 202 using the address 5 and the delay selection control signal GGAM.
The signal is output to the PWM modulation section 119 of.

Specifically, four types of gamma conversion characteristics I to ■ as shown in FIG. 51 are selected depending on the contents of the delay selection control signal DGAM.
”, the gamma conversion characteristic I is applied to the non-character edge part.
is selected.

In addition, when the delay selection control signal DGAM is "1" (in the case of color character edge detection), in addition to the 0 side and 255 side for each range from 0 to 255, the input corresponding to the j section center is 0 and 255. output with a slope of 255/
The conversion characteristic (gamma conversion characteristic ir) is connected by a straight line (255-2j). This is a low concentration input “0”
Video signal VIDE with lower density for nearby inputs
0 is output from the gamma conversion ROM 118a, a higher density video signal VIDEO is output for an input near 255 which is a high density input, and a video signal VIDEO of a higher density is output for an input near 128 which is an intermediate density input. is output, emphasizing the density change of the input near 128, which is the intermediate density, and making it possible to record character edges more chirply.

Furthermore, when the delay selection control signal DGAM is "2" (in the case of intermediate chroma character edge detection), the gamma conversion characteristic II
I is selected, and the interval is made larger than the interval j when the delay selection control signal DGAM is "1",
Character edges will be recorded sharply. However, since the linearity between input and output collapses, color tone cannot be guaranteed. Therefore, it is selected when recording medium saturation character edges.

Further, when the delay selection control signal DGAM is "3", the gamma conversion characteristic (2) having a section .degree. C. which has a larger value depending on the section is selected, and is applied to black character edges where sharpness is required.

The delay selection control signal DGAM, which becomes this gamma conversion characteristic switching signal, is generated by transmitting the selection control signal GAM from the gamma switching signal generation section 110 shown in FIG. 52, which will be described later, to the delay memory 12.
1 is delayed by one line and one clock.

FIG. 52 shows the gamma switching signal generating section 1 shown in FIG.
110a is a gamma switching conversion ROM, and a 5-bit judgment signal EDGE is shown.
, BL1. Using UNKI, C0LI, CANI and the edge signal EDGE as addresses, the selection control signal CAM is output to the delay memory 121 when the relationship as shown in FIG. 53 is satisfied, for example.

For example, when detecting a black character edge (the edge signal EDGE is
1" and the judgment signal BLI is "1"), the selection control signal CAM becomes "3", and when an intermediate saturation character edge is detected (the edge signal EDGE is "1" and the judgment signal UNK is "1"), the selection control signal CAM becomes "3". ), the selection control signal CAM becomes "2", but in either case, the judgment signal BLI becomes "1" due to color shift.
” or when the determination signal CANl indicating that the determination signal UNKI has been reached is “1”, the selection control signal CAM is set to “0” so as not to emphasize the character edges.

Next, the configuration of the PWM modulator 119 shown in FIG. 7 will be explained with reference to FIGS. 54 and 55.

54 is a circuit block diagram illustrating the detailed configuration of the PWM modulator 119 shown in FIG. 7, and FIG.
5 is a timing chart illustrating the operation shown in the figure.

The configuration and operation will be explained below.

- The video signal VIDEO is converted into an analog image signal AV by the D/A converter 3701 and sent to the comparator 3708,
3709 is input. The pixel clock CLK synchronized with this video signal VIDEO and the screen clock CLK4 with twice the frequency are supplied to a toggle flip-flop 37.
At 02.3703, the pixel clock C is divided into 1/2 in synchronization with the horizontal synchronization signal HSYNC and has a duty of 50%.
LK and screen clock CLK4.

These two pixel clocks CLK and screen clock CLK4 are transformed into triangular waves by integrators 3704 and 3705 composed of resistors R and capacitors C, and then
Amplifiers 3706 and 3707 adjust the wave height to match the output dynamic range of the A/D converter, and analog comparators 3708 and 3709 respectively compare and process the analog image signals AV.
Type of pulse width modulation signal PW4. Converted to PW. Thereafter, in the selector 3710, one of the pulse width modulation signals PW4°PW is selected by the delayed screen control signal DSCR output from the delay memory 121, and then inputted as the laser drive signal LDR to the drive circuit of the laser unit (not shown). be done.

Specifically, as shown in FIG. 55, the triangular wave TRI4, which is obtained by integrating the clock CLK4F obtained by dividing the screen clock CLK4 into 1/2, is a triangular wave having a period of one pixel of the image.

Since this triangular wave TRI4 changes almost realistically over the entire output range of the D/A converter 3701, by comparing this triangular wave TRI4 and the analog image signal AV, the analog image signal AV is is pulse-width-modulated with one period as a pulse-width modulated signal PW4.

Similarly, since the triangular wave TRI is created by the clock CLKF obtained by dividing the pixel clock CLK by 172, the triangular wave TR
By I, the analog image signal AV becomes a pulse width modulated signal PW which is pulse width modulated with one cycle being a two-pixel section of the image.

In this way, the pulse width modulated signal PW4 pulse width modulated in one pixel period is recorded by a deprinter with the same resolution as the pixel clock CLK. However, when an image is recorded using the pulse width modulation signal PW4, the basic density unit is as small as one pixel, so the gradation expression cannot be said to be sufficient due to the characteristics of the electrostatic photographic process used in the printer.

On the other hand, since the pulse width modulation signal PW reproduces the density in units of two pixels, the gradation expression is sufficient, but the recording resolution is half that of the pulse width modulation signal PW4.

Therefore, in this embodiment, the pulse width modulation signal P is controlled by the delayed screen control signal DSCR depending on the type of image.
W4. By selectively controlling one of the pulse width modulated signals PW4.PW. Switch PW for each pixel.

Specifically, the pulse width modulation signal PW4 is selected to emphasize color tone for black character edges, intermediate saturation character edge parts, and non-edge parts that require resolution, and the color tone is selected for color character edge parts and non-edge parts. The pulse width modulation signal PW is selected based on the meaning of emphasis. However, it has been experimentally confirmed that it is better to adopt the pulse width modulation signal PW4, which emphasizes the resolution of color character edges, even at the expense of color tone, even for originals composed of thin color characters such as maps. Although it is.

The delayed screen control signal DSCR is generated from the screen control signal SCR output from the screen switching signal generating section 111 shown in FIG. The configuration and operation of the screen switching signal generating section 111 shown in FIG. 7 will be described below with reference to FIG. 56.

FIG. 56 is a circuit block diagram illustrating the configuration of the screen switching signal generating section 111 shown in FIG.
1a and 1llb are or gates, 11c and 111d are nante gates, and 111e.

111f is an AND gate, and 111g to 1111 are inverters.

As can be seen from this figure, the determination signal EDGE.

BLI, UNKI, C0LI, CANI are processed by gate IA and the judgment output is output from Nantes gate 110KC.
)-111f, the screen control signal SCR is set to "1" or "0" by the mode signals MODO and MODI outputted to one input of the AND gate 111f.
Controlled by setting.

Specifically, in the character mode, the mode signals MODO, M
The ODI becomes rQ", rl", and the screen control signal S
CR is always “0” and the pulse width modulation signal PW4 is always
let them choose.

Also, in photo mode, mode signal MODO.

MODI becomes riJ, rQJ, screen control signal SCR is always "1", and pulse width modulation signal pw is always
let them choose.

Furthermore, mode signals MODO and MODI are "1" in text/photo mode and map mode.

rl”, rQ4. "OJ", when the black character area, that is, the judgment signal BLI is "1", selects the pulse width modulation signal PW4, and otherwise selects the pulse width modulation signal P
Let W4 be selected.

As a result, the delayed screen control signal DSCR shown in FIG. 55 becomes LOW in the portion corresponding to black or intermediate saturation character edge portion and the portion corresponding to black toner development.
Only in this period, the pulse width modulation signal PW4 is output as the laser drive signal LDR. At this time, in the case of a character edge part that is determined to be a character edge part but has a color shift (determination signal CANl is "1"), in order to prevent the quality of the recorded image from deteriorating due to the color shift being emphasized, The pulse width modulation signal PW4 is not selected.

That is, sharp character edges are required for black character edges, and for color character edges, reproduction of the color tone of the original is important.

On the other hand, as shown in FIG. 41(a),
M, C, and Y toners do not exist in the black character edge portion. Further, due to the function of the masking processing unit 105, black toner no longer exists.

Furthermore, as shown in FIG. 41(b), a moderate amount of black toner, yellow toner, magenta toner, and cyan toner are also present in the intermediate saturation character edge portion.

Based on the above considerations, in this embodiment, the character edge determination unit 107 is driven by laser only when using black toner.
The pulse width modulation signal PW4 of the pixel period can be used.

As a result, the edges of black characters, which originally have few color components, can achieve the same sharpness as in the first embodiment, and the edges of color characters, which contain a small amount of color components, can be sharply recorded with only the black component, and the color components are Since the tonality is maintained, color reproducibility is also guaranteed. In the above embodiment, the screen switching signal generating section 11
1, the screen control signal SCR is determined by the judgment signal EDGE.
, BLI, UNKI, C0LI.

Although the case where the face signal PHASE outputted from the control unit 401 has been described is obtained by gate processing of CANI,
It may be configured such that the screen control signal SCR is output by decoding that the developed color is black.

In addition, in the above embodiment, for a general original containing a mixture of text and photographs, it is possible to select whether the copy reproduces the text clearly or faithfully reproduces the photograph.
For example, as shown in FIG. 57, as shown in the operation section 1870, a priority reproduction type setting key 4214 serving as an unbiased weight setting means is provided.
a, 4214b is provided, and the text-in-photo priority mode or the photo-in-photo priority mode is set in stages, and the set polarization degree is analyzed and the lower slice level shown in FIG. 30(a) is set. 1 to T3 may be subordinated to vary the detection conditions so that the reproduction directivity of characters or photographs in a photographic manuscript can be selected.

FIG. 57 is a plan view illustrating the configuration of an operating section in an image reading device showing another embodiment of the present invention;
Reference numeral 14 denotes a reproduction directivity indicator composed of, for example, an LED, which displays the degree of polarization when a priority reproduction type setting key 4214a or 4214b is pressed.

Note that when the leftmost display part of the reproduction directionality display 4214 is lit by the priority reproduction type setting key 4214a, the mode is determined to be the "most character-oriented mode", and reading is performed in the same manner as the character mode in the above embodiment. Process the image signal.

Furthermore, when the display section at the earliest position of the reproduction directionality display 4214 is lit by the priority reproduction type setting key 4214a, it is determined that the mode is the "most photographic mode" and the reading is performed in the same manner as the photographic mode in the above embodiment. Process the image signal.

4209a is a map mode setting key, and when this map mode setting key 4209a is pressed, the mode display 4
213 lights up to indicate that the "map mode" is being set.

In other words, the priority reproduction type setting key 4 serves as a bias setting means.
214a, when the document reading mode characteristics set by the mode setting means, for example the map mode setting key 4209a, are continuously set to emphasize character reading mode characteristics or photograph reading mode characteristics, the condition setting means (cpU 1871 in this embodiment) is the priority reproduction type setting key 42
At least one judgment condition of the first detection means and the third detection means is variably set depending on the degree of bias of the mode characteristic biased by 14a, and text or a photograph is selected from a document containing a mixture of text and photographs. It is possible to faithfully separate and determine sexual desire.

As a result, when the map mode setting key 4209a is pressed, the "map mode" is set, and when it is not, the read image signal is determined and processed in the same manner as the text/photo mode in the above embodiment.

However, in this case, as mentioned above, the priority reproduction type setting key 4
214a is biased, and depending on the degree of bias, the slice levels T, ~T3 shown in FIG. 30(a) are subordinated to vary the detection conditions. , by making it possible to select the reproduction directionality of characters or photographs in a photographic manuscript, and in particular, by increasing the values of slice levels T1 to T3 as the degree of emphasis on photography is higher, erroneous judgments due to noise in photographs are reduced. becomes possible.

(Effects of the Invention) As explained above, the present invention includes a mode setting means for setting and inputting a plurality of original G4 reading modes corresponding to the type of document, and a mode setting means for setting and inputting a plurality of original G4 reading modes corresponding to the type of document, and determining the density level of surrounding pixels including the pixel of interest in an image signal. a conversion means for comparing the density level and converting it into a predetermined density level signal; a judgment level variable means for varying the judgment density level based on the document reading mode set by the mode setting means; and a character area separation processing means for separating character edge regions from the image signal based on a predetermined density level signal output from the conversion means in accordance with the set judgment density level, so that the density of the read image signal is It becomes possible to vary the level conversion standard depending on the original gJ reading mode, greatly reducing misjudgment between characters and halftone dots, and generating an image signal capable of reproducing a faithful original image.

The determination level variable means is configured to variably set the determination density level so that the detection ratio of high-density level pixels to low-density pixels is given priority to low-density pixels, so that low-density originals can be detected with extremely high sensitivity and read. Pixels can be accurately identified as character edges.

Furthermore, the conversion means is configured to convert the density levels of surrounding pixels including the pixel of interest in the image signal into a predetermined density level signal based on the conversion lookup table, so it can be input with a very simple configuration. It is possible to quickly convert an image signal to a desired density level signal.

Furthermore, since the conversion means is configured to convert the density levels of surrounding pixels including the pixel of interest of the color image signal into a predetermined density level signal based on the conversion lookup table, It has excellent effects such as being able to faithfully separate and detect a character original from an intermediate saturation original, and generating and outputting an image signal that can faithfully reproduce desired color characters and color halftone images.

[Brief explanation of drawings]

FIG. 1 is a sectional view showing an example of an image reading device according to an embodiment of the present invention, FIG. 2 is a plan view illustrating the configuration of an operating section in the image reading device according to the invention, and FIG. FIG. 4 is a flowchart illustrating an example of the back tank processing procedure in four types of original modes, and FIG. 4 is a block diagram illustrating the configuration of the image scanner section shown in FIG.
5 is a block diagram explaining the configuration of the signal processing section shown in FIG. 1, FIG. 6 is a tying chart explaining the operation of FIG. 5, and FIG. FIG. 8 is a detailed block diagram of the main parts of the processing section. FIG. 8 is a configuration block diagram explaining the configuration of the color determination section shown in FIG. 7. FIG. 9 is a characteristic explaining the received light wavelength and visibility in the image scanner section figure,
FIG. 10 is a circuit block diagram illustrating the detailed configuration of the color determination section shown in FIG. 7, and FIG.
FIG. 12 is a circuit block diagram illustrating the detailed configuration of the AX/M I N detector, and FIG. 12 is a correlation diagram illustrating the output conditions of the input image signal and each determination signal shown in FIG. 11. FIG. 13 is a circuit block diagram explaining the detailed configuration of the selector shown in FIG. 10, FIG. 14 is an explanatory diagram explaining the operational functions of FIG. 13, and FIG. 15 is an image reading system according to the present invention. Schematic diagram illustrating pixel color determination divisions of the device, No. 16
The figure is a correlation diagram explaining the relationship between the area divisions shown in FIG. 15 and the output signals of the MAX/MIN detector, and FIG.
FIG. 4 is a spatial characteristic diagram showing the color balance characteristics of the 3-line sensor, FIG. 18 is a circuit block diagram explaining the configuration of the area processing section shown in FIG. 8, and FIG.
An operation explanatory diagram illustrating the operation of the counting means showing the centroid,
FIG. 20 is a cancel signal generation circuit diagram of the color determination section shown in FIG. 7, FIG. 21 is a schematic diagram illustrating the state of occurrence of color fringing, and FIG. 22 is a diagram at the specific position shown in FIG. 21. FIG. 23 is a characteristic diagram illustrating the light quantity characteristics, and FIG.
25 is a block diagram illustrating the configuration of the character edge determination section shown in FIG. 7, FIG. FIG. 26 is a principle explanatory diagram illustrating the principle of character edge determination processing according to the present invention, FIG. 27 is a schematic diagram illustrating the configuration of a pixel block according to the present invention, and FIG. 28 is a diagram illustrating the density change characteristic determination pattern according to the present invention. 29 is a circuit block diagram illustrating the detailed configuration of the character edge determination section shown in FIG. 7, and FIG. 30(a) is a schematic diagram illustrating the detailed configuration of the detector shown in FIG. 29. 30(b) is a schematic diagram illustrating the conversion table delta of the density conversion ROM shown in FIG. 30(a), and FIG. 30(C) and (d) are circuit block diagrams to be explained. 30(e) is a characteristic diagram illustrating the concentration conversion characteristics in FIG. 30(d); FIG. 31 is a circuit block diagram illustrating still another detailed configuration of the detector shown in FIG. 29. FIG. FIG. 29 is a schematic diagram illustrating a specific pixel pattern detected by the density change processing unit shown in FIG. 33 is a circuit block diagram explaining the detailed configuration of the halftone dot processing section shown in FIG. 24, FIG. 34 is a schematic diagram explaining the halftone detection pattern according to the present invention, and FIG.
5(a) to (d) are schematic diagrams illustrating the halftone determination pattern of the pixel of interest shown in FIG. 34, and FIG.
37 and 38 are schematic diagrams explaining the operation of FIG. 36, and FIG. 39 is a circuit block diagram explaining the configuration of the halftone signal area processing section shown in FIG. FIG. 40 is a circuit block diagram illustrating the configuration of the edge determination section, and FIG. 41 is a characteristic diagram illustrating the characteristics of the character determination signal output from the feature extraction section shown in FIG. FIG. 42 is an explanatory diagram showing an example of arithmetic processing of the adder of the color signal processing section shown in FIG. 7, and FIG. 44 is a block diagram illustrating the configuration of the multiplication coefficient generating section shown in FIG.
45 is a block diagram illustrating the detailed configuration of the multiplier shown in FIG. 7, and FIG. 46 is a filter circuit shown in FIG. 7. FIG. 47 is a circuit block diagram explaining the detailed configuration of the filter control signal generator shown in FIG. 7, and FIG. 48 is a circuit block diagram explaining the detailed configuration of the filter control signal generator shown in FIG. FIG. 49 is a correlation diagram illustrating the relationship between the control signal output from the generator and filter processing, FIG. 49 is a block diagram illustrating smoothing filter processing in the image reading device according to the present invention, and FIG. 51st block diagram illustrating the configuration of the gamma conversion unit shown in FIG.
52 is a characteristic diagram illustrating the gamma conversion characteristic in the image reading device according to the present invention, FIG. 52 is a block diagram illustrating the configuration of the gamma switching signal generating section shown in FIG. FIG. 54 is a diagram explaining the functional processing of the gamma switching signal generation section shown in the figure.
55 is a circuit block diagram explaining the detailed configuration of the WM modulation section, FIG. 55 is a timing chart explaining the operation of FIG. 54, and FIG. 56 is a circuit block diagram explaining the configuration of the screen switching signal generation section shown in FIG. 7. FIG. 57 is a plan view illustrating the configuration of an operating section in an image reading device showing another embodiment of the present invention. In the figure, 103 is a light amount signal-density signal converter, 104 is a black extraction unit, 105 is a masking processing unit, 112, 113 are line delay memories, 114° and 115 are multipliers, 116 is an adder, 117 is a filter circuit, 210 is an image scanner section, 211 is an image processing unit, 401 is a control section, 402 is a color signal processing section, 403 is a feature extraction section, 404
is a color processing control signal generation section. Fig. 3 H5YNC Fig. 8 Fig. 06 Fig. 'iii Length (mouth m) Fig. 1 Fig. 12 Fig. 5 Fig. 3 Fig. 4 Fig. 20 Fig. 5YNC in CL Fig. 1 Fig. Color signal 1 Fig. 2 Figure 24 Figure 6 Flash Figure 7 Figure 7 Figure 30 (b) Input address (A7-AO) Figure 0 (d) - Personnel address 31 Figure 2 Figure 4 Figure 254 Figure 5 Figure m-'' Figure 35 Figure 41 (a) (b) Q 0 Yuyu (c) Su [F]■ Figure 42 Figure 5 Figure 3 Crystal diagram 48 Figure 9 Figure 46 Figure 0 18a Figure 51 Input Figure 2 Figure 3 Figure 56

Claims (4)

[Claims] (1) An image sensor that reads the original image while converting the reflected image obtained by exposure scanning the original into an electrical signal;
In an image reading device including an image processing unit that processes image signals output from the image sensor,
mode setting means for setting and inputting a plurality of original reading modes corresponding to the type of the original; and a mode setting means for setting and inputting a plurality of original reading modes corresponding to the type of the original, and comparing density levels of neighboring pixels including the pixel of interest of the image signal with a determination density level to obtain a predetermined density level signal. a conversion means for converting, a judgment level variable means for varying the judgment density level based on the document reading mode set by the mode setting means, and a conversion means according to the judgment density level variably set by the judgment level variable means. An image reading apparatus comprising: a character area separation processing means for separating a character edge region from the image signal based on the predetermined density level signal output from the image reading apparatus. (2) Claim (1) characterized in that the determination level variable means variably sets the determination density level so that the detection ratio of high-density level pixels to low-density pixels is given priority to low-density pixels.
The image reading device described. (3) The converting means converts the density levels of neighboring pixels including the pixel of interest of the image signal into a predetermined density level signal based on a conversion lookup table. Image reading device. (4) The converting means converts the density levels of neighboring pixels including the pixel of interest of the color image signal into a predetermined density level signal based on a conversion lookup table. Image reading device.