JP2003259115A - Image processor, image reader, image forming device, and color copying machine - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processor, an image reader, an image forming device, and a color copying machine which do not emphasize image data in a specified period. <P>SOLUTION: A filter 321 computes coefficient groups A and B (Fig. 6) and outputs (minimum value of arithmetic result/16 + pixel of interest) to edge extraction 322 and outputs (maximum value of arithmetic result/16 + pixel of interest) to white area extraction 323. The coefficient groups A and B are Laplacian, so they are divided by a specified coefficient (16) and added to a pixel of interest for correction. The minimum value of the arithmetic result is outputted to the edge extraction 322 since a white level can possibly not be extracted sufficiently when the structure of a character is in a hair line shape. The maximum value of the arithmetic result is outputted to the white area extraction 323 to make it easy to extract a pattern when the pattern is in the structure of a hair line pattern (e.g. output of a copying machine). <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus, an image reading apparatus, an image forming apparatus and a color copying apparatus which detect characters in image data and perform appropriate image processing.

[0002]

2. Description of the Related Art In recent years, color printers capable of digitally processing color image data and outputting in color, and color copying in which color originals are color-separated and electrically read, and the acquired color image data is copied onto paper. Image processing devices capable of color printing, such as printing machines, are widely used. With the spread of image processing apparatuses such as these color printers and color copying machines, it is required to improve the output quality of color images. For example, if the original is a single color of black, the signals of yellow, magenta, cyan, and black are generated as the signals for reproducing black. Blurring may occur and the image may appear blurry. In order to prevent such a reduction in output quality, a method has been proposed in which the characteristics of black characters and black lines in an image are utilized by switching filters for image correction and image quality adjustment. As for the switching of the filter, there are a method of switching by a character or a photograph and a method of controlling the coefficient of the filter by the edge amount (Laplacian). It is necessary to set the filter coefficient so that the line pattern is not detected as a character edge and the line pattern is not emphasized. Further, there is also a method of performing correction by switching a plurality of filter coefficients under specific conditions.

[0003]

However, due to the MTF (frequency characteristic) of the lens of the image reading unit,
The scanned image data itself may be blurred,
When the image data itself is unraveled in this way, it is common to apply a filter that emphasizes the image data and then reproduce the image data. Will be done. In particular,
If the line pattern used in the output of the copying machine is emphasized, it will be difficult to distinguish it from a character. Further, in the image correction performed by switching the filter, 600 dp
i reading, 400 dpi line pattern and 600d
It is difficult to set a filter coefficient that does not emphasize both the line patterns of pi, and when using a plurality of filter coefficients, switching may be erroneously determined. Therefore, a first object of the present invention is to perform an operation with a plurality of types of filters (coefficients) so as not to emphasize image data having a specific frequency characteristic, an image processing apparatus, an image reading apparatus, an image forming apparatus, and A color copying apparatus is provided. A second object of the present invention is to carry out a filter operation with a filter coefficient that does not emphasize a 400-dpi line cycle and a filter coefficient that does not emphasize a 600-dpi line cycle with a plurality of filters, and perform image processing to switch the filter coefficient according to the result. To provide an apparatus, an image reading apparatus, an image forming apparatus, and a color copying apparatus. A third object of the present invention is to select a filter that does not emphasize black in the line pattern in the edge extraction means and a filter that does not emphasize white in the white background area detection to select the line pattern as a character. An image processing apparatus, an image reading apparatus, an image forming apparatus, and a color copying apparatus can be provided.

[0004]

According to a first aspect of the present invention, image data receiving means for receiving image data, correction means for correcting the image data received by the image data receiving means by a plurality of filters, and Filter operation means for performing filter operation of the plurality of filters, filter coefficient storage means for storing filter coefficients of the plurality of filters, and when the correction means performs correction based on the operation result by the filter operation means Since the filter coefficient determining means for determining the filter coefficient and the edge extracting means for performing the edge extracting process of the image data based on the correction result by the correcting means are provided, the first and second objects are achieved. To achieve. According to the invention of claim 2,
In the invention according to claim 1, the filter coefficient determination means determines the filter coefficient based on the minimum value of the image density of the calculation result by the filter calculation means, thereby achieving the first and second objects. To achieve.

According to the third aspect of the invention, the image data receiving means for receiving the image data, the correcting means for correcting the image data received by the image data receiving means by a plurality of filters, and the filter of the plurality of filters. A filter calculation unit that performs a calculation, a filter coefficient storage unit that stores the filter coefficients of the plurality of filters, and a filter that determines a filter coefficient when the correction unit makes a correction based on the calculation result of the filter calculation unit. The third object is achieved by including the coefficient determining means and the area extracting means for performing the white background area extracting process of the image data based on the correction result by the correcting means. Claim 4
According to a third aspect of the invention, in the third aspect of the invention, the filter coefficient determining means determines the filter coefficient based on the maximum value of the image density of the calculation result by the filter calculating means, thereby the third coefficient. Achieve the purpose.

According to a fifth aspect of the invention, the image processing apparatus according to any one of the first, second, third, and fourth aspects separates the original image into colors and reads the image data to obtain image data. By further comprising color image data reading means for generating and transmitting to the image data receiving means, the first, second and third objects are achieved. According to a sixth aspect of the present invention, a color image formation is performed in the image processing apparatus according to any one of the first, second, third and fourth aspects, in which the image data is formed as a color image on a sheet. By further comprising means, the first and second
And to achieve the third object. According to a seventh aspect of the present invention, the image processing apparatus according to any one of the first, second, third, and fourth aspects performs color separation of an original image and reads the original image to generate image data. By further including color image data reading means for transmitting the image data to the image data receiving means, and color image forming means for forming a color image of the image data on a sheet, whereby the first, second, and the first Achieve the purpose of 3. According to an eighth aspect of the invention, the image forming apparatus according to the sixth aspect or the color copying apparatus according to the seventh aspect further includes print instruction receiving means for receiving a print instruction from an external device, and the image data receiving means is The image data is received from the external device, and the color image forming unit executes the color image formation by the color image forming unit based on the print instruction received by the print instruction receiving unit. 1, achieve the second and third objects.

[0007]

BEST MODE FOR CARRYING OUT THE INVENTION Preferred embodiments of the present invention will be described in detail below with reference to FIGS. 1 to 30. FIG. 1 is a diagram showing a schematic configuration of an image processing apparatus according to this embodiment. In FIG. 1, a digital full-color copying machine (hereinafter, simply referred to as a color copying machine) is shown as an example of the image processing apparatus. A color image reading device (hereinafter referred to as a scanner) 200 forms an image of an original 180 on a contact glass 202 on a color sensor 207 via an illumination lamp 205, mirror groups 204A, 204B, 204C, etc., and a lens 206. The color image information of the original is read for each color-separated light of blue (hereinafter referred to as B), green (hereinafter referred to as G) and red (hereinafter referred to as R), and converted into an electrical image signal. In this embodiment, the color sensor 207 is composed of a 3-line CCD (photoelectric conversion element) sensor, and reads B, G, and R images for each color. Based on the B, G, and R color-separated image signal intensity levels obtained by the scanner 200, color conversion processing is performed by an image processing unit (not shown), and black (hereinafter referred to as Bk), cyan (hereinafter referred to as C) ), Magenta (hereinafter referred to as M)
Color image data including recording color information of yellow and yellow (hereinafter, Y) is obtained.

Using the color image data, a color image recording device (hereinafter referred to as a color printer) 4 described below is used.
00, Bk, C, M, and Y images are formed on the intermediate transfer belt in an overlapping manner, and then transferred to a transfer paper. The scanner 200 receives a scanner start signal in time with the operation of the color printer 400, and an illumination / mirror optical system including an illumination lamp 205 and mirror groups 204A, 204B, and 204C scans the document in the direction of the left arrow,
Image data of one color is obtained for each scanning. Then, each time, while being visualized by the color printer 400, these are superposed on the intermediate transfer belt to form a full-color image of four colors.

A writing optical unit 401 as an exposing means of the color printer 400 converts color image data from the scanner 200 into an optical signal and performs optical writing corresponding to an original image, and an electrostatic latent image is formed on the photosensitive drum 414. Form an image. The writing optical unit 401 includes a laser light emitter 441, a light emission drive controller (not shown) that drives the laser light emitter 441, a polygon mirror 443, a rotation motor 444 that rotationally drives the polygon mirror 443, an fθ lens 442, and a reflection mirror 4.
It is composed of 46 and the like. The photoconductor drum 414 rotates in a counterclockwise direction as shown by an arrow in the figure. A potential sensor 414D that detects a potential, a selected developing device of the revolver developing device 420, a development density pattern detector 414P, an intermediate transfer belt 415, and the like are arranged.

The revolver developing device 420 includes a Bk developing device 420K, a C developing device 420C, an M developing device 420M, a Y developing device 420Y, and a revolver for rotating each developing device in a counterclockwise direction as shown by an arrow in the figure. The rotary drive unit (not shown) is included. These developing devices rotate the developing sleeves 420KS and 42KS by bringing the brush of the developer into contact with the surface of the photoconductor drum 414 to visualize the electrostatic latent image.
0CS, 420MS, 420YS, and a developing paddle that rotates for assembling and stirring the developer. In the standby state, the revolver developing device 420 is set to Bk.
It is set to the position where development is performed with the developing device 420K,
When the copy operation is started, the scanner 200 starts reading the Bk image data at a predetermined timing,
Based on this image data, optical writing and latent image formation by laser light starts. Hereinafter, the electrostatic latent image based on the Bk image data is referred to as a Bk latent image. Similarly, each of the C, M, and Y image data is referred to as a C latent image, an M latent image, and a Y latent image. This B
In order to enable development from the tip of the k latent image, the Bk developing device 42
Before the latent image front end reaches the development position of 0K, the developing sleeve 420KS starts rotating to develop the Bk latent image with Bk toner. After that, the developing operation of the Bk latent image area is continued, but when the trailing edge of the latent image passes the Bk latent image position, the development of the next color from the developing position of the Bk developing device 420K is promptly performed. The revolver developing device 420 is driven and rotated to the position. This rotation operation is completed at least before the leading edge of the latent image by the next color image data arrives.

When the image forming cycle is started, the photosensitive drum 414 rotates counterclockwise as indicated by an arrow, and the intermediate transfer belt 415 rotates clockwise by a drive motor (not shown). Move. Intermediate transfer belt 41
With the rotation of 5, the Bk toner image formation, the C toner image formation, the M toner image formation, and the Y toner image formation are sequentially performed, and finally, on Bk, C, M, and Y in this order on intermediate transfer belt 415. And a toner image is formed thereon. The Bk image is formed as follows. That is, the charger 41
9 corona discharges the photosensitive drum 414 to a negative charge and uniformly charges the photosensitive drum 414 to about −700V. Then, the laser light emitter 441 performs raster exposure based on the Bk signal.
When the raster image is exposed in this manner, in the exposed portion of the photosensitive drum 414 that is initially uniformly charged, the charge proportional to the amount of exposure light disappears and an electrostatic latent image is formed. The toner in the revolver developing device 420 is negatively charged by stirring with the ferrite carrier, and the Bk developing sleeve 4 of the revolver developing device 420 is charged.
20 KS is biased by a power supply circuit (not shown) with respect to the metal base layer of the photoconductor drum 414 to a potential in which a negative DC potential and an alternating current are superimposed. As a result, the toner does not adhere to the portion of the photoconductor drum 414 where the electric charge remains, and the Bk toner is adsorbed to the portion having no electric charge, that is, the exposed portion, and Bk visible similar to the latent image. An image is formed. The intermediate transfer belt 415 is driven by the drive roller 415.
D, a transfer opposing roller 415T, a cleaning opposing roller 415C, and a driven roller group are stretched and driven to rotate by a drive motor (not shown). Photoconductor drum 4
Bk toner image formed on the photosensitive drum 414
The image is transferred by a belt transfer corona discharger (hereinafter, referred to as a belt transfer unit) 416 onto the surface of the intermediate transfer belt 415 which is driven at a constant speed in contact with. Hereinafter, the toner image transfer from the photoconductor drum 414 to the intermediate transfer belt 415 will be referred to as belt transfer. A small amount of untransferred residual toner on the photoconductor drum 414 is cleaned by the photoconductor cleaning unit 421 in preparation for reuse of the photoconductor drum 414. The toner collected here is stored in an exhaust toner tank (not shown) via a collection pipe.

On the intermediate transfer belt 415, the Bk, C, M, and Y toner images sequentially formed on the photoconductor drum 414 are sequentially aligned on the same surface to form a belt transfer image of four-color superposition. After being formed, it is collectively transferred onto a transfer paper by a corona discharge transfer device. By the way, the photosensitive drum 414
On the side, the process of forming a Bk image is followed by the process of forming a C image. At a predetermined timing, the reading of the C image data by the scanner 200 starts, and the C latent image of the C latent image is written by the laser light writing by the image data. Form. The C developing device 420C rotates the revolver developing device 420 with respect to the developing position after the trailing end of the previous Bk latent image has passed and before the leading end of the C latent image reaches. C
Develop the latent image with C toner. After that, the development of the C latent image area is continued, but at the time when the trailing edge of the latent image passes, the revolver developing device 420 is driven as in the case of the previous Bk developing device to send out the C developing device 420C, and M developing device 420
Position M at the development position. This operation is also the next M
This is performed before the leading edge of the latent image reaches the developing section. The steps of forming the M and Y images are the same as the steps of reading the image data, forming the latent image, and developing according to the steps of the Bk image and the C image described above.

The belt cleaning device 415U comprises an inlet seal, a rubber blade, a discharge coil, and a contact / separation mechanism for the inlet seal and the rubber blade. 1
During the belt transfer of the images of the second, third, and fourth colors after the Bk image of the color is transferred onto the belt, the entrance seal, the rubber blade, etc. are separated from the surface of the intermediate transfer belt by the blade contact / separation mechanism. The paper transfer corona discharger (hereinafter, referred to as a paper transfer device) 417 transfers AC + DC or DC component to the transfer paper and the intermediate transfer by the corona discharge method in order to transfer the superimposed toner image on the intermediate transfer belt 415 to the transfer paper. It is applied to the belt 415.

Transfer paper cassettes 482 in the paper feed bank contain transfer papers of various sizes, and the paper feed roller 4 is loaded from the cassette that stores papers of a specified size.
The sheet is fed and conveyed in the direction of the registration roller pair 418R by 83. Reference numeral 412B2 indicates a paper feed tray for manually inserting OHP paper or thick paper. At the time when image formation is started, the transfer paper is fed from any of the transfer paper cassettes 482 described above, and the registration roller pair 41
Standby at 8R nip. And the paper transfer device 4
When the front end of the toner image on the intermediate transfer belt 415 approaches 17 the registration roller pair 418R is driven so that the front end of the transfer paper coincides with the front end of this image, and the paper and image are aligned. In this way, the transfer paper is superposed on the color superposed image on the intermediate transfer belt and passes over the paper transfer device 417 which is connected to the positive potential. At this time, the transfer paper is positively charged by the corona discharge current, and most of the toner image is transferred onto the transfer paper. Then, the paper transfer device 417
When passing through a separation static eliminator by a static eliminator brush (not shown) disposed on the left side of the sheet, the transfer sheet is destaticized, separated from the intermediate transfer belt 415, and transferred to the paper transport belt 422. The transfer paper on which the four-color superposed toner images are collectively transferred from the surface of the intermediate transfer belt 415 is fixed to the fixing device 423 by the paper transport belt 422.
The toner image is melted and fixed at the nip portion of the fixing roller 423A and the pressure roller 423B, which are controlled to a predetermined temperature, is sent to the outside of the main body by the discharge roll pair 424, and is stacked face up on a copy tray (not shown). It

Incidentally, the photosensitive drum 414 after the belt transfer.
Is cleaned on its surface by a photoconductor cleaning unit 421 including a brush roller and a rubber blade,
Further, the static elimination lamp 414M uniformly eliminates the static electricity. Also,
Intermediate transfer belt 415 after the toner image is transferred onto the transfer paper
Again presses the blade with the blade contacting / separating mechanism of the belt cleaning device 415U to clean the surface. In the case of repeat copy, the operation of the scanner and the image formation on the photosensitive member proceed to the first color image process of the second sheet at a predetermined timing after the fourth color image process of the first sheet. The intermediate transfer belt 415 transfers the second Bk toner image to the area where the surface is cleaned by the belt cleaning device 415U, following the batch transfer process of the first four-color superimposed image onto the transfer paper. To be done. After that, the same operation as the first sheet is performed.

The color copying machine shown in FIG. 1 is connected to a LAN (Local) from a host device such as a personal computer.
Area Network) or parallel I / F
When print data is given through the (interface), it can be printed out (image output) by the color printer 400, and the image data read by the scanner 200 can be transmitted to a remote facsimile and the received image data can also be printed out. It is a color copier with multiple functions. This color copier is a private branch exchange PBX.
Connected to the public telephone network via
Facsimile communication and communication with the management server of the service center are possible.

FIG. 2 is a block diagram showing a schematic configuration of an electric system of the color copying machine shown in FIG. In FIG. 2, the control device of the color copying machine is shown centering on the main controller 10. The main controller 10 controls the entire color copying machine. The main controller 10 includes an operation / display board OPB for displaying information to the operator and controlling input of function settings from the operator, an editor 15, a scanner 200, and an optional ADF (Au).
to Document Feeder), control for writing a document image in the image memory, control for forming an image from the image memory, and the like.
2. Control of the printer controller 16, the image processing unit (hereinafter referred to as IPU) 300, and the image forming engine in the color printer 400 that performs charging, exposure, development, paper feeding, transfer, fixing, and transfer paper transportation. A distributed control device such as an engine controller 13 for performing the operation is connected. Each distributed control device and main controller 10
The machine status and operation commands are exchanged as necessary. Further, the main motor and various clutches required for paper conveyance and the like are also connected to a driver (not shown) in the main controller 10.

In the color printer 400, charging of the photosensitive drum 414, charging of the photosensitive drum 414,
There are electric circuits and control circuits for driving mechanical elements that perform image exposure, development, transfer, fixing and paper discharge by a laser writing unit, and various sensors. The printer controller 16 analyzes an image from an external device such as a personal computer and a command to print, develops a bitmap into image data as a printable state, and drives the color printer 400 via the main controller 10 to drive the image. Print out the data. In order to receive and operate images and commands via LAN and parallel I / F, LAN control 19 and parallel I / F
There is an F control 18. FAX controller 17
When there is a facsimile transmission instruction, the scanner 200 and the IPU 30 are transmitted via the main controller 10.
0 is driven to read the image of the original, and the image data is sent to the facsimile communication line via the communication control 20 and the private branch exchange PBX. When receiving a facsimile call from the communication line and receiving image data, the color printer 400 is driven through the main controller 10 to print out the image data.

FIG. 3 shows an image processing unit (IPU) 30.
It is a block diagram showing the composition of 0. The R, G, B image data generated by the scanner 200 is stored in the interface 3
It is given to the IPU 300 via 51. When the BR unit 355 instructs recording of B or R single color, R, G, B image data is selected and assembled.
Description of the image recording process in this mode is omitted. IPU3
The R, G, B image data given to 00 is converted from the reflectance data (R, G, B data) to the density data (R, G, B data) by the RGBγ correction 310. In the document recognition 320, based on the density R, G, B data, the image area to which these data are addressed is a character edge area (edge area of a character or a line drawing), a halftone dot area, a low frequency halftone dot area,
It is determined whether the area is a picture area (a picture or picture area & not a character area & not a halftone area), and C /
The P signal and the B / C signal are given to the main controller 10 via the RGB filter 330 and the interface 353. Here, the C / P signal is a 2-bit signal, "3" is a low frequency halftone dot area, "2" is a halftone dot area, "1" is a character edge area, and "0" is a pattern. It shall indicate the area. The B / C signal is a 1-bit signal, and “H” (= 1) indicates an achromatic area and “L” (= 0) indicates a chromatic area.

FIG. 4 is a block diagram showing the functional arrangement of the document recognition 320 shown in FIG. The document recognition 320 performs character edge detection, pattern detection, and chromatic / achromatic detection to generate a C / P signal representing a character edge area or a pattern area and a B / C signal representing a chromatic area / achromatic area. Occur. The document recognition 320 is roughly divided into a filter 321, an edge extraction 322, a white area extraction 323, a halftone dot extraction 324, a color determination 325, and a comprehensive determination 326. Here, a case where the reading density of the scanner 200 is about 600 dpi will be described as an example.

The filter 321 corrects the G image data generated by the scanner 200 mainly for extracting the edge of the character. Here, since the data read by the scanner 200 is blurred due to the characteristics of the lens, an edge emphasis filter is applied. However, here, it is necessary to simply emphasize the image edge on the original and not to emphasize the line pattern that is widely used in copying machines for gradation expression. If the line pattern is emphasized, the pattern (gradation expression area by the line pattern) may be extracted as an edge, and it may be erroneously determined to be a character edge, so it is necessary not to emphasize it. . Also, 600 read at 600 dpi
As shown in FIG. 5, which shows a parallel line pattern of 600 dpi or 400 dpi, a parallel line pattern A of 600 dpi and 400 dpi
Since the line pattern B of No. 2 has a different repetition period, it is difficult to prevent the line pattern B from being emphasized by the same filter coefficient. Therefore, either the maximum value or the minimum value of the calculation results of the two coefficients is used according to the subsequent feature amount detection (edge extraction, white area extraction). In FIG. 5, the sum of the width of one white block in the main scanning direction x and the width of one black block in contact with it is the line pitch (constant width: a predetermined number of pixels), that is, the line cycle, and the low density intermediate In tones, the white block width widens and the black block width narrows. The white block width becomes narrower and the black block width becomes wider as the density becomes higher.

In the present embodiment, the pixel matrix of the filter 321 is set to the number of pixels in the main scanning direction x 7 × the sub scanning direction y.
Number of pixels 5 (mechanical scanning direction of scanner 200)
As shown in the block of the filter 321, the weighting factors a1 to a7 and b1 to b are assigned to the respective pixels.
7, c1 to c7, d1 to d7, and e1 to e7, there are two sets of coefficient groups (coefficient matrices) A and B. Figure 6
FIG. 4 is a diagram showing an example of coefficient groups. The coefficient group A shown in FIG. 6 is a filter processing coefficient that suppresses the emphasis of the line pattern A of 600 dpi in FIG. 5 and further emphasizes the edge of the character. On the other hand, the coefficient group B shown in FIG. 6 is a filter processing coefficient that suppresses the emphasis of the 400 dpi line pattern B of FIG. 5 and emphasizes the edge of the character.

The coefficient groups A and B shown in FIG.
Is arranged in the horizontal scanning direction x and the vertical direction is arranged in the sub scanning direction y. The coefficients of the first row in the coefficient groups A and B are the coefficients a1 to a7 of the first row of the coefficient matrix of the block of the filter 321 in FIG. 20 ”is the third row c1 to c7 of the coefficient matrix of the block of the filter 321.
Is the coefficient of the pixel in the center of, i.e., the coefficient c4 of the pixel of interest. The product obtained by multiplying each coefficient of the coefficient matrix by the value represented by the image data of the pixel addressed to it (total 7 × 5 = 35
The sum (product sum value) of the individual pixels is given to the edge extraction 322 and the white area extraction 323 as an image data value obtained by processing the pixel of interest (the pixel addressed to c4) by the filter 321. Here, the pixel of interest is a pixel that is currently a processing target, and is sequentially updated to have a different position in the x direction and the y direction.

The coefficient group A is 600 dp shown in FIG.
Negative coefficients (coefficients with small values) are distributed at the parallel pitch of the parallel pattern A of i, 0 (coefficients with slightly large values) are distributed between them, and the pixel of interest is focused on the pixel of interest for edge enhancement. 20 (very large coefficient) is addressed. As a result, when the image data (pixel of interest) is the black / white edge of the area of the line pattern A, the weighted average value (product sum value) derived to it is a character edge that is not the line pattern A. It is much lower than when it is. on the other hand,
Coefficient group B has negative coefficients (coefficients with small values) distributed at the line pitch of line pattern B of 400 dpi shown in FIG. 5, and 0s (coefficients with slightly large values) are distributed between them, and 20 (very large coefficient) is assigned to the pixel of interest for edge enhancement. As a result, when the image data (pixel of interest) is the black / white edge of the area of the line pattern B, the weighted average value (product sum value) derived to it is a character edge that is not the line pattern B. It is much lower than when it is.

In the filter 321, the coefficient group A and the coefficient group B are calculated, and the edge extraction 322 is performed.
Is output to (minimum value of calculation result / 16 + pixel of interest), and white area extraction 323 displays (maximum value of calculation result / 16)
+ (Target pixel) is output. Since the coefficient groups A and B are Laplacian, they are divided by a predetermined coefficient (16) and added to the pixel of interest for correction. The minimum value of the calculation result is output to the edge extraction 322 in order to avoid that the white level may not be sufficiently extracted when the character structure has a parallel line shape. The maximum value of the calculation result is output to the white region extraction 323 when the pattern has a line pattern structure (for example, output of a copying machine), the maximum value is output so that the pattern can be extracted more easily. This makes it easier to pick up edges in edge extraction, and easier to pick up as a pattern in white area extraction. In the present embodiment, two coefficients are described as an example, but the present invention is not limited to this, and the same effect can be obtained with three or more coefficients.

Next, the edge extraction 322 will be described. As shown in FIG. 4, the edge extraction 322 is a ternarization 3
22a, black pixel continuous detection 322b, white pixel continuous detection 32
2c, neighboring pixel detection 322d, and isolated point removal 322e
It consists of Note that FIG. 4 shows an example in which G image data is referred to in the edge processing, but the present invention is not limited to this G data and may be luminance data, as long as it is a signal expressing dark and light. And In the character area, there are many high-level density pixels and low-level density pixels (hereinafter referred to as black pixels and white pixels), and at the edge portion, these black pixels and white pixels are continuous. The edge extraction 322 detects a character edge based on the continuity of each of such black pixels and white pixels.

First, the ternarization 322a will be described.
In the ternarization 322a, the two types of thresholds TH1 and TH2 are used to ternarize the G image data (input data of the edge extraction 322) that has been filtered by the filter 321 for character edge enhancement. The thresholds TH1 and TH2 are set as TH1 = 20 and TH2 = 80, for example, when the image data represents 256 gradations from 0 to 255 (0 = white). In the ternarization 322a, when the input data <TH1, the pixel to which the data is addressed is a white pixel and TH1
When ≦ input data <TH2, the halftone pixel and T
When H2 ≦ input data, the input data is converted into ternary data which is expressed as black pixels.

The black pixel continuous detection 322b and the white pixel continuous detection 322c detect the continuous black pixels and the continuous white pixels by pattern matching based on the ternary data. In this embodiment, for this pattern matching, the patterns BPa to BPd and WPa to WP of the 3 × 3 pixel matrix shown in FIG. 7 are used.
d is used. In the pattern shown in FIG. 7, black circles indicate the above black pixels, white circles indicate the above white pixels, and blank pixels without any circles indicate black pixels,
It indicates that it does not matter whether it is a halftone pixel or a white pixel. The pixel at the center of the 3 × 3 pixel matrix is the target pixel. Black pixel continuous detection 322b
When the distribution of the contents of the ternary data matches any of the black pixel distribution patterns BPa to BPd shown in FIG. Give to. Similarly, the white pixel continuous detection 322c is performed by the white pixel distribution pattern WPa shown in FIG.
When any one of ˜WPd is matched, the pixel of interest at that time is set as a “white continuous pixel”, and data representing it is given to the pixel of interest.

The next adjacent pixel detection 322d determines whether or not there is a black continuous pixel or a white continuous pixel in the vicinity of the target pixel in the adjacent pixel detection 322d, based on the detection results of the black pixel continuous detection 322b and the white pixel continuous detection 322c. By checking whether the pixel of interest is in the edge area or the non-edge area, it is determined. That is, in the present embodiment, when a block of a 5 × 5 pixel matrix has at least one black continuous pixel and one white continuous pixel, the block is determined to be an edge region, and if not, , Determine that the block is a non-edge region. Further, since character edges exist continuously, isolated point removal 3
22e corrects an isolated edge into a non-edge region. Then, “1” is set for the pixel determined to be the edge region.
The edge signal of (edge region) is output, and the edge signal of “0” (non-edge region) is output to the pixel determined to be the non-edge region.

Next, the white area extraction 323 will be described. FIG. 8 is a block diagram showing the configuration of the white area extraction 323. First, the binarization 323a will be described. Two
The binarization 323a is performed by the image density data (G
2) the edge emphasis output of (image data) with a threshold value thwsb
The digitized white pattern matching 323d (see FIG.
2 for generation of white data referred to in step 7) of 1.
Generate a binarized white judgment signal. The edge emphasis output is
In this embodiment, there are 256 gradations from 0 to 255, and
Is white with no density, and an example of the threshold value thwsb is 50
If the value of the edge emphasis output is smaller than thwsb = 50, the binarization 323a is determined to be "binarization white",
A binarized white determination signal "1" is generated. When the value of the edge emphasis output is thwsb = 50 or more, the binarized white determination signal "0" is generated.

The RGB white extraction 323b is 1. ) RGB white background detection, 2. 2.) Color background detection and 3. ) Valley white pixel detection is performed to determine whether the image data is a white area. 1. ) RGB white background detection In the RGB white background detection 323b, the white background separation operation is activated by detecting a white background area in the R, G, and B image data. That is, the process of separating the white background is activated. Specifically, as shown in the pattern WBP of FIG.
If all of the R, G, B image data of the 3 × 3 pixel matrix is smaller than the threshold value thwss, the pixel of interest (3 × 3
It is determined that the central pixel of the pixel matrix is a white region, and white pattern matching 323d (step 3 in FIG. 11 described later).
(White background determination signal) referred to by is set to active "1".
This is to detect whether there is a white pixel area having a certain extent of spread. Note that each of the R, G, and B image data also has 256 gradations from 0 to 255 in the present embodiment, 0 is a base level with no density, and the threshold value thw
ss <thwsb, and an example of thwss is 40
And all of the R, G, B image data is thwss
If it is less than 40, it is determined that the white background is present and a white background determination signal "1" is generated. When any of the R, G, and B image data is thwss = 40 or more, the white background determination signal “0” is generated.

2. ) Color gamut detection In order to prevent a light color from being determined as a white background, a color ground is detected. A. First, the code of each pixel of the 5 × 5 pixel matrix centered on the pixel of interest is set to the pattern MPp in FIG.
, The center pixel c3 (M
If the RGB difference (difference between the maximum value and the minimum value of the R, G, and B image data for one pixel) of Ca to MCd) is larger than the threshold value thc, the color pixel determination signal a is set to “1” (color). Pixel), and when the threshold value is less than or equal to thc, “0” (black and white pixel)
And B. Peripheral pixel group on one side of the pixel of interest Δ (MCa-
If the R, G, B image data of any of the pixels (in the MCd) are all less than or equal to the threshold value thwc, the white determination signal b on one side is set to "1" (white pixel), and if it exceeds the threshold value thwc, " 0 ”(non-white pixel). Here, the threshold thw
It is assumed that c is 20 as an example. C. Peripheral pixel group on the other side of the target pixel □ (MCa in FIG.
When the R, G, B image data of any of the pixels (in the MCd) is less than or equal to the threshold value thwc, the white determination signal c on the other side is set to "1" (white pixel), and when the threshold value thwc is exceeded. It is set to “0” (non-white pixel). D. In any of the patterns MCa to MCd in FIG. 10, a AND (exclusive NOR of b and c) =
If “1” is satisfied, that is, if a = “1” (the target pixel is a color pixel) and b and c match (white pixels on both sides of the target pixel, or non-white pixels on both sides), The color background determination signal d addressed to the pixel is set to "1" (color background).
This color background determination signal d is the white pattern matching 323d.
(Step 6 of FIG. 11 described later) is referred to.

Process of Pattern Matching A. ~
D. The reason for doing so is that when the surroundings of black characters are slightly colored due to the deviation of the RGB reading position, they are not picked up as colors. At the colored position around the black character, the exclusive NOR of b and c is “0” (one of both sides of the target pixel is a white pixel, the other is a non-white pixel), and in this case, the color-ground determination signal d = “0”. "(Non-colored background). In addition, when the pixel of interest surrounded by a white background is a color pixel, the color-ground determination signal d = “1” (color-ground) is set, and a light-color pixel is detected as a color-ground even when a line is complicated. You can That is, when a line is complicated, an originally white part cannot be completely read as white, but the above-mentioned process A. Therefore, if the RGB difference is small, it is not judged as a color pixel.
The threshold value thwc is set to be stricter than the white background to be viewed in terms of density (for example, thwss = 40 and thwsb = 50, but thwc = 20), and the process B. ~ D. Therefore, it is possible to accurately check whether or not it is a white background, and accurately detect a light color pixel as a color background.

3. ) Valley white pixel detection In the valley white pixel detection, valley white pixels in a small white area that cannot be detected by the above-described RGB white background detection are detected based on the 5 × 5 pixel matrix distributions RDPa and RDPb of the G image data shown in FIG. To do. Specifically, based on the 5 × 5 pixel matrix distribution RDPa, miny = min (G [1]
[2], G [1] [3], G [1] [4], G [5]
[2], G [5] [3], G [5] [4]) are calculated. That is, the 5 × 5 pixel matrix distribution R shown in FIG.
The minimum density miny in the pixel group marked with a black circle of DPa is extracted. Then, maxy = max (G [3]
[2], G [3] [3], G [3] [4]) are calculated. That is, the 5 × 5 pixel matrix distribution R shown in FIG.
The maximum density maxy in the pixel group marked with a white circle of DPa is extracted.

Next, mint = min (G [2]
[1], G [3] [1], G [4] [1], G [2]
[5], G [3] [5], G [4] [5]) are calculated. That is, the lowest density mint in the pixel group marked with black circles of another 5 × 5 pixel matrix distribution RDPb shown in FIG. 9 is extracted. Then, maxt = max (G
[2] [3], G [3] [3], G [4] [3]) are calculated. That is, the maximum density max in the pixel group with white circles in the 5 × 5 pixel matrix distribution RDPb shown in FIG.
Extract t. Here, min () is a function for detecting the minimum value. Further, max () is a function for detecting the maximum value. Next, OUT = ((miny-max
y)> 0) # ((mint-maxt)> 0) is calculated. That is, (miny = maxy) and (mint-
maxt) is a positive and larger value as a valley detection value OUT, and if the value of this OUT is equal to or larger than a threshold value, the pixel of interest (center pixel of RDPa or RDPb) is detected as a valley white pixel. . In this way, the valley state of the image is detected, and 1. ) In the RGB white background detection, a part that is difficult to detect is supplemented.

Next, the white judgment 323c will be described.
Here, the state variables MS and SS [I] used for white determination are updated. FIG. 11 shows a state variable MS used for white determination,
7 is a flowchart showing a processing procedure for updating SS [I]. Here, the state variable MS is for the pixel of the processing target line (target line), and the state variable SS [I] is for the pixel of one line before the processing target line (processed line). It is 4-bit white background information indicating the degree of whiteness, and is generated by the processing procedure of FIG. 11. The maximum value represented by the state variable MS and the state variable SS [I] is set to 15, which means the whitest degree, and the minimum value is 0. That is, the state variable MS
The state variable SS [I] is data indicating the degree of white, and the larger the value it represents, the stronger the white.
At the start of the copying operation, the state variable MS and the state variable SS
Both [I] are initialized to 0.

First, the state variable one line before the target pixel to be processed, that is, the state variable of the pixel one pixel before on the same line as the white background information SS [I] and the target pixel (preceding pixel: processed pixel) That is, when the white background information SS [I] of one line before is compared with the white background information MS (step 1) is larger (step 1; Y), it is set as the temporary white background information MS of the target pixel ( Step 2). White background information SS
When [I] is not larger (step 1; N), the state variable MS of the preceding pixel is set as the temporary white background information MS of the pixel of interest. This means that information that is closer to white than white background information of peripheral pixels is selected. After starting the copying operation, the above 1. ) When a white area, that is, a white background is detected by RGB white background detection, [1. ) White background determination signal output from RGB white background detection = “1”], white background information SS [I] of the pixel one line before the target pixel is updated to 15 (steps 3 and 4), and the white background information of the target pixel is updated. The MS is also set to 15 (step 5). Then, the white background information MS of the target pixel is written in the main scanning position (F) of the target pixel of the line memory for the current line (target line) of the line memory LMP shown in FIG. The information SS [I] is written in the main scanning position (F) of the pixel of interest in the line memory for the previous one line of the line memory LMP shown in FIG. 12 (steps 3 to 5). Next, the white background information SS [I] addressed to the pixel one line before is propagated to the pixel one line before as follows (steps 14 to 17). Note that [I] means the main scanning position of the pixel of interest, and [I-1] is the pixel one pixel before in the main scanning direction x (the pixel immediately before the pixel of interest).
Means the position of.

When SS [I-1] <SS [I] -1,
SS [I-1] = SS [I] -1 is set in the line memory (steps 14 and 15). That is, in the line one line before the target pixel, the white background information S one pixel before (E) the position (F) of the target pixel in the main scanning direction.
White background information S at the position (F) of the pixel of interest rather than S [I-1]
When the value "SS [I] -1" obtained by subtracting 1 from S [I] is larger (white degree is stronger), the pixel (pixel) one pixel before the position (F) of the target pixel on the line one line before ( The white background information SS [I-1] addressed to E) is updated to a value obtained by reducing the white intensity by 1 from the white background information SS [I] at the position (F) of the pixel of interest.

Next, when J = J + 1 (step 16), J
= MS does not hold (step 17; N), so return to step 14 and if SS [I-2] <SS [I] -2,
SS [I-2] = SS [I] -2 is set in the line memory (step 15). Next, again, J = J + 1 (step 16), and J = MS is not satisfied (step 1
7; N), returning to step 14, SS [I-3] <SS
If [I] -3, SS [I-3] = SS [I] -3 is set in the line memory (step 15). In the same manner, finally, SS [I-15] <SS [I] -15
In the case of, SS [I-15] = SS [I] -15 is set in the line memory (step 15). The lower limit value MIN of the value of the white background information SS [I] is 0, and when it is less than 0, it is kept at 0. This is step 1 below
The same is true for 3.

By the processing of these steps 14 to 17, the white background information SS one line before and the main scanning position of the target pixel is shifted from the white background information MS of the target pixel by one pixel in the main scanning direction x. The white background information of the pixel of interest is propagated to the rear of the main scan in the main scanning direction x one line before at the reduction ratio (white propagation processing). This is a case where the white background information one line before has a smaller value. For example, if the pixel one line before is 1. ) When the white background (white area) is detected by the RGB white background detection, the white background information of the white background is 15, which is the maximum value, and therefore rewriting is not performed.

When the pixel of interest is updated and becomes a non-white background [1. ) White background determination signal output from RGB white background detection = "0"], the process proceeds from step 3 to step 6 and thereafter, and the pixel of interest is the color background [see 2. ) Not the color-ground determination signal d = “1”] which is the output of the color-ground detection (a non-color-ground) and the output of the binarized white [the above-mentioned binarization 323a 2
The binarized white determination signal = “1”] and step 1
And the state variable of the pixel of interest provisionally determined in step 2, that is, the white background information MS is equal to or greater than the threshold value thw1 (for example, 13),
If it is, the white background information MS addressed to the pixel of interest is incremented by 1 (steps 6 to 10). That is, the white level is updated to a strong value by one. The maximum value max of the white background information MS is set to 15, and when it exceeds 15, it is limited to 15 (steps 9 and 10). Even when this route has been advanced, the above-mentioned step 5 and steps 14 to 17 are executed. That is, white propagation processing is performed.

Although the pixel of interest is a non-colored background and is binarized white, the white background information MS is less than thw1 (for example, 7), thw
When the number of pixels is 2 (for example, 1) or more and the number of pixels is a valley white pixel, the state variable MS is held at the same value (step 8, step 11, step 12). Even when this route has been advanced, the above-mentioned step 5 and steps 14 to 17 are executed. That is, white propagation processing is performed. When none of the above conditions are met, that is, when the pixel of interest is a color background or non-binarized white, the white background information MS of the pixel of interest is set to -1.
(Step 13). That is, the white background information is updated to have a weak white degree of 1. The minimum value MIN of the white background information MS is 0, and when it is less than 0, it is kept at 0. Even when this route has been advanced, the above-mentioned step 5 and steps 14 to 17 are executed. That is, white propagation processing is performed.

By generating the white background information MS as described above, the white information can be propagated to the peripheral pixels on the line memory LMP via the state variable (white background information) MS. The generation of the white background information MS is performed by the color data (R,
Generation of color-corresponding white background information MS of the system of steps 3 to 5 and steps 14 to 17 of FIG. In addition, the edge emphasis output of the density data (G image data) (filter 3
21) is smaller than the threshold value thwsb = 50, the white background information MS corresponding to the density of the system of steps 7 to 13, step 5 and steps 14 to 17 of FIG. 11 based on the white background and the binarized white determination signal. Including generation.

The white judgment 323c is first made in the step 1 in the RGB white extraction 323b. ) Until the white area is detected by RGB white background detection, that is, the above 1. ) RGB white background detection generates a white background determination signal "1" and correspondingly generates white background information MS corresponding to color (steps 3 to 5 and steps 14 to 17).
The operation (execution of step 4) is not performed until the start of. This is because the area that cannot be determined as a white area is filtered by the filter 3
This is to prevent erroneous determination as a white pixel (white block) in the white pattern matching of the G image data after the edge enhancement process 21 of FIG.

Edge enhancement filter 3 for light colored letters
When 21 is applied, the data around the character has a lower value (white) than the original image data (color background). Therefore, if white pattern matching is performed on the data after the edge enhancement processing of the filter 321, that is, density correspondence Generation of white background information MS (steps 7 to 13, step 5 and step 1)
If the white area is determined based only on 4 to 17), it is easy to erroneously determine the area around the characters on the color ground as a white background, but the white background information MS corresponding to the above-described color is generated (steps 3 to 5, step 14 to
17) The white background information MS is set to the highest value so that white pattern matching for determining a white pixel (white block) described later is applied to an area that can be determined to be a white area by 17),
When it is not a white background in step 3, the white background condition is checked in detail in step 6 and subsequent steps, and the white background information MS, which is one parameter for determining whether to apply the white pattern matching, is adjusted. Therefore, the edge of the filter 321 is adjusted. The white pattern (white block) is prevented from being erroneously determined by the white pattern matching of the post-emphasizing G image data, which will be described later.

For example, when the possibility of color pixels is high,
The white background information MS is lowered (step 13), and when there is a suspicion of a color pixel, the white background information MS is held (no change) (steps 11 to 13), and a white pixel (white block) is erroneously determined by white pattern matching described later. Prevent judgment,
This prevents the data around the characters from becoming a lower value (white) than the original image data (color background). Where the characters are dense, the above-mentioned processing (steps 3-5, 6-)
10 and steps 14 to 17), white background information MS
Is updated and propagated, the possibility that a dense character area is erroneously determined as a pattern is reduced. Also, among characters such as complicated characters (for example, “calligraphy”), 1. ) White may not be detected by RGB white background detection. At that time, 3. ) It is detected as white by the valley white pixel detection, and the white background information MS is detected in step 1
Since the YES output of 2 holds in the straight path to step 5 and keeps the tendency toward the white background, the possibility that a complicated character is erroneously determined as a pattern is reduced.

Further, as described above, when the pixel of interest is a color pixel surrounded by a white background, the above 2. ) Color gamut determination signal d = “1” (color ground), which is the output of the color gamut detection, can detect a light color pixel as a color ground even when a line is complicated, and whether the surroundings of the pixel of interest are white Since the viewing threshold thwc is set low (thwc = 20), it is possible to strictly check whether or not the periphery of the light color pixel (pixel of interest) is the white background and detect the light color pixel as the color background.
It is possible to further reduce the possibility that a complicated character is erroneously determined as a design.

Further, since the light color pixel can be detected more strictly as the color background, when it is detected as the color background, FIG.
Go to step 13 from step 6
In addition to reducing the possibility that the color background is determined to be white, the threshold value thwss (for example, 40) at the time of generating the white background determination signal referred to in step 3 is compared with step 7
Threshold value th in the case of generating the binarized white determination signal referred to in
If wsb (for example, 50) is set to a large value and it is not determined to be a color background (step 6; N), binarization 323a
Then, the probability of being considered as white is increased, and the process proceeds from step 7 to step 10 in FIG. 11 to increase the state variable MS to increase the possibility of determining a white area.

That is, 1. ) Threshold t for RGB white background detection
When hwss = 40, a strict white determination with a low probability of white determination is performed, and when the white background is determined, the state variable MS is increased and the character background is determined to be white by the processing from step 3 to step 4 in FIG. Is more likely to.
If the white background is not determined by the strict white determination, conversely, the highly reliable strict color background determination in which light color pixels are also detected as the color background, that is, 2. ) Referring to the result of color background detection,
When it is not judged as a color background, once again, if it is a white judgment with reference to the binarization 323a, that is, a white judgment with a threshold thwsb = 50 that has a high probability of being judged as white, that is, the state is white, The variable MS is increased to increase the possibility that the character background is determined to be white (steps 6-1).
0). Since this processing (steps 6 to 10) is performed, when the background density unevenness is lighter than the light color pixel detected as the color background, for example, when there is unevenness on the background of the original such as show-through, fine background unevenness of the original. It is possible to prevent the state variable MS from greatly changing in a binary manner in conjunction with the above, and to prevent the determination as to whether or not it is a white pixel in the next white pattern matching 323d from finely changing in the scanning direction. As a result, when the background is a light-colored background, fine color loss (white background) does not appear in association with fine background unevenness of the document such as show-through.

Next, the white pattern matching 323d will be described. Whether or not the background is white is determined depending on whether or not there are continuous white pixels in a block of 5 × 5 pixels centering on the pixel of interest. Therefore, when the following formula is satisfied for the target pixel, the target pixel is provisionally defined as a white pixel,
Perform white pattern matching. (Non-color pixel & (white background information MS ≧ thw1 (13)) & 2
Quantized white) # (non-color pixel & (white background information MS ≧ thw2 (1)) & valley white pixel & binarized white) Here, the pixel of interest for which it is checked whether this conditional expression is satisfied is the step of FIG. 5 and the white propagation processing target of steps 14 to 17 has been through the processing process, and the “white background information MS” in the above conditional expression is the white background information MS [of the target pixel to be checked after the white propagation processing. I]. This MS [I] is the white background information after the white propagation processing is completed, and its I is the position in the main scanning direction x of the target pixel to be checked, and the target pixel for calculating the state variable MS in the white determination 323c. Is different from the position in the main scanning direction x.

The "non-color pixel" in the above condition is 2. )
The color-ground determination signal d that is the output of the color-ground detection is "0", and the "binarized white" is "1" (binarized white) for the binarized white judgment signal of the binarization 323a. And “Tanihaku pixel” are 3. ) It means that the detection result of the valley white pixel detection is a valley white pixel, and # means a logical sum (OR: or). The white pattern matching 323d corresponds to the output (whether it is a white pixel or not) determined by the above-mentioned conditional expression.
It is checked whether any of the two vertical, horizontal, and diagonal continuity patterns PMPa to PMPd corresponds. White circles attached to the patterns PMPa to PMPd mean that the pixels are white pixels. It does not matter whether the other blank pixels are white pixels or not. The white pixel distribution of the 5 × 5 pixel matrix centering on the target pixel is the pattern PMPa, PMP of FIG.
If it corresponds to b, PMPc, or PMPd, it is determined that the pixel of interest is a white pattern pixel.

Next, the gray judgment will be described. R,
Hue division of G, B, Y, M, C, and Bk is performed, and a pixel with low density is detected for each hue. The hue division is the same as the color determination described later. Here, the filtered G data is compared with thgr and is greater than the G data, or
If either of the color pixels is detected in the RGB white extraction color pixel detection, the following calculation is performed, and if the following conditions are satisfied, the pixel is determined to be a gray pixel. The reason why the threshold is changed for each color is that the maximum density of each ink is different. 4.1). RY hue region boundary (ry) R-2 ^* G + B> 0 4.2). Y-G hue region boundary (yg) 11 ^* R-8 ^* G-3 ^* B> 0 4.3). G-C hue region boundary (gc) 1 ^* R-5 ^* G + 4 ^* B <0 4.4). CB hue region boundary (cb) 8 ^* R-14 ^* G + 6 ^* B <0 4.5). B-M hue region boundary (bm) 9 ^* R-2 ^* G-7 ^* B <0 4.6). MR hue region boundary (mr) R + 5 ^* G-6 ^* B <0

4.7). Y pixel pixel determination (gry) (color pixel) & (ry == 1) & (yg == 0) &
(Maximum RGB value <thmaxy) 4.8). G pixel determination (grg) (is a color pixel) & (yg == 1) & (gc == 0) &
(Maximum RGB value <thmaxg) 4.9). C pixel determination (grc) (color pixel) & (gc == 1) & (cb == 0) &
(Maximum value of RGB <thmaxc) 4.10). B pixel determination (grb) (which is a color pixel) & (cb == 1) & (bm == 0) &
(Maximum value of RGB <thmaxb) 4.11). M pixel determination (grm) (color pixel) & (bm == 1) & (mr == 0) &
Maximum value of RGB <thmaxm) 4.12). R pixel determination (grr) (color pixel) & (mr == 1) & (ry == 0) &
(Maximum RGB value <thmaxr) 4.13). When not a color pixel (grbk) (RGB maximum value <thmaxbk) 4.14). Gray pixel determination 4.7). ~ 4.14). When any one of the conditions is satisfied, a gray pixel is set.

Next, gray pattern matching 323f
Will be described. The following pattern matching is performed, where D is a gray pixel and bk is darker than the gray pixel. The copy manuscript has a thin 200-line parallel pattern,
Since it is a line of 300 lines, the following pattern is adopted so that the copy original is also detected in gray. Those that match the following pattern become gray pixels (Fig. 13).
(See (a), (b), (c)).

Next, 4. ) The white pattern correction will be described. Isolated by white pixel pattern matching (1x1, 1x
2, 2 × 1, 2 × 2, 1 × 3, 3 × 1 white pixels) are deactivated. This removes isolated pixels. 5. ) The result of correction of white expanded white pixel pattern matching is set to O of 7 × 41.
Perform R. 6. ) AND of 1 × 33 of the result of white contraction and white expansion. By performing white expansion and white contraction, expansion and removal of inactive pixels existing in a small area are performed for the correction result of white pixel pattern matching. This determination result is a result including a boundary area on the non-white background side with respect to the white background and the boundary portion. In other words, the area is larger than the white background. 7. ) White correction In white block correction, when one or more white candidate blocks are present in each 6 × 4 pixel area at four corners in 15 × 11 pixels centered on a pixel of interest marked with × in the block pattern BCP. , White block correction data is given to the block of interest. As a result, the area surrounded by the white background becomes the white area.

8. ) 11 × 11 for the result of gray expanded gray pattern matching
OR processing of. This makes the area slightly larger than the gray area. 9. ) A white background is used when the judgment white correction result is active or when the contraction result is active and the gray expansion result is inactive. It can be expressed as the following equation. Result of white correction # (Result of white contraction &! Result of gray expansion) Here, in the result of white correction, the area written on the white background is definitely judged as a white area, and the result of white contraction &! In the result of the gray expansion, the dark black character periphery is a white area, and the light density area is a non-white area.

The black protrusions surrounded by circles Bp1 to Bp4 in FIG. 14 are one or more white candidate blocks in each 6 × 4 pixel area at four corners in 15 × 11 pixels centered on the above-mentioned block of interest. , The white block can be replaced with the white block by the white block correction that gives the white block correction data to the target block. Round Bp1 to Bp
Setting a black area surrounded by a white background such as a black protrusion surrounded by 4 as a white area reduces the possibility that the black area is determined as a pattern portion. In a comprehensive determination 326 described later, the non-white area is determined to be a pattern, but the possibility of erroneously determining a black area surrounded by a white background such as a black protrusion surrounded by circles Bp1 to Bp4 to be a pattern is reduced. Further, the white contraction result and the gray expansion result determine the boundary between the black background and the white background as a white area (character area), so a dark character edge is determined as a white background regardless of the character thickness. It is possible to correctly determine the character edge. Areas with low density are no longer judged as character edges.

Next, the adjustment of the character / photo judgment level will be described. In the white area extraction 323, the white background determination signal and the color background determination signal d of the RGB white extraction 323b in the white determination 323c.
And white background information MS, which is a state variable representing the degree of white, corresponding to the valley white pixel determination signal and the binarized white determination signal of the binarization 323a. Then, in the white pattern matching 323d, the color background determination signal d and the white background information M
Based on the S-binarized white determination signal and the valley white pixel determination signal, it is temporarily determined whether the pixel of interest is a white pixel, and white pixel distribution pattern matching is performed on the pixel matrix including the pixel of interest to determine whether the pixel is a white pixel. Using this result and the result of the gray pattern matching 323f, the white correction 3
The target pixel of 23 g is a boundary between a black background and a white background (white area:
Character area) is determined.

The white background determination signal of RGB white extraction 323b (step 3 in FIG. 11) is "1" when all of the R, G, and B image data of the target pixel are smaller than the threshold value thwss = 40.
(White background). When the threshold value thwss is increased, the probability that the white background information MS having a large value is determined increases, and the probability that the “white area” (border between the black background and the white background: character area) is extracted increases. That is, the probability of extracting the pattern area is reduced. If the threshold value thwss is reduced, the opposite is true. Two
The binarized white determination signal of the binarization 323a (step 7 in FIG. 11) is “1” (2 if the edge emphasis output of the G image data of the filter 321 is smaller than the threshold thwsb = 50.
Quantized white). If the threshold value thwsb is increased, the probability that the white background information MS having a large value is determined increases, and the probability that the “white area” is extracted increases. That is, the probability of extracting the pattern area is reduced. If the threshold value thwsb is reduced, the opposite is true.

Since the image data of the "white area" is subjected to the image processing for clearly expressing the character image in the later step, if the thresholds thwss and thwsb are increased, the image processing having a high priority is applied to the character. To be done. Since the image data for the non-white area, that is, the picture (photo) area is subjected to image processing for faithfully representing the picture or the picture in a later step, the threshold value th
If wss and thwsb are reduced, the design (photo)
Is subjected to image processing with high priority. By the way, RGB
When the color background determination signal d (step 6 in FIG. 11) of the white extraction 323b is "1" (color background), the white background information MS is lowered and the probability of extracting the "white area" is lowered. That is,
The probability of extracting the picture area increases. 2. ) Process for generating color background determination signal d by color background detection B. And C.I. If the threshold value thwc (for example, 20) used in step 1 is reduced, the probability that peripheral pixels (triangle (Δ) and square (□) in FIG. 10) are simultaneously detected as color pixels, that is, (exclusive NOR of b and c) = “1” The probability that the color ground determination signal d
= The probability of obtaining "1" (color background) is also high, and the probability of extracting the "white area" is low. That is, the probability of extracting the pattern area is increased.

Therefore, in this embodiment, the operation / operation of FIG.
"Character / Photograph" in the parameter adjustment that is adjusted by operating the keypad (parameter specification key and up / down key) on the menu screen of the input mode by the key input on the display unit OPB and the menu screen displayed on the liquid crystal display. Threshold th based on the adjustment of “level”
Adjust wss, thwsb and thwc as shown in FIG.

That is, the operator operates the operation / display unit OP.
The standard value (default) of the parameter "text / photo level" adjusted and set in B is "3", and this default value is the text / photo level and the thresholds thwss, thwsb.
And a conversion table showing the relationship with thwc,
The data is written in the ROM (Read Only Memo) 358 shown in FIG. 3, and when the IPU 300 shown in FIG. 3 is powered on and the CPU (Central Processing Unit) 357 initializes the IPU 300, the CPU 357 causes the ROM 358. The default value of the character / photo level is read from, the corresponding threshold values thwss, thwsb, and thwc are read from the conversion table, and are written in the respective threshold value registers of the RAM (random access memory) 356. Used for processing. After that, the character / photo level is adjusted by input from the operation / display unit OPB, and the adjusted value A is sent from the main controller 10 to the CP.
When given to U357, the CPU 357 reads the values of the parameters thwss, thwsb, and thwc corresponding to the adjusted value A from the conversion table of the ROM 358, and writes them in the parameter-addressed register of the RAM 356.

The threshold value is a standard value thwss = 40, thws
When b = 50 and thwc = 20 are set, if the operator uses the operation / display unit OPB to increase the value of “text / photo level” by i (for example, 1) and “Up (up)”, the threshold value is increased. thwss, thwsb and th
wc is set to a value changed in the character priority direction by 2i (2). On the contrary, when the operator decreases the value of “text / photo level” by i (for example, 1) and “Down”, the thresholds thwss, thwsb and thwc are 2
i (2) is set to a value changed in the photo priority direction.

Next, the halftone dot extraction 324 will be described.
FIG. 16 is a diagram showing the configuration of the halftone dot extraction 324. The first halftone dot peak detection unit 324a uses the G image data to form a pixel that forms part of a halftone dot dot from the pixel density information in the two-dimensional local area of a predetermined size (referred to as a halftone dot peak pixel). ) Is a circuit for detecting. Regarding the local area, when the following two conditions are satisfied at the same time, the central pixel of the area is detected as a halftone dot peak pixel. Condition 1 is that the density level of the central pixel is maximum (peak peak) or minimum (valley peak) in the local region. The condition 2 is that the absolute value of the difference between the average of the density levels of the pixel pairs and the density level of the central pixel is equal to or more than the threshold value Th for all pairs of pixels having a point-symmetrical relationship with the central pixel. To do.

Here, the detection processing of the first halftone dot peak detection 324a will be specifically described with reference to FIG. Figure 1
7 is an example in which a mask of a 5 × 5 pixel matrix (M × M pixel matrix in generalized form) is adopted as the local area. The code of each pixel of the 5 × 5 pixel matrix is shown in FIG.
If the pattern MPp of 0 is used, the density Lc of the central pixel c3, which is the pixel of interest, is the density L1 of the surrounding pixels
~ Maximum or minimum compared to L8 and ab
s (2Lc-L1-L8) ≧ Lth, and abs (2
Lc-L2-L7) ≧ Lth, and abs (2Lc-
L3-L6) ≧ Lth, and abs (2Lc-L4-
When L5) ≧ Lth, the central pixel (Lc) of the mask is detected as a halftone dot peak pixel. The abs function means to take an absolute value. Lth is a threshold value (fixed value).

Specifically, the surrounding pixels are pixels to which a quadrangle of the surrounding pixel distribution pattern MPa or MPb shown in FIG. 17 is added. Surrounding pixel distribution pattern MPa and MP
Either of the halftone dot peak pixel detections described above based on b
When it is detected as a halftone dot peak pixel, a detection signal representing the halftone dot peak pixel is given to the target pixel (center pixel c3) at that time. The reason why two patterns are used is to widely correspond to the number of halftone dots. The pattern MPa is L1 = b
2, L2 = b3, L3 = b4, L4 = c2, L5 = c
4, L6 = d2, L7 = d3, L8 = d4. Here, L1 = b2 means the density of the pixel b2,
This means setting the value to L1 in the above-mentioned halftone dot peak pixel detection calculation. The pattern MPb has L1 = b2 and L2 = a.
3, L3 = b4, L4 = c1, L5 = c5, L6 = d
2, L7 = e3 and L8 = d4.

In the case of copying, enlargement in the sub-scanning direction y,
Since the reduction is performed at low and high document scanning speeds of the scanner 200, the scanner 200 supplies the image data that has been enlarged or reduced in the sub-scanning direction y. Therefore, at the time of reduction, instead of the patterns MPa and MPb described above,
The patterns MPc and MPd shown above are used. At the time of enlargement, the patterns MPe and MPf shown in FIG. 17 are used. In addition, in the patterns MPe and MPf, the pixels given the triangle marks may be added to the above-mentioned “surrounding pixels”.

The second halftone dot peak detector 324b detects halftone dot peaks using B data, and has the function of:
This is the same as the first halftone dot peak detector 324a. Since the first halftone dot peak detector 324a uses G image data, it reacts with most colors, but does not react with Y, so the second halftone dot peak detector 324b uses B image data. In addition, it is an auxiliary object for detecting the halftone dot peak of Y.

The halftone dot area detecting unit 324d determines the peak and valley halftone dot peak pixels detected by either the first halftone dot peak detecting unit 324a or the second halftone dot peak detecting unit 324b to have a predetermined size. Counting is performed for each two-dimensional small area, and the sum of halftone dot peak pixels of peaks and valleys is set as the count value P of the small area. When the count value P is larger than the threshold value Pth, all the pixels in the small area (or only the central pixel of the small area in the case of pixel unit processing) are determined as the halftone dot area. The determination result is stored in the temporary storage unit 324f.

Here, the detection processing of the third halftone dot peak detector 324c will be specifically described with reference to FIG. The third halftone dot peak detector 324c is intended to detect 100 lines or less and 65 lines (newspaper dots). This is an example in which a mask of a 7 × 7 pixel matrix (M × M pixel matrix in generalized form) is used as the local area. If this is shown in the pattern of FIG. 19, the density Lc of the central pixel group that is the pixel of interest is maximum or minimum as compared with the density groups L1 to L8 of the surrounding pixels, and abs (2Lc-L1- L8) ≧ Lth, and
abs (2Lc-L2-L7) ≧ Lth, and abs
(2Lc-L3-L6) ≧ Lth, and abs (2L
When c−L4−L5) ≧ Lth, the central pixel (Lc) of the mask is detected as a halftone dot peak pixel. The abs function means to take an absolute value. Lth is a threshold value (fixed value).

The halftone dot peak pixel when the central pixel (Lc) of the mask has the maximum value is output to the cycle check 324g as the halftone dot peak pixel. Further, the halftone dot peak pixel when the central pixel (Lc) of the mask has the minimum value is output to the cycle check 324h as the halftone dot peak pixel. Specifically, the surrounding pixels are pixels having the surrounding pixel distribution pattern shown in FIG. When any one of the above-described halftone dot peak pixel detections based on the surrounding pixel distribution pattern is detected as a halftone dot peak pixel, a detection signal representing the halftone dot peak pixel is given to the target pixel (center pixel d4) at that time.
The reason why two patterns are used is to widely correspond to the halftone dot area ratio.

For the density of Lc, Lc =
It is calculated as Min (d4, d3, d5, c4, e4). When this Lc is the maximum value with respect to the peripheral pixels, the pattern is L1 = Max (a1, a2, b1), L2 = Ma.
x (a3, a4, a5), L3 = Max (a6, a7,
c7), L4 = Max (c1, d1, e1), L5 = M
ax (c7, d7, e7), L6 = Max (f1, g
1, g2), L7 = Max (g3, g4, g5), L8
= Max (g6, g7, f7). Here, L1 = Max (a1, a2, b1) means the pixel a
This means that the maximum value of the densities of 1, a2, and b1 is set as the value of L1 in the above-mentioned halftone dot peak pixel detection calculation. Lc = M
in (d4, d3, d5, c4, e4) means d4, d
It means the minimum value of the density (image data) of 3, d5, c4, and e4.

Further, Lc = Max (d4, d3, d5,
c4, e4), if Lc is the minimum value for the surrounding pixels, the pattern is L1 = Min (a1, a2, b
1), L2 = Min (a3, a4, a5), L3 = Ma
x (a6, a7, c7), L4 = Max (c1, d1,
e1), L5 = Max (c7, d7, e7), L6 = M
ax (f1, g1, g2), L7 = Max (g3, g
4, g5) and L8 = Max (g6, g7, f7).

In the case of copying, enlargement in the sub-scanning direction y,
Since the reduction is performed at low and high document scanning speeds of the scanner 200, the scanner 200 supplies the image data that has been enlarged or reduced in the sub-scanning direction y. Therefore, at the time of reduction, the pattern shown in FIG. 18 (b) is used. When enlarging, the pattern shown in FIG. 18 (a) is used. Third
The calculation formula of the halftone dot peak detection unit 324c does not calculate with the data of one pixel, but refers to the target pixel with a plurality of pixels (calculation of min and max). This is because halftone dots with a low number of lines have a large shading period. Therefore, when determining with one pixel, by referring to peripheral pixels, the influence of noise (dust) is reduced and the amount of arithmetic calculation is reduced. , Since we are making it possible to use arithmetic expressions in common with other blocks,
Easy to harden.

In the period check 324g, the period between the halftone dot peak and the halftone dot peak is checked. Check the interval of the halftone dot peaks in the one-dimensional main scanning direction. Since the number of detected halftone dots in halftone dot peak detection is 100 or less, the halftone dot peak is 8 or more. If it is less than 8, it is almost always the case that a small character is misjudged (600
(when reading dpi). As shown in FIG. 20A, L1
Is 6 and L2 is 10, the halftone dot peaks at L1 intervals are mostly small characters, so the halftone dot peaks are removed. Since the L2 interval is mostly a low frequency halftone dot, it is not removed. Figure 2 in the cycle check
0 (a) is corrected as shown in FIG. 20 (b). In the cycle check 324h, the halftone dot valley peak is checked in the cycle check 3
Do the same as for 24 g. The halftone dot peak and the halftone dot peak are performed independently because the halftone dot peak and the halftone dot peak appear alternately in the vicinity of the halftone dot area ratio of 50%, and an accurate cycle does not appear independently. Is going. Cycle check 324
g, the output of the cycle check 324h is logically ORed by the OR 324i and input to the halftone dot area detection unit 324e.

The halftone dot area detection unit 324e ORs the period checks 324g and 324h with the OR 324i, counts the halftone dot peak pixels of the peaks and valleys for each two-dimensional small area of a predetermined size, and calculates the peaks. Let the sum of the halftone dot peak pixels of the and valley be the count value P of the small area. When the count value P is larger than the threshold value Pth, all the pixels in the small area (or only the central pixel of the small area in the case of pixel unit processing) are determined as the halftone dot area. The result of the determination is stored in the temporary storage unit 324f. Further, the halftone dot area detection unit 324e has a period check 324g,
324h is ORed by OR 324i, and the peak peak pixels of the peaks and valleys are counted for each two-dimensional small area of a predetermined size, and the sum of the peak peak pixels of the peaks and valleys is the count value of the small area. P
And When the count value P is larger than the threshold value Pth,
All the pixels in the small area (or only the central pixel of the small area in the case of pixel-based processing) are determined as the halftone dot area. The determination result is stored in the temporary storage unit 324j.

If either the halftone dot area detection unit 324d or the halftone dot area detection unit 324e is the halftone dot area, the halftone dot / non-halftone dot judgment result (periphery) of the processed area in the vicinity of the small area of interest. The threshold value Pth is changed in accordance with the feature information). In the present embodiment, two values TH1 and TH2 (TH1> TH2) are prepared as the threshold value Pth, and according to the determination result of the processed area in the vicinity of the target small area stored in the temporary storage unit 324f, Select one of these values. That is, when the neighboring area is determined to be the non-halftone dot area, it is highly likely that it is a line drawing area. Therefore, TH1 which has a stricter condition is selected as the threshold Pth in order to reduce erroneous detection. On the other hand, when it is determined that the neighborhood area is the halftone dot area, it is highly likely that the neighborhood area is the halftone dot area.
Used as. The initial value of the threshold Pth is TH
Select 1.

The distribution of the small areas described above is shown in the AMP of FIG. Each of S1 to S4 of the small area distribution pattern AMP is, for example, a small area (block) having a size of 4 × 4 pixels, and the small areas of interest in S4 are S1 and S2.
And S3 are processed small areas. S1, S
When it is determined that 2 and S3 are all halftone dot areas, TH2 is used as the threshold Pth for the determination of S4. If even one of S1, S2, and S3 is determined to be a non-halftone dot area, TH1 is set as the threshold Pth.
Is selected. A halftone dot area detection signal ht of "1" when it is determined to be a halftone dot area and "0" when it is determined to be a non-halftone dot is output from the halftone dot extraction 324. Note that this is an example, and TH2 is selected when any one of the small areas S1, S2, and S3 is determined to be a halftone dot area, and TH1 is selected only when all are determined to be non-halftone dot areas. May be selected. Further, the neighboring area to be referred to may be only S1 or only S2. This output result is a halftone dot extraction result.

If the halftone dot area detection unit 324e is a halftone dot area, the halftone dot / processed area near the small area of interest /
The threshold Pt is determined according to the non-halftone dot determination result (surrounding feature information).
Change h. In the present embodiment, the threshold Pth
As a result, two values TH1 and TH2 (TH1> TH2) are prepared, and according to the determination result of the processed area near the small area of interest stored in the temporary storage means 324j,
Select one of these values. That is, when the neighboring area is determined to be the non-halftone dot area, it is highly likely that it is a line drawing area, and therefore the condition becomes severe in order to reduce erroneous detection.
1 is selected as the threshold Pth. On the other hand, when it is determined that the neighboring region is the halftone dot region, it is highly likely that the neighboring region is the halftone dot region, and therefore TH2 where the condition is loose is used as the threshold Pth. Note that TH1 is selected as the initial value of the threshold Pth.

The distribution of the small areas described above is shown in the AMP of FIG. Each of S1 to S4 of the small area distribution pattern AMP is, for example, a small area (block) having a size of 4 × 4 pixels, and the small areas of interest in S4 are S1 and S2.
And S3 are processed small areas. S1, S
When it is determined that 2 and S3 are all halftone dot areas, TH2 is used as the threshold Pth for the determination of S4. If even one of S1, S2, and S3 is determined to be a non-halftone dot area, TH1 is set as the threshold Pth.
Is selected. A halftone dot area detection signal ht of "1" when it is determined to be a halftone dot area and "0" when it is determined to be a non-halftone dot is output from the halftone dot extraction 324. Note that this is an example, and TH2 is selected when any one of the small areas S1, S2, and S3 is determined to be a halftone dot area, and TH1 is selected only when all are determined to be non-halftone dot areas. May be selected. Further, the neighboring area to be referred to may be only S1 or only S2. This output result is a low-line halftone dot extraction result. The difference from the halftone dot extraction result is that in the halftone dot peak detection, the halftone dots are extracted only by inputting a low number of lines, so only the low number of halftone dots can be extracted.

Next, the color judgment 325 will be described. When detecting a color (chromatic) pixel or a black (achromatic) pixel in the document, a relative read deviation of R, G, and B exists due to sampling of each color image data and mechanical accuracy. . The image density signal will be described with reference to FIG.
FIG. 21A shows an image density signal. Ideally, the black density signal is ideal black when the levels of the R, B, and G density signals match. However, the actual image data is
An image is formed on the CCD with a lens, and the image signal of the CCD is digitized. FIG. 21B shows an ideal high and low waveform. Since a general scanner uses a 3-line CCD sensor, R, G, and B images of image data are not read simultaneously at the same time, but the R, G, and B line sensors are equally spaced. Since they are arranged and they cannot be read at the same time in terms of time, a reading position shift will inevitably occur. For example, the R, G, and B color density signals representing the black of the level change shown in FIG. 21B are relatively displaced as shown in FIG. 21C. If this deviation is large, color deviation appears at the periphery of the black area.

Next, the hue division 325a will be described. The color determination 325 is to find a chromatic color area. The input data R, G, and B are processed by the hue division 325a.
c, m, y and color determination w (white) signals are converted. As an example of the hue division, the boundary of each color is obtained, and the difference between the maximum value and the minimum value of the image data of each of R, G, and B in one pixel is defined as an RGB difference, and is as follows. Here, the R, G, B image data becomes black (darker) as the number increases. 1). RY hue region boundary (ry) R-2 ^* G + B> 02). Y-G hue region boundary (yg) 11 ^* R-8 ^* G-3 ^* B> 03). G-C hue region boundary (gc) 1 ^* R-5 ^* G + 4 ^* B <04). CB hue region boundary (cb) 8 ^* R-14 ^* G + 6 ^* B <05). BM hue region boundary (bm) 9 ^* R-2 ^* G-7 ^* B <06). MR hue region boundary (mr) R + 5 ^* G-6 ^* B <0

7). Color determination w (white) pixel determination: (R <
If thwa) & (G <thwa) & (B <thwa), then y = m = c = 0. Note that thwa is a threshold value. 8). Y pixel determination: (ry == 1) & (yg == 0) &
If (RGB difference> thy), then y = 1 and m = c = 0. In addition, thy is a threshold value. 9). G pixel determination: (yg == 1) & (gc == 0) &
If (RGB difference> thg), c = y = 1 and m = 0. Note that thg is a threshold value. 10). C pixel determination: (gc == 1) & (cb == 0)
& (RGB difference> thc), c = 1 and m = y = 0. Note that thc is a threshold value.

11). B pixel determination: (cb == 1) &
If (bm == 0) & (RGB difference> thb), then m = c
= 1 and y = 0. Note that thb is a threshold value. 12). M pixel determination: (bm == 1) & (mr == 0)
& (RGB difference> thm), m = 1 and y = c = 0. Note that thm is a threshold value. 13). R pixel determination: (mr == 1) & (ry == 0)
& (RGB difference> thr), y = m = 1 and c = 0. Note that thr is a threshold. 14). BK pixel determination: 7) above. ~ 13). When the above does not apply, y = m = c = 1.

Further, the w pixel for color judgment is judged. The conditions are as follows. (R <thw) & (G <th
w) & (B <thw), it is set as w pixel for color pixel, and w
Output as. Here, thw is a threshold value. 7).
~ 14). As for the priority order of, the smaller number has priority. The above thresholds thwa, thy, thm, thc, thr,
thg and thb are threshold values determined before copying (processing). The relationship between thw and thwa is thw> tha. The output signal is 1-bit 3-bit data for each of c, m, and y, and 1-bit for w pixel detection for color determination. The reason why the threshold value is changed for each hue here is that the threshold value is determined according to the hue area when the chromatic range is different for each hue area. This hue division 325a is an example, and any formula may be used. Hue division 325a
Output c, m, y, w of the line memories 325b to 325b.
Five lines are stored in 5e and input to the color pixel determination 325f.

Next, the color pixel determination 325f will be described. FIG. 22 is a diagram showing the configuration of the color pixel determination 325f. The data of c, m, y, and w for four lines are pattern matching 325f5 to 325f7 and count 3
25f1 to 325f3. First, B / C
Pattern matching 325f6 in the flow of obtaining a signal
Will be described. When the w pixel for the color pixel exists, the pixel is corrected to c = m = y = 0. With this correction,
The white level of the 5 × 5 pixel matrix centering on the pixel of interest becomes large. Next, for the pixel of interest, all of c, m, and y of the pixels determined by the hue division 325a are 1 (c = m = y = 1).
Alternatively, by checking whether all are pixels (color pixels) other than 0 (c = m = y = 0) and whether the 5 × 5 pixel matrix matches the following pattern,
judge. 1). Color pixel pattern group 1-1). Pattern 1-1 (pm1) D23 & D33 & D43 1-2). Pattern 1-2 (pm2) D32 & D33 & D34 1-3). Pattern 1-3 (pm3) D22 & D33 & D44 1-4). Pattern 1-4 (pm4) D24 & D33 & D42 The central pixel (pixel of interest) is D33.

FIG. 23 shows the above patterns pm1 to pm4.
It is the figure which showed. The white circles on these patterns are c,
At least one of m and y is 1. The reason why pattern matching is adopted is to prevent isolated points from being picked up. Conversely, when detecting a small area color such as a halftone dot, pixels (color pixels) whose central pixel is other than 1 (c = m = y = 1) or all 0 (c = m = y = 0) It may be determined depending on

2). Color thin line pattern group A color line surrounded by white is detected. The pattern used for this is shown in FIG. In FIG. 24, the pixels with white circles are
Pixels in which c, m, and y are all 0. Data (c, 5x5 pixel matrix centered on the pixel of interest (center pixel))
m, y) distribution is the pattern pw11a to pw of FIG.
If any of 14d is matched, the pixel of interest (center pixel) at that time is regarded as a color line pixel. 2-1). Pattern 2-1 (pw11a~pw11d) ((D12 & D13 & D14) & (D42 & D43 & D44)) # ((D12 & D13 & D14) & (D52 & D53 & D54)) # ((D22 & D23 & D42) & (D42 & D43 & D44)) # ((D22 & D23 & D42) & (D52 & D53 & D54)) 2- 2). Pattern 2-2 (pw12a~pw12d) ((D21 & D31 & D41) & (D24 & D34 & D44)) # ((D21 & D31 & D41) & (D25 & D35 & D45)) # ((D22 & D23 & D24) & (D24 & D34 & D44)) # ((D22 & D23 & D24) & (D25 & D35 & D45)) 2- 3). Pattern 2-3 (pw13a to pw13d) ((D11 & D21 & D12) & (D35 & D44 & D53)) # ((D11 & D21 & D12) & (D45 & D44 & D55)) # ((D13 & D22 & D31) & (D35 & D44 & D53)) # ((D13 & D && D & 3 & D22 & D22 & D & 3). 4). Pattern 2-4 (pw14a to pw14d) ((D13 & D24 & D35) & (D41 & D51 & D52)) # ((D14 & D15 & D25) & (D41 & D51 & D52)) # ((D13 & D24 & D35) & (D31 & D42 & D53)) # ((D14 & D && 15 & D15)).

3). Pattern matching is performed when the white area pattern groups c, m, and y are all 0. The pattern used for this is shown in FIG. In FIG. 25, pixels with white circles are pixels in which c, m, and y are all 0. When the distribution of the data (c, m, y) of the 5 × 5 pixel matrix centered on the target pixel (center pixel) matches any of the patterns pw21a to pw24d in FIG. 25, the target pixel (center pixel at that time) ) Is regarded as a white area pixel. 3-1). Pattern 3-1 (pw21a to pw21d) (D21 & D31 & D41) # (D22 & D32 & D42) # (D24 & D34 & D44) # (D25 & D35 & D45) 3-2). Pattern 3-2 (pw22a to pw22d) (D12 & D13 & D14) # (D22 & D23 & D24) # (D42 & D43 & D44) # (D52 & D53 & D54) 3-3). Pattern 3-3 (pw23a to pw23d) (D52 & D51 & D41) # (D53 & D42 & D31) # (D35 & D24 & D13) # (D25 & D15 & D14) 3-4). Patterns 3-4 (pw24a to pw24d) (D54 & D55 & D45) # (D53 & D44 & D35) # (D31 & D22 & D13) # (D21 & D11 & D12).

4). Determination of Color Pixel Candidate 2 If the pattern matching result extracted above matches the following pattern, the pixel of interest is set to color determination candidate color pixel 2. ((Pm1 == 1) & ((pw11 == 1) # (pw21! = 1))) # ((pm2 == 1) & ((pw12 == 1) # (pw22! = 1))) # ((Pm3 == 1) & ((pw13 == 1) # (pw23! = 1))) # ((pm4 == 1) & ((pw14 == 1) # (pw24! = 1))) # Here, (pm1 == 1) means that the data distribution centered on the pixel of interest matches the pattern pm1, and (pw11 == 1) means the patterns pw11a to pw.
means to match any of w11d,
(Pw21! = 1) is the pattern pw21a to pw21.
It means to match any one of d. & Means logical product, and # means logical sum. By this pattern matching, the color pixel surrounded by the white area is set as the color pixel candidate, and when the white area exists in other areas, it is not set as the color pixel. A color pixel pattern matching without a white region is a color pixel candidate.

Next, the count 325f1 will be described. Within the 5 x 5 pixel matrix centered on the pixel of interest,
If there is a w pixel for color determination, the hue division 3 of that pixel
The c, m, y data determined in 25a is corrected to c = m = y = 0. This correction increases the white level of the pixel matrix. Then, the number of 1s (c = 1, m = 1, y = 1) of c, m, and y of each pixel in the pixel matrix is counted. If the difference between the maximum value and the minimum value of the count values for each of c, m, and y is greater than or equal to thcnt and the minimum value is less than thmin, color pixel candidate 1 is set. th
cnt and thmin are threshold values set before copying (processing). Planes are expanded to y, m, and c, and the number is counted for each plane in the N × N matrix.
The minimum value is assumed to be black. As a result, even if the reading of the black pixel is omitted, the correction can be performed. Then, the chromatic pixel is determined based on the difference between the maximum value and the minimum value. This corrects pixels where black pixels are out of reading,
Extract chromatic pixels. If there is a fixed chromatic pixel in the 5 × 5 pixel matrix centered on the target pixel, the target pixel is regarded as a chromatic pixel.

The color pixel determination 325f8 determines whether or not the pixel is a color pixel in the color pixel determination 325f8 based on the output of the pattern matching 325f6 and the count 325f1. If it is the color pixel candidate 1 and the color pixel candidate 2, the color pixel 1 is set. Blocking 325f9 is performed by the color pixel determination 325.
Block the output of f8. Blocking is 4x
In a 4-pixel matrix, 1 or more color pixels 1
If so, the entire 4 × 4 pixel matrix is output as one block of color pixels. In the processing after the block formation 325f9, 4 × 4 pixels are set as one block and output in block units. The isolated point removal 325f10 removes the blocked data as an isolated point in the isolated point removal 325f10 if there is no one color pixel block in the block adjacent to the target block.

The expansion 325f11 is the isolated point removal 325f.
In the expansion 325f11, the output of 10 is expanded to 5 × 5 blocks when one color pixel block exists. The reason for expansion is to prevent black character processing around the color pixels. Here, the B / C signal to be output outputs L (chromatic) in the case of one color pixel block, and outputs H (achromatic) in other cases.

Next, the count 325f2 will be described. When the w pixel for color determination exists in the 5 × 5 pixel matrix centered on the pixel of interest, the hue division of that pixel 32
The c, m, and y data determined in 5a are corrected to c = m = y = 0. This correction increases the white level of the pixel matrix. Then, the number of 1s (c = 1, m = 1, y = 1) of c, m, and y of each pixel in the pixel matrix is counted. If the difference between the maximum value and the minimum value of the count values for each of c, m, and y is greater than or equal to thacnt and the minimum value is less than thamine, the pixel of interest is set to color pixel candidate 1. Thacnt and thamine are threshold values set before copying (processing).

The color pixel determination 325f12 determines whether or not it is a color pixel in the color pixel determination 325f12 based on the output of the pattern matching 325f6 and the count 325f2. If it is the color pixel candidate 1 and the color pixel candidate 2, the color pixel 2 is set. Blocking 325f13 is performed by the color pixel determination 32
The output of 5f12 is blocked. That is, 4 × 4
If there is one or more color pixels 2 in the pixel matrix, the entire 4 × 4 pixel matrix is output as a color pixel 2 block. In the processing after the block formation 325f13, 4 × 4 pixels are set as one block and output in block units. The density of 325f14 indicates that an active condition (color pixel 2
If there are three or more blocks and the target block is active (color pixel), the target block is set as the active block (color pixel 2 block).

Next, the count 325f3 will be described. The number of 1s (c = 1, m = 1, y = 1) of c, m, and y in each pixel in the 5 × 5 pixel matrix centering on the pixel of interest is counted. The difference between the maximum value and the minimum value of the count value for each of c, m, and y is greater than or equal to thal1cnt, and the minimum value of the counted c, m, and y is thal.
If it is less than min, the color pixel candidate 3 is selected. thal1cn
t and tha1min are threshold values set before copying (processing).

The pattern matching 325f5 determines whether the pixel (c, m, y) determined by the color pixel detection is a color pixel by 5 ×.
It is determined by pattern matching using a 5-pixel matrix. The pattern is the same as that of the pattern matching 325f6. Pixels that match in pattern matching are color pixel candidates 4. Color pixel determination 325f15 determines color pixel 3 if it is color pixel candidate 3 and color pixel candidate 4.
The blocking 325f16 blocks the output of the color pixel determination 325f15. That is, if there is one or more color pixels 3 in the 4 × 4 pixel matrix, the entire 4 × 4 pixel matrix is output as a color pixel 3 block. The process after blocking 325f16 is
4 × 4 is set as one block and output in block units. The density of 325f17 is 3 × 3 because of the isolation block removal.
There are three or more active conditions (3 blocks of color pixels) in the block, and the block of interest is active (3 color pixels)
If so, the block of interest is the active block (color pixel 3
Block).

Next, the count 325f4 will be described. 1 of c, m, and y determined by the hue division 325a of each pixel in the 5 × 5 pixel matrix centered on the pixel of interest
Count the number of (c = 1, m = 1, y = 1). c,
If the minimum value of the count values of m and y is not less than thabk, the pixel of interest is set to black pixel candidate 1. thabk is a threshold value set before copying (processing).

The pattern matching 325f7 is c = m in the 5 × 5 pixel matrix centered on the target pixel.
= Y = 1 pixel pattern matching is performed. 1-1). Pattern 1-1 (pm1) D23 & D33 & D43 1-2). Pattern 1-2 (pm2) D32 & D33 & d34 1-3). Pattern 1-3 (pm3) D22 & D33 & D44 1-4). Pattern 1-4 (pm4) D42 & D33 & D24 These patterns are shown in FIG. 23, and the pixels circled in the figure are pixels with c = m = y = 1. When any one of these patterns matches, the target pixel is set as the black pixel candidate 2.

In the achromatic determination 325f18, if the pixel of interest is the black pixel candidate 1 and the black pixel candidate 2, it is determined to be a black pixel. The blocking 325f19 blocks the output of the black pixel. Blocking here means that if there is one or more black pixels in a 4 × 4 pixel matrix, the entire 4 × 4 pixel matrix is output as a black pixel block. The processing after blocking 325f19 is
4 × 4 pixels are set as one block and output in block units.
The expansion 325f20 makes the target block non-active (non-black pixel block) if the target block is active (black pixel block) and its peripheral pixels are non-active (non-black pixel block) in the matrix of 3 × 3 blocks. To do.

In the total color pixel determination 325f21, if the target block is not determined to be active (color pixel 2) in the color pixel determination 325f12 and is not active (black pixel) in the achromatic determination 325f18, the target block is colored. It is determined to be (color block). Also, color pixel determination 325f
When 15 is active (color pixel), it is determined to be a color. The expansion 325f22 is performed in the total color pixel determination 325f21 in order to regard small characters as continuous with respect to the block determined to be a color. Therefore, if even one block is active in the matrix of 9 × 9 blocks centering on the block of interest. , The block of interest is the active block. Here, the reason for greatly expanding is to fill the gap between the characters. In the continuous count 325f23, the continuity of the color pixel blocks is checked to determine whether it is a color original document or a monochrome original document. By counting the number of consecutive color pixels in the output data (color pixel block) of the expansion 325f22,
Determine if it is a color original.

FIG. 26 is a flow chart showing a processing procedure for color pixel determination. The contents of the determination process will be described with reference to FIG. If the pixel of interest is in the color pixel block, the number of consecutive color pixels of the pixel of interest is calculated by referring to the number of consecutive color pixels of the upper left, upper, upper right and left pixels of the pixel of interest (steps 21 to 26). Here, if the pixel of interest is, for example, c3 pixel of the 5 × 5 pixel distribution pattern MPp of FIG. 10, the upper left, upper, upper right, and left pixels are respectively
The pixels are b2, b3, b4 and c2. If the pixel of interest is not in the color pixel block, it is given a continuous number of color pixels of 0 (steps 21 to 27).

When the pixel of interest is in the color pixel block,
First, the number of consecutive color pixels of the upper pixel (b3) of the target pixel (c3) is checked (step 22), and if it is 0, the reference value A is set to 1 for the number of consecutive color pixels of the upper right pixel (b4). Is added (step 24), and if the number of consecutive color pixels of the upper pixel (b3) is 0, the reference value A is added to the upper right pixel (b
The continuous number of color pixels of 4) is given (step 23). next,
The reference value B is given a value obtained by adding 1 to the number of consecutive color pixels of the upper left pixel (b2), the reference value C is given a value obtained by adding 1 to the number of consecutive color pixels of the upper pixel (b3), and the reference value The left pixel (c
A value obtained by adding 1 to the number of consecutive color pixels in 2) is given (step 25). Then, the highest value of the reference values A, B, C and D is set as the number of continuous color pixels of the target pixel (c3) (step 26). When the number of consecutive color pixels is given to the pixel of interest (c3) as described above, the number of consecutive color pixels is set to the set value THACS.
Check if this is the case (step 28), and THA
If it is greater than or equal to CS (step 28; Y), it is determined that the original is a color original (step 29), and the processing of the continuous count 325f23 ends there. When the number of continuous color pixels is less than the set value THACS (step 28; N),
The pixel of interest is updated to the next pixel in the scanning directions x and y, and the above processing is repeated. As a result of the above-described processing for the entire surface of the original, when the number of continuous color pixels is less than the set value THACS until the end (steps 30 to 34), the original is determined to be a monochrome image.

The continuous number of color pixels is the sum of the vertical colored line segment and the horizontal colored line segment. The number of consecutive color pixels on the upper right is different from the others because double counting is prevented. FIG. 27
FIG. 6 is a diagram showing specific data on the number of continuous color pixels.
The small squares with the numbers shown in FIG. 27 are color pixels, and the numbers are the number of continuous color pixels given to the pixels. A block in which a number of small squares are connected is a color pixel group, and even if the number of continuous color pixels in any one of the color pixel groups on the same document is greater than or equal to the set value THACS, and it is a color document. , Color or monochrome is determined (step 28, step 29).

The reason for separating the color pixel determination from 325f8 to 325f15 is to increase the accuracy of determination of a color original or a black and white original. The color pixel determination for black character processing is local and not so noticeable even if an erroneous determination is made. On the other hand, the determination as to whether the original is a color original or a black-and-white original affects the entire original if an erroneous determination is made. Therefore, the counts 325f1 to f4 are made independent. Originally, it is better to be independent from the hue division 325a, but if the hue division 325a is made independent, the memory of the pattern matching 325f5 to f7 increases, which is not preferable. By increasing the parameters (color pixel candidates 1 and 3, black pixel candidate 1) of the counts 325f1 to f4 and changing the color pixel parameters (color pixels 1 to 3), the increase in the memory amount is reduced. Color pixel determination 325f12,
325f15 is provided to detect a low density color such as yellow of a fluorescent pen. Furthermore, achromatic determination (black pixel determination) 325f18 is provided when erroneous detection is made when the density is low. This is to make a correction. Colors with low density such as a highlighter can be corrected with black data within a certain width without any problem. When extracting a plurality of color pixels, the w (white) level is simply changed.
It is not necessary to have a memory for one, and it is possible with a capacity of one plus one line.

In the continuous count 325f23, since the count value is counted by referring to the count data of the previous line and the count data of the current line, it is possible to accurately count the continuity of the peripheral pixels, so that the number of color pixels can be accurately counted. It becomes possible to count the series. In the present embodiment, R, G, B
Although the hue determination is performed on the image data, it is not limited to the R, G, B image data, and the luminance color difference (for example,
It is easy to determine the hue for Lab).

Next, the comprehensive judgment 326 will be described.
FIG. 28 is a block diagram showing the structure of the comprehensive judgment 326. The comprehensive judgment 326 includes a character judgment 326a, an expansion processing 326b, and a decoding 326c. First, the character determination 326a will be described. In the character determination 326a, when the result of the edge extraction 322 is an edge, the result of the halftone dot extraction 324 is no halftone, and the result of the white area extraction 323 is a white area, it is determined to be a character edge. If not, it is determined as a non-character edge (in the pattern or character).

Next, the expansion process 326b will be described. The expansion process 326b outputs the result of the character determination 326a to 8
OR (logical sum) processing of × 8 blocks, then 3 × 3
An AND (logical product) process of blocks is performed to perform a dilation process of 4 blocks. That is, 8 centered on the block of interest
If any of the × 8 blocks is a character edge, it is assumed that the block of interest is also a character edge block, and if all 3 × 3 blocks centering on the block of interest are character edges, The block is defined as a character edge, and the block of interest and three blocks adjacent to it are regarded as the character edge. OR
AND processing is performed after processing, especially in the case of black characters,
When a small non-black character area exists around the black character area, a sense of discomfort may be felt due to the difference in processing. For example, black looks light. To prevent this, the non-black character area is enlarged by OR processing. The AND process is performed to obtain the desired expansion amount.

By the way, the color copying machine scans four times to make one copy.
The character determination result is slightly different. In particular, if non-black character determination is performed during black image formation and black character determination is performed during non-black image formation, this black character area becomes thin, so
At the time of bk, OR processing of 8 × 8 blocks is performed, and then AND processing of 3 × 3 blocks is performed. Further, at the time of image formation other than bk, OR processing of 5 × 5 blocks is performed, and thereafter, AND processing of 1 × 1 blocks is performed. 1 x 1 A
The ND processing is synonymous with the fact that no processing is performed because the result is the same as that before the processing. The result of the expansion process 326b is output to the decode 326c as a character edge signal. In this way, the expansion process 32
By performing 6b, the separation result is not different and the character area is not thinned. Further, the expansion process 326b may darken the middle part of the character, but if the character is thin with respect to the edge of the character, the density will be saturated, so that there is no discomfort.

FIG. 29 is a schematic enlarged view showing the overlap of color coloring agents in color copying. FIG. 29
(D) shows an ideal case where black characters are processed for all four colors. In FIG. 29E, black characters are processed for all four colors,
It shows the case where the correction is not applied and the correction is applied to the areas other than bk to make the area thinner. FIG. 29 (f) shows a suitable case in which only bk is subjected to black character processing according to the present embodiment, and FIG. 29 (g) shows that only bk is subjected to black character processing according to the present embodiment and only bk is corrected. No, it shows a suitable case where the correction is applied for other than bk. FIG. 29A shows an ideal case where the expansion amount is the same and black character processing is performed.
FIG. 29B shows a case where the expansion amount is the same and black character processing is performed to shift the printing position (whitening out). FIG. 29
(C) shows the case where the expansion amount of bk is large, and the printing position is shifted by the black character processing according to the present embodiment.

Next, the decoding 326c will be described. The C / P signal finally output from the decode 326c is as shown in the table shown in FIG. In the table,
X may take any value. Here, FIG. 3 is referred to again. C / P signal and B generated by document recognition 320
The / C signal is an RGB filter 330, color correction 340, scaling 350, interface 352, UCR 360, C
The MYBk filter 370, the CMYBkγ correction 380, and the gradation processing 390 are applied to the cascade in synchronization with the image data.

The RGB filter 330 is a filter for MTF-correcting RGB data, and is composed of a coefficient matrix corresponding to an N × N pixel matrix and a logic for multiplying each coefficient by each image data to obtain a weighted average value. There is. When the C / P signal represents one (character edge area), a coefficient matrix for sharpening processing is used,
0 or 2 or 3 (pattern area, low-frequency halftone dot area,
In the case of a halftone dot area), a weighted average value is derived using a coefficient matrix for smoothing processing and is output to the color correction 340. When the smoothing filters here are arranged in descending order of the smoothing amount, they become a low frequency halftone dot area, a halftone dot area, and a pattern area. This is because the halftone dot structure will remain unless the smoothness is strengthened, which causes moire. Further, the low frequency halftone dots need to be smoothed more strongly than the high frequency halftone dots.
The color correction 340 converts the R, G, B data into C, M, Y data by a primary masking process or the like. Magnification change 350
The image data is enlarged / reduced in the main scanning direction x or is subjected to equal-magnification processing.

The UCR 360 is for improving the color reproduction of the image data, and UCR (additive removal) processing is performed on the common portion of the C, M and Y data input from the color correction 340 to generate Bk data. , C, M, Y, Bk data are output. Here, when the C / P signal is other than 1 (character edge area) (in a character area or a picture area), skeleton black processing is performed. When the C / P signal is 3 (character edge area), full black processing is performed. Further C / P
When the signal is 1 (character edge area) and the B / C signal is H (achromatic area), the data of C, M and Y are erased. This is because in the case of a black character, it is expressed only by the black component.

Also, the output image signal IMG of the UCR 360
Indicates that the temporary point is one color of C, M, Y, and Bk, and is a frame sequential one-color output. That is, full-color (four-color) data is generated by reading the document four times. Further, in the case of black-and-white copying, one Bk image formation is sufficient, so that one-time document reading is sufficient. If there is a color original or black-and-white original determination mechanism, the number of times of reading depending on the original is sufficient, so that the operator does not need to determine whether the original is a color original or a black-and-white original and make a copy. In the present embodiment, the B / C signal is a signal that is referred to when determining whether a color original or a monochrome original. When the B / C signal is H (achromatic area) on the entire surface of the original, the main controller 10 determines that the original is black and white.

The CMYBk filter 370 is N × N according to the frequency characteristic of the color printer 400 and the C / P signal.
Smoothing and sharpening processing is performed using the spatial filter of. C
The MYBkγ correction 380 changes and processes the γ curve according to the frequency characteristic of the color printer 400 and the C / P signal. When the C / P signal is other than 1 (other than the character edge area), the γ curve that faithfully reproduces the image is used, and when the C / P signal is 1 (the character edge area), the γ curve is raised to increase the contrast. Emphasize. The gradation processing 390 performs quantization such as dither processing and error diffusion processing according to the gradation characteristics of the color printer 400 and the C / P signal. In the case of Bk image formation, when the C / P signal is other than 1 (other than the character edge area), gradation-oriented processing is performed, and otherwise resolution-oriented processing is performed. In the case of an image other than Bk, when the C / P signal is 0 (pattern area), gradation-oriented processing is performed, and in other cases, resolution-oriented processing is performed. The image data that has been subjected to the above processing is supplied from the video control 359 having a buffer memory to the color printer 400 in synchronization with the image data writing operation.

The IPU 300 performs a smoothing process by the RGB filter 330, a skeleton black process by the UCR 360, and a linear (gradation by the CMYBkγ correction 380) in a case other than the character area (C / P signal = 1). )
Select a curve that emphasizes CMYBk filter 370
In the gradation processing 390, processing with emphasis on gradation is performed.
On the other hand, in the case of character processing (C / P signal = 1 and B / C signal = L), the RGB filter 330 performs edge enhancement processing, the UCR 360 performs full black processing, and CMYB
In the kγ correction 380, a curve that emphasizes contrast is selected, and the CMYBk filter 370 and gradation processing 390 are selected.
Then, the processing that emphasizes the resolution is performed.

Black character processing (B / B when C / P signal = 1)
C signal = H), C, M, Y data is not printed at the time of C, M, Y image formation except Bk. This is to prevent the surroundings of black characters from being colored due to misalignment. In addition, the RGB filter 330 of the Bk data at this time
You can make the edges sharper and sharper than with colored letters. In this way, the IPU 300 carries out four types of processing including a pattern, a character edge, a halftone dot, and a low-frequency halftone dot.

[0118]

According to the first aspect of the present invention, there are provided a filter calculation means for performing a filter calculation of a plurality of filters, and a filter coefficient determination means for determining a filter coefficient when the correction means makes a correction based on the calculation result. Since it is equipped with 400
Filter coefficient not emphasizing the parallel line period of dpi and 600d
The filter operation can be performed with a filter coefficient of pi that does not emphasize the line cycle, and the filter coefficient can be switched according to the result. The pattern can be non-letter.

According to the second aspect of the invention, the filter coefficient determining means determines the filter coefficient based on the minimum value of the image density of the calculation result of the filter calculating means.
The filter coefficient that does not emphasize the line cycle of 00 dpi and 60
The filter coefficient can be switched with a filter coefficient that does not emphasize the parallel line cycle of 0 dpi, and the filter coefficient can be switched according to the result. The line pattern can be made non-character.

According to the third aspect of the invention, the filter calculation means for performing the filter calculation of the plurality of filters, and the filter coefficient determination for determining the filter coefficient when the correction means performs the correction based on the calculation result by the filter calculation means. And a region extracting unit that performs a white background region extracting process of the image data based on the correction result by the correcting unit. so,
Filter calculation is performed, and the filter coefficient can be switched according to the result. In the white area detection, a filter that does not emphasize white can be added to prevent the line pattern from becoming characters.

According to the fourth aspect of the invention, the filter coefficient determining means determines the filter coefficient based on the maximum value of the image density of the calculation result of the filter calculating means.
The filter coefficient that does not emphasize the line cycle of 00 dpi and 60
A filter operation can be performed with a filter coefficient that does not emphasize the line interval of 0 dpi, and the filter coefficient can be switched according to the result. In the white area detection, a filter that does not emphasize white is added to add a line pattern character. You can choose not to.

According to the fifth aspect of the present invention, since the original image is color-separated and read, color image data reading means for generating image data and transmitting the image data to the image data receiving means is further provided. It can also be applied to a processing device.

According to the sixth aspect of the invention, since it further comprises a color image forming means for forming a color image of the image data on the sheet, it can be applied to a color printer.

According to the seventh aspect of the present invention, a color image data reading means for color-separating an original image to read the image, generating image data and transmitting the image data to the image data receiving means, and forming a color image on the paper with the image data. It is also possible to apply the present invention to a color copying machine by further including a color image forming means.

According to an eighth aspect of the invention, there is further provided a print instruction receiving means for receiving a print instruction from an external device, wherein the image data receiving means receives the image data from the external device and forms the color image. Since the means executes the color image formation by the color image forming means based on the print instruction received by the print instruction receiving means, it can be applied to a color copying machine.

[Brief description of drawings]

FIG. 1 is a diagram showing a schematic configuration of an image processing apparatus according to an embodiment.

2 is a block diagram showing a schematic configuration of an electric system of the color copying machine of FIG.

FIG. 3 is a block diagram showing a configuration of an image processing unit (IPU).

FIG. 4 is a block diagram showing a functional configuration of document recognition in FIG.

FIG. 5 is a line pattern A of 600 dpi and 400 dpi.
It is the figure which showed the parallel line pattern B of.

FIG. 6 is a diagram showing an example of coefficient groups.

FIG. 7 is a pattern of 3 × 3 pixel matrix BPa to BP.
It is the figure which showed d and WPa-WPd.

FIG. 8 is a block diagram showing a configuration of white area extraction.

FIG. 9 is a pattern WBP, RDPa, RDPb, BBP.
It is the figure which showed.

FIG. 10 shows patterns MPp, MCa, MCb, MCc,
It is the figure which showed MCd.

FIG. 11 is a flowchart showing a processing procedure for updating state variables MS and SS [I] used for white determination.

FIG. 12: Line memory LMP, patterns PMPa, P
MPb, PMPc, PMPd and block pattern B
It is the figure which showed CP.

FIG. 13 is a diagram showing a line pattern of 200 lines and 300 lines.

FIG. 14 is a diagram showing black protrusions Bp1 to Bp4.

FIG. 15 is a diagram showing adjustment values of thresholds thwss, thwsb, and thwc.

FIG. 16 is a diagram showing a configuration of halftone dot extraction.

FIG. 17 is a diagram showing a case where a mask of a 5 × 5 pixel matrix is adopted as a local area.

FIG. 18 is a diagram showing a case where a mask of a 7 × 7 pixel matrix is adopted as a local area.

FIG. 19 is a diagram showing a pattern of a 7 × 7 pixel matrix.

FIG. 20 is a diagram showing halftone dot peaks corrected by a cycle check.

FIG. 21 is a diagram showing an image density signal.

FIG. 22 is a diagram showing a configuration of color pixel determination.

FIG. 23 is a diagram showing patterns pm1 to pm4.

FIG. 24 is a diagram showing patterns pw11a to pw14d.

FIG. 25 is a diagram showing patterns pw21a to pw24d.

FIG. 26 is a flowchart showing a processing procedure for color pixel determination.

FIG. 27 is a diagram showing specific data on the number of continuous color pixels.

FIG. 28 is a diagram showing a configuration of comprehensive determination.

FIG. 29 is a schematic enlarged view showing the overlap of color coloring materials by color copying.

FIG. 30 is a diagram showing a C / P signal output by a decode.

[Explanation of symbols]

320 Document recognition 321 Filter 322 Edge extraction 323 White area extraction 324 Halftone dot extraction 325 color judgment 326 Overall judgment

   ─────────────────────────────────────────────────── ─── Continued front page    F term (reference) 5B057 CE06 CE16 CH09 CH11 DC16                 5C077 MP05 MP07 MP08 PP01 PP32                       PP33 PP43 PP47 PQ22 TT06                 5C079 HB01 HB03 HB12 LB02 MA01                       NA02 PA02

Claims (8)

[Claims] 1. An image data receiving means for receiving image data, a correcting means for correcting the image data received by the image data receiving means with a plurality of filters, and a filter calculating means for performing a filter operation of the plurality of filters. A filter coefficient storage unit that stores filter coefficients of the plurality of filters; a filter coefficient determination unit that determines a filter coefficient when the correction unit performs correction based on a calculation result by the filter calculation unit; An image processing apparatus comprising: an edge extraction unit that performs an edge extraction process of the image data based on a correction result by the correction unit. 2. The image processing apparatus according to claim 1, wherein the filter coefficient determination means determines the filter coefficient based on a minimum value of the image density obtained as a result of the calculation by the filter calculation means. 3. An image data receiving means for receiving image data, a correcting means for correcting the image data received by the image data receiving means with a plurality of filters, and a filter calculating means for performing a filter operation of the plurality of filters. A filter coefficient storage unit that stores filter coefficients of the plurality of filters; a filter coefficient determination unit that determines a filter coefficient when the correction unit performs correction based on a calculation result by the filter calculation unit; An image processing apparatus comprising: an area extracting unit that performs a white background area extracting process of the image data based on a correction result by the correcting unit. 4. The image processing apparatus according to claim 3, wherein the filter coefficient determination means determines the filter coefficient based on the maximum value of the image density obtained as a result of the calculation by the filter calculation means. 5. A document image is color-separated and read by the image processing apparatus according to any one of claims 1, 2, 3, and 4, and image data is generated to generate the image data. An image reading apparatus further comprising color image reading means for transmitting to a receiving means. 6. The image processing apparatus according to claim 1, further comprising color image forming means for forming a color image of the image data on a sheet. An image forming apparatus characterized by being provided. 7. An image processing apparatus according to claim 1, wherein the original image is color-separated and read, and image data is generated to generate the image data. Color image data reading means for transmitting to the receiving means,
A color copying apparatus, further comprising: a color image forming unit that forms a color image of the image data on a sheet. 8. The image forming apparatus according to claim 6 or the color copying apparatus according to claim 7, further comprising print instruction receiving means for receiving a print instruction from an external device, wherein the image data receiving means is the external device. A color copying apparatus, wherein the color image forming unit receives the image data from the apparatus, and the color image forming unit executes the color image formation by the color image forming unit based on the print instruction received by the print instruction receiving unit. .