JP2007104009A - Device and method for forming image - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve the precision of gradation correction in an image-forming device for forming a toner image half-toned by an error diffusion method. <P>SOLUTION: Reference patterns P1-P3 having different dot patterns are formed on an intermediate transfer belt (S1), and the density of a portion (patch) corresponding to each gradation value of respective reference patterns P1-P3 is detected for sampling (S2). Then, the average value of three density values sampled corresponding to each gradation value is obtained (step S3), and gradation characteristics, which show the relationship between input image data and the density of an output image from the average density, are obtained. A γ curve for correcting so that the gradation characteristics become linear is obtained (step S4), and the gradation of the input image data is corrected, based on the γ curve, thus forming an image having excellent gradation reproducibility. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an image forming apparatus and an image forming method for forming a toner image expressed in gradation by an error diffusion method on a transfer medium.

  In general, the electrophotographic image forming apparatus has a problem that the gradation of the input image is not reproduced linearly in the output image due to the photosensitive characteristics of the photosensitive drum and the color development characteristics of the toner. For this reason, conventionally input image data is subjected to gradation correction based on a gradation correction curve (hereinafter referred to as “γ curve”) to improve the gradation reproducibility of the output image.

However, the amount of toner adhering to the photosensitive drum during development changes due to deterioration of the photosensitive drum, changes in temperature and humidity in the apparatus, and changes in charging characteristics due to toner deterioration. It is desirable to periodically update the contents of the key correction.
Therefore, for example, in the invention described in Patent Document 1, a toner pattern is periodically formed on a transfer medium based on image data having a predetermined gradation value, and the density of the toner pattern is detected by a photoelectric sensor. The tone reproducibility is improved by obtaining a γ curve each time based on the detection result and correcting the tone value of the input image data using the γ curve.

  On the other hand, an error diffusion method has become widespread as a method for expressing gradation in toner images (for example, Patent Document 2). In this error diffusion method, the target pixel of the input image is reduced (for example, binarized) by a certain threshold value, and the difference (error) between the reduced value and the gradation value of the input image is measured. This is a method of expressing the pseudo gradation by reducing the value of the target pixel one after another while dispersing and compensating for the pixels, and is particularly suitable for multi-tone image expression.

Even in such an image forming apparatus that expresses gradation by the error diffusion method, it is desirable to periodically update the gradation correction as described above, as long as the image is formed by transferring toner. Not too long.
Japanese Patent Laid-Open No. 5-14728 Japanese Patent Publication No. 7-93684

  However, in the error diffusion method, pseudo gradation expression is performed by the distribution (hereinafter referred to as “dot pattern”) of pixels (white pixels and black pixels) whose values are reduced (for example, binarized) as described above. In addition, since the black pixels are randomly distributed, the detection value varies depending on the detection position of the toner pattern even in the region showing the same gradation value. Therefore, accurate gradation correction becomes difficult.

Particularly, in the low density portion, since black pixels are considerably unevenly distributed as shown in FIG. 14, when the reading position of the pattern detection sensor is different from, for example, A, B, and C in FIG. The correlation (gradation characteristics) between the gradation data level of the input image and the density detection value of the output image varies as shown in FIG.
Therefore, the γ curve obtained to complement this gradation characteristic also varies as shown in FIG. 16 in the low density region, and is formed on the recording sheet based on the image data corrected by the γ curve. As a result, the gradation reproducibility of the image is not satisfactory in the low density region as shown in FIG.

  The present invention has been made in view of the above circumstances, and an object thereof is to improve the accuracy of gradation correction in an image forming apparatus that expresses gradation using an error diffusion method.

  In order to achieve the above object, an image forming apparatus according to the present invention performs error diffusion processing under different conditions on each of image data for patch formation for forming a plurality of toner patches having different gradation values, Toner patch forming means for forming a plurality of toner patches having different dot patterns for each gradation value, and for the plurality of toner patches formed for each gradation value, the density of each partial region is detected and Representative density acquisition means for acquiring a representative density of a toner patch indicating each gradation value from the detected density, and gradation correction for generating gradation correction data based on the representative density value and correcting the gradation of the input image data And image forming means for processing the input image data subjected to gradation correction by an error diffusion method to form an image.

  Here, the toner patches indicating the respective gradation values may be formed continuously in the sub-scanning direction, or may be formed separately at a predetermined interval. “To detect density” of a toner patch is not only when the density value is directly detected by a detection means such as a photoelectric sensor, but also by obtaining the toner adhesion amount from the detection signal of the detection means, and using this as the density value. The case of conversion is also included.

  In the present invention, a plurality of toner patches having different dot patterns are formed for the same gradation value by performing error diffusion processing on the patch formation image data for each gradation value under different conditions. Then, the density of a partial region of each toner patch is detected, and the representative density of the toner patch indicating each gradation value is obtained from the detected density. In this way, a plurality of toner patches are formed for the same gradation value, and since the dot patterns thereof are different, the uneven distribution states of dots in the detection areas of the respective toner patches are also different.

  Therefore, if the representative density of the toner patch indicating the gradation value is acquired based on the detected density, the uneven distribution state of the dots can be evaluated with some degree of uniformity, and at least as in the conventional case, each level is evaluated. To obtain a toner patch density that is less susceptible to the uneven distribution of dots than when only one specific toner patch formed for a tone value is detected to create gradation correction data. Therefore, it is possible to perform gradation correction with high accuracy.

Here, the representative density acquisition unit acquires an average value of the densities detected for the plurality of toner patches formed for each gradation value as a representative density of the toner patch indicating the gradation value, It is possible to stably acquire the toner patch density by equalizing the uneven distribution of dots due to the error diffusion process.
In addition, the toner patch forming unit performs error diffusion processing on the data of a specific pixel in the image data for patch formation, and an error between a gradation value of the pixel and a value obtained by lowering the value. By changing this, toner patches having different dot patterns may be formed for the same gradation value. At this time, the specific pixel may be the first pixel read from the patch image data.

Further, the toner patch forming means, when performing error diffusion processing on the image data for patch formation, varies the position of the pixel from which the error diffusion processing is started, so that different dots are generated for the same gradation value. A pattern toner patch may be formed.
Further, the toner patch forming means affects the sum of the gradation values of each pixel in the predetermined pixel block unit for the patch forming image data when performing error diffusion processing on the patch image data. It is also possible to form toner patches having different dot patterns for the same gradation value by overlapping noise that is not given.

Hereinafter, an embodiment of an image forming apparatus according to the present invention will be described by taking a tandem type color digital copying machine (hereinafter simply referred to as “copying machine”) as an example.
<First Embodiment>
(1) Outline of Configuration of Copying Machine FIG. 1 is a diagram showing a configuration of a main part of a copying machine 1 according to the present embodiment.

The copying machine 1 is roughly composed of an image reading unit 100 and an image forming unit 200.
As the image reading unit 100, a known unit that reads a document image with a CCD sensor and generates R, G, B image data is used, and a mirror that scans the document placed on the document table by moving the scanner. There are a scanning type and a sheet-through type that reads a document image while moving the document with a document conveying device while fixing the scanner, but there is no particular limitation as long as the document image can be read in color.

  In this embodiment, the image forming unit 200 employs a tandem type and a secondary transfer system, and includes an intermediate transfer unit 10, an image forming unit 20, a paper feeding unit 30, a secondary transfer unit 40, and a fixing unit. 50 and a control unit 60, and a full-color image is formed by multiple transfer of cyan (C), magenta (M), yellow (Y), and black (K) reproduced toner images. (Hereinafter, each reproduction color of cyan, magenta, yellow, and black is simply expressed as C, M, Y, K, and this C, M, Y, K is added to the number of the component related to each reproduction color. Add as).

  The intermediate transfer unit 10 is configured by stretching an intermediate transfer belt 15 around a driving roller 11 and driven rollers 12, 13, and 14, and positions inside the intermediate transfer belt 15 corresponding to the photosensitive drums 22Y to 22K. Are provided with primary transfer chargers 16Y to 16K. The intermediate transfer belt 15 is configured to run at a constant speed by the driving roller 11. A cleaner 17 for removing toner remaining on the surface of the intermediate transfer belt 15 is disposed below the driving roller 11.

The image forming unit 20 is formed by arranging four electrophotographic image forming units 21 </ b> Y to 21 </ b> K along the traveling direction of the intermediate transfer belt 15.
Here, for example, the image forming unit 21Y includes a photosensitive drum 22Y, a charging charger 23Y, an LED array 24Y, a developing unit 25Y, and a cleaner 26Y. The surface of the photosensitive drum 22Y from which the residual toner on the surface has been removed by the cleaner 26Y is uniformly charged by the charging charger 23Y, and this is exposed and scanned by the LED array 24Y, whereby an electrostatic latent image is formed on the surface of the photosensitive drum 21Y. Is formed. The electrostatic latent image is developed with yellow toner supplied from the developing unit 25Y, and the developed toner image is primarily transferred onto the surface of the intermediate transfer belt 15 by the transfer charger 16Y.

The other image forming units 21M, 21C, and 21K basically have the same configuration except that the color of the toner stored in the developing unit is different.
The sheet feeding unit 30 is configured to supply a sheet feeding cassette 31 that stores recording sheets, and a timing at which the recording sheet is sent out from the sheet feeding cassette 31 one by one to the pickup roller 32 and the intermediate roller 33 and the secondary transfer unit 40. A registration roller 34 and the like are provided.

The secondary transfer unit 40 secondarily transfers the toner image on the intermediate transfer belt 15 onto the recording sheet S by electrostatic force generated by the voltage applied to the transfer roller 41.
The fixing unit 50 fixes the toner image transferred onto the recording sheet by thermocompression bonding onto the recording sheet.
The control unit 60 performs various processes including gradation correction on the image data acquired by the image reading unit 100 and the image data received from an external terminal, and reproduces C, M, Y, and K reproduction color signals. In addition, the operation of each part of the image reading unit 100 and the image forming unit 200 is controlled to perform a smooth image forming operation.

  An operation panel 70 (FIG. 2) is provided at a position on the upper surface of the copying machine 1 that is easy to operate. On the operation panel 70, a key group for setting the number of copies, a key group such as a start key for starting copying, and a setting for receiving input such as a document reading mode and a reading magnification (enlargement / reduction ratio). A liquid crystal display unit for displaying messages such as a screen, a button, and a paper out or a paper jam is arranged.

  In the copying machine 1 having the above-described configuration, the image data of the R, G, and B color components obtained by reading the document by the image reading unit 100 is subjected to various data processing by the control unit 60, and further the C , M, Y, and K are converted to image data of each reproduction color, and the exposure LED arrays in the image forming units 21Y to 21K in the image forming unit 30 are driven with a predetermined time lag, and the respective photosensitive drums 22Y to 22Y are driven. A toner image of each reproduction color is formed on 22K.

The C, M, Y, and K toner images formed by these image forming units 21 </ b> Y to 21 </ b> K are sequentially superimposed on the same position of the intermediate transfer belt 15 and transferred, so that a full color image is formed on the intermediate transfer belt 15. The toner image is formed.
The paper feeding unit 30 feeds the recording sheet S to the secondary transfer unit 40 at the timing when the toner image formed on the intermediate transfer belt 15 reaches the secondary transfer unit 40, where the intermediate transfer is performed. The toner image on the belt 15 is secondarily transferred onto the recording sheet S. After that, the toner image is fixed by the fixing unit 50 and then discharged onto the paper discharge tray 55.

  A pattern detection sensor 18 for detecting a toner pattern formed on the intermediate transfer belt 15 is disposed at a position along the intermediate transfer belt 15 on the downstream side of the secondary transfer unit 40. As the pattern detection sensor 18, for example, a reflective photoelectric sensor including a light emitting element such as an LED and a light receiving element such as a photodiode is used, and gradation correction is performed based on the detection result of the pattern detection sensor 18. The Details will be described later.

(2) Configuration of Control Unit 60 Next, the configuration of the control unit 60 will be described with reference to FIG.
As shown in the figure, the control unit 60 includes a CPU 61, an image signal processing unit 62, an image memory 63, an LED driving unit 64, a RAM 65, a ROM 66, an EEPROM 67, and the like.
The image signal processing unit 62 converts R, G, and B electrical signals obtained by scanning a document to generate image data composed of multi-valued digital signals, and performs various processes to be described later. , C, M, Y, K reproduction color image data is generated and output to the image memory 63. The image memory 63 stores the image data for each reproduced color.

Under the control of the CPU 61, the LED array driving unit 64 reads image data for each scanning line from the image memory 63, and drives the LED arrays in the image forming units 21Y to 21K.
The RAM 65 temporarily stores various control variables and provides a work area when executing the program.

In the ROM 66, in addition to control programs for the respective units of the image reading unit 100 and the image forming unit 200, printing data for toner patterns of each color is stored.
The EEPROM 67, which is a non-volatile writable memory, stores correction data and the like in gradation correction data creation processing described later.
FIG. 3 is a block diagram showing a configuration of the image signal processing unit 62 of FIG.

The image signal photoelectrically converted by the CCD sensor of the image reading unit 100 is converted by the A / D conversion unit 621 into r, g, b multi-value digital image data.
The A / D converted image data is subjected to predetermined shading correction in a shading correction unit 622. This shading correction eliminates the unevenness of the exposure lamp of the scanner and the unevenness of the sensitivity of the CCD sensor in the image reading unit 100, and reads the white reference plate on which the original glass table is placed during prescanning. The multiplication ratio of each pixel is determined from the read image data and stored in an internal memory (for example, the RAM 65), and is corrected by multiplying each image data by the multiplication ratio stored in the internal memory when reading a document. is there.

Since the shading-corrected image data is the document reflectance data, the LOG converter 623 converts the image data into actual image density data Dr, Dg, and Db.
The copying machine 1 can be connected to a network such as a LAN via an interface (I / F) 632 and can receive image data from other terminals connected to the network. When image data is received from an external terminal, they are often R, G, and B density data. Therefore, the data processing in the A / D conversion unit 621, shading correction unit 622, and LOG conversion unit 623 described above is particularly important. unnecessary.

  The density data Dr, Dg, and Db are input to a BP processing unit (black printing processing unit) 624, and black data K is generated. The reason for generating black here is that even if black is reproduced by overlaying cyan, magenta, and yellow necessary for full color reproduction, it is difficult to reproduce clear black due to the spectral characteristics of each toner. It is. Therefore, a certain ratio of the overlapping density of the density data Dr, Dg, and Db of the R, G, and B components obtained from the LOG conversion unit 623 is output as the black density data K.

The UCR processing unit (under color removal processing unit) 625 multiplies the density data Dr ′, Dg ′, Db by multiplying a value obtained by subtracting the black density data from the density data Dr, Dg, Db by a predetermined coefficient. Output as Db ′.
R, G, B and C, M, Y are complementary colors and should have the same density. Actually, however, the transmission characteristics of the filters R, G, B in the CCD sensor and the toner C of the printer unit , M, and Y do not change linearly, the color correction unit 626 multiplies the density data Dr ′, Dg ′, and Db ′ by a predetermined masking coefficient and applies linear correction to obtain C, M and Y density data are generated and output.

In this manner, the C, M, Y, and K density data obtained by the BP processing unit 624, the UCR processing unit 625, and the color correction unit 626 are input to the next stage γ correction unit 627 and stored therein. The gradation is corrected based on the gradation correction data (hereinafter referred to as “γ correction data”).
The tone-corrected C, M, Y, and K density data are subjected to error diffusion processing in the error diffusion processing unit 628 and then stored in the image memory 63 for each reproduced color.

FIG. 4 is a block diagram showing a configuration of the error diffusion processing unit 628.
As shown in the figure, the error diffusion processing unit 628 includes an input correction unit 6281, a lowering unit 6282, a difference calculation unit 6283, an error diffusion matrix unit 6284, an error value calculation unit 6285, and an error data update unit 6286. Composed.
The input correction unit 6281 adds an error distributed and accumulated from surrounding pixels to the target pixel of the image data. The lowering unit 6282 outputs the image data of the target pixel as quantized data using the threshold data.

The difference calculation unit 6283 is an error between the tone value of the image data output from the input correction unit 6281 and the tone value of the quantized data output from the lowering unit 6282 (hereinafter referred to as “lowering error”). )).
The error diffusion matrix unit 6284 is a functional element that distributes the lowering error of the image data according to the weight assigned to each peripheral pixel. Here, weights are formed in a matrix in order to distribute the error of the gradation value calculated by the difference calculation unit 6283 to surrounding pixels. For example, the weights immediately above and right next to the target pixel are formed. The pixel needs to reflect the error of the target pixel as much as possible, and the weight is set to “10 times” larger than other peripheral pixels. The error distribution value of each pixel weighted in this way is divided by the sum of the weights (“64” in the present embodiment) in the error calculation unit 6285 and sent to the error data update unit 6286.

The error data updating unit 6286 accumulates the error distribution value due to the lowering error of the other target pixel so far for each pixel and stores it as error data, and stores the current error distribution value as the stored error. Update in addition to the distribution value.
When the image data of the next target pixel is input to the input correction unit 6281, the error data stored for the pixel is output to the input correction unit 6281. The input correction unit 6281 The error data is added to the gradation value, and the process of lowering the value and distributing the error is repeated based on the value.

  In this way, the error diffusion process lowers the gradation value of each pixel of interest, and adds the error between the lowered data and the gradation value before the reduction to the surrounding pixels by weighted averaging. As a result, the gradation value of the next pixel is corrected, so that each pixel is different from the gradation value of the input pixel, but when viewed as a set of pixels, The gradation value can be sufficiently reflected.

Returning to FIG. 3, the image signal processing unit 62 further includes a pattern storage unit 629, a detection processing unit 630, a γ correction data calculation unit 631, and an error initial value determination unit 633 that are used in γ correction data creation processing described later.
The pattern storage unit 629 stores reference pattern image data (hereinafter referred to as “pattern data”) as shown in FIG. 5 for each of C, M, Y, and K.

  In FIG. 5, the gradation of the reference pattern seems to continuously change in the sub-scanning direction from the maximum density portion to the minimum density portion, but actually, for example, 250, 240, 230, 220,. .., 40, 30, 20, 10, 0, etc., which are divided into 26 equal parts for every 10 gradation values and change in stages (in this specification, a portion having a width for each 10 gradation values is Is referred to as “patch” for convenience, and the image data of the patch portion is referred to as “patch data.” Therefore, the reference patterns in FIG. 5 are 250, 240, 230, 220,. , It can be expressed that patches of a predetermined width indicating a gradation value of 0 are formed continuously in the sub-scanning direction, and the pattern data is composed of patch data indicating a plurality of gradation values).

The width of one patch in the sub-scanning direction is equal to or larger than the detection width of the pattern detection sensor 18 in the sub-scanning direction, and the density of the patch is detected by the pattern detection sensor 18 while the intermediate transfer belt 15 is running. Is set to the required width.
The error initial value determination unit 633 adjusts the lowering error of the first target pixel of the pattern data read from the pattern storage unit 629.

  In the error diffusion method, the error of the pixel of interest read sequentially is diffused one after another to the surrounding pixels, and the gradation is expressed in a pseudo manner by a large number of pixels. Even if the adjustment is made, the dot pattern changes although it hardly affects the gradation value that is pseudo-represented by the toner image formed based on the patch data (hereinafter referred to as “toner patch”).

  Therefore, after the pattern data of the reference pattern read from the pattern storage unit 629 is input to the error diffusion processing unit 628 and subjected to error diffusion processing, the pattern data is temporarily stored in the image memory 63, and when the γ correction data is generated, the pattern data The data is read and three reference patterns P1 to P3 are formed at predetermined intervals in the sub-scanning direction as shown in FIG. 6. At this time, the error initial value determination unit 633 performs the error diffusion process. By making the error initial value different between the reference patterns P1 to P3, each patch has a different dot pattern while displaying the same gradation value.

FIG. 7 schematically shows how toner patches having different dot patterns are formed. In this figure, for easy understanding, it is assumed that each of the reference patterns P1 to P3 is composed of five patches indicating gradation values of five levels (for example, gradation values of 30, 80, 130, 180, and 230).
When the pattern data is subjected to error diffusion processing, by setting the initial error values to different “a”, “b”, and “c” for the reference patterns P1 to P3, for example, the gradation value 30 of each reference pattern P1 to P3 is set. When the patches P11 to P31 shown are respectively enlarged, it becomes like P11 to P31 shown on the left, and it can be seen that the dot patterns are clearly different. Here, the circles in the enlarged patches P11 to P31 show examples of detection areas by the pattern detection sensor 18, and in this way, the patches having the same gradation value have relatively the same detection areas. The dot pattern is different.

The reference patterns P <b> 1 to P <b> 3 are formed for each color of C, M, Y, and K, and the density at the relatively same position is detected by the pattern detection sensor 18 for each patch.
The γ correction data calculation unit 631 obtains gradation characteristics from the relationship between the gradation value to be indicated by each patch and the representative density of the actual patch, and calculates γ correction data so that this relationship is linear. The γ correction data in the γ correction unit 627 is updated with the γ correction data calculated here, and thereafter, tone correction of the image data is executed based on the updated γ correction data.

  Thus, based on the patch data showing the same gradation value, toner patches having different dot patterns are formed, the density values detected for these are averaged, and this is represented as a representative toner patch showing the gradation value. In terms of density, it is possible to obtain a patch density that is less susceptible to the uneven distribution of dot patterns compared to the case where only one density is detected for one toner patch showing one gradation value as in the prior art. Can do.

However, in order to obtain a more appropriate patch density, the initial error values a and b are set so that none of the dot patterns in the detection region of each patch is biased to either “dense” or “sparse”. , C should be appropriately adjusted in advance.
(3) Content of γ Correction Data Creation Process The procedure of the γ correction data creation process will be described below with reference to the flowchart of FIG. 8, taking as an example the case where the image data to be corrected is cyan. Note that this γ correction data creation processing is executed, for example, every time power is turned on to the copying machine 1, every time a predetermined time elapses, or every time the cumulative number of image forming operations exceeds a predetermined number. .

  First, error diffusion processing is performed on cyan pattern data to form a plurality of reference patterns (three in this embodiment, P1 to P3) on the intermediate transfer belt 15. At this time, the patch data of each gradation value of the reference pattern is subjected to error diffusion processing while changing the initial error value as described above, thereby forming reference patterns P1 to P3 having different dot patterns for each patch (step S1). .

  More specifically, for example, when forming the reference pattern P1, the cyan pattern data is read from the pattern storage unit 629, and the lowering error for the first pixel is determined for each patch value of each gradation value. Regardless of the value obtained by the calculation value of the difference calculation unit 6283 (see FIG. 4), the value “a” read from the error initial value determination unit 633 is set, and the data of the second and subsequent pixels for the patch is set. Normal error diffusion processing is performed.

Since the size of the patch data of each gradation value is known in advance, by counting the number of pixels from the start of pattern data reading, the pixel data input to the error diffusion processing unit 628 is converted to each gradation value. It can be determined whether or not it is the first pixel of the patch data.
Based on the pattern data subjected to error diffusion processing in this way, after forming the reference pattern P1, the pattern data is read out at a predetermined interval and this time, the initial error is set to “b” and error diffusion processing is performed. Thus, the reference pattern P2 is formed. Thereafter, the reference pattern P3 is formed in the same manner with the initial error value set to “c”.

Then, the density of the patch indicating each gradation value of each of the reference patterns P1 to P3 is detected and sampled (step S2).
Specifically, for example, the number of clocks after the edge of the highest density portion side of each of the reference patterns P1 to P3 in FIG. 6 is detected by the pattern detection sensor 18 is counted.
The gradation value to be indicated by each patch of the reference pattern is stored in advance in a table (not shown) corresponding to the number of clocks. For example, the number of gradations of the input image data is 256, and 250, 240 , 230,..., 30, 20, 10, 0, and 10 when the density of the patch is detected, the same detection position as the relative low of the patch corresponding to the gradation value of the reference pattern is the number of clocks. N1 to N26 (N1 <N26) in advance, and by counting the number of clocks, the reference pattern P1 to P3 has a dot pattern of the dot pattern while being relatively the same detection area of patches of the same gradation value. Three density detection values in different cases can be sampled.

These sampling values are stored in a table (not shown) in the detection processing unit 630 corresponding to the corresponding gradation values. Here, the average value of each of the three detection values sampled corresponding to each gradation value. Is obtained (step S3).
For example, if the detection values d1 to d3 are obtained for the patches having the gradation value 60 of the reference patterns P1 to P3, the average value (d1 + d2 + d3) / 3 is obtained, and this value is calculated as the gradation value 60. Is a representative density value of the patch to be displayed, and is stored in a predetermined table in correspondence with the gradation value.

  Here, since the patches of the reference patterns P1 to P3 are formed by performing error diffusion processing using different initial error values for the same pattern data, the overall dot patterns are different. Accordingly, the dot patterns in the detection regions at relatively the same positions are relatively different for a plurality of patches showing one gradation value. By averaging these density values, at least the density of the toner patch is detected at one location. Compared to the conventional case, it is possible to obtain a toner patch density that is less susceptible to the uneven distribution of dot patterns.

Then, the γ correction data calculation unit 631 obtains γ correction data (γ curve) by the same method as described in the background art based on the representative density values of the patches corresponding to the respective gradation values (step) S4).
That is, the tone characteristic is first obtained from the tone value of each input image and the representative density value of the patch formed corresponding thereto, and then the γ correction data is set so that the tone characteristic is linearly corrected. The calculated lookup table is stored in the nonvolatile memory in the γ correction unit 627 in FIG. 3 and updated.

Such processing for obtaining γ correction data is also performed for other Y, M, and K, and each is stored in a non-volatile memory in the γ correction unit 627.
At the time of image formation, gradation correction is performed based on the corresponding γ correction data for each of C, Y, M, and K input image data, thereby forming an image with excellent gradation reproducibility. Is something that can be done.

  In this embodiment, for the first pixel of each patch data, the lowering error is changed to form a different dot pattern. However, as long as the dot pattern in the detection area of each patch is different. Therefore, the pixel for adjusting the lowering error is not necessarily limited to the first pixel of each patch pattern, and may be executed for a specific pixel that can affect the dot pattern of the detection region. Further, the present invention is not limited to only one pixel, and lowering errors of a plurality of pixels may be changed for one patch data within a range that does not greatly change the gradation value to be pseudo-displayed.

In addition, the initial error value is not changed for only one reference pattern (for example, P1), the conventional error diffusion process is performed for all the pixels as it is, and the initial error is changed only when another reference pattern is formed. Good.
A program for executing the above-described gradation correction method is stored in the ROM 66 (FIG. 2) in advance, and accurate gradation correction is performed by the CPU 61 reading and executing this program.

  Such programs include, for example, magnetic disks such as magnetic tape and flexible disks, optical recording media such as DVD, CD-ROM, CD-R, MO, and PD, Smart Media (registered trademark), COMPACT FLASH (registered trademark), and the like. It can be recorded on various computer-readable recording media such as a flash memory recording medium, and may be produced, transferred, etc. in the form of the recording medium, or wired including the Internet in the form of a program. In some cases, the data is transmitted and supplied via various wireless networks, broadcasting, telecommunication lines, satellite communications, and the like.

<Second Embodiment>
The second embodiment is different from the first embodiment only in the method of creating a reference pattern having a different dot pattern in the image signal processing unit 62. Therefore, the configuration unique to the present embodiment is described below. Only will be described.
That is, in the first embodiment, when the pattern pattern data of the reference pattern is subjected to error dissolution processing, the error initial value is changed to change the dot patterns of the reference patterns P1 to P3. In this embodiment, when error diffusion processing is performed on each patch data, the lowering error is not diffused to surrounding pixels for the first several pixels (that is, only lowering of the target pixel is performed, The feature is that the dot pattern is also changed by controlling the distribution of errors).

FIG. 9 is a diagram illustrating a configuration of a main part of the image signal processing unit 62 in this case. As shown in the figure, the present embodiment is different in that an error diffusion start position determining unit 634 is provided instead of the error initial value determining unit 633 in the image signal processing unit 62 of FIG.
The error diffusion start position determination unit 634 stores information on how many pixels in the patch data the error diffusion process starts, specifically the values of the pixel numbers M1, M2, and M3.

  For example, when the CPU 61 instructs to form the reference pattern P1, the error diffusion processing unit 628 reads the pattern data from the pattern storage unit 629 and executes error diffusion processing in units of patch data. The determination unit 634 counts the number of pixels input for each patch data, and the difference calculation unit 6283 is prohibited from calculating the difference up to the (M1-1) th pixel, and the error is “0”. Is output, and only the quantized data whose value is reduced by the lowering unit 6282 is output (hereinafter referred to as “simple lowering”). When the pixel becomes the M1th pixel, the above-described difference calculation is prohibited. Cancel and perform normal error diffusion processing. The number of pixels at the error diffusion start position is set as the reference pattern P2. In the error diffusion process of P3, the dot pattern of the toner patch showing the same gradation value can be changed for each of the reference patterns P1 to P3 by changing to “M2” and “M3”.

FIG. 10 is a diagram showing how toner patches having different dot patterns are formed in this way. For easy understanding, each reference pattern P1 to P3 has five gradation values as in FIG. It is assumed that it is composed of five patches indicating (tone values of 30, 80, 130, 180, 230).
As shown in the figure. When the pattern data is subjected to error diffusion processing, the error diffusion processing is performed for each of the reference patterns P1 to P3 from the first pixel of each patch data to “M1 (M1 = 0 in this example)”, “M2”, “M3” -th. For example, when the patches P11 to P31 indicating the gradation values 30 of the respective reference patterns P1 to P3 are enlarged by changing the number of pixels in the simple lowering portion, It can be seen that the dot pattern is different as shown in the square frame shown.

The image data is normally read from the pixel at the upper left corner of the image, but in this embodiment, a toner patch is formed from the high density side (from the lower side of FIG. 10) in the traveling direction of the intermediate transfer belt 15. Thus, the lower right corner of each patch is the first pixel of the patch data.
Accordingly, the dot pattern is different even in the detection area of each toner patch by the pattern detection sensor 18 (circle in each enlarged view indicates an example of the detection area), and relative in each toner patch showing the same gradation value. Even if the detection areas are exactly the same, different detection results can be obtained.

If these detected values are averaged to obtain the representative density of the toner patch indicating the gradation value, the dot pattern is unevenly distributed as compared with the case where only the density value of only one location of the patch is detected as in the prior art. A patch density that is not significantly affected can be obtained.
Here, in order to obtain a more appropriate representative density of toner patches, the dot patterns in the detection areas of a plurality of patches showing one gradation value are not biased to either “dense” or “sparse”. Thus, it may be desirable to adjust the values of the initial error values M1, M2, and M3.

<Third Embodiment>
In the third embodiment, only a dot pattern of a toner patch indicating each tone value is changed without changing a pseudo tone value by a different method.
That is, in the present embodiment, different noise data prepared in advance is overlapped with the same patch data, and then error diffusion processing is performed under the same conditions, and finally the toner patches having different dot patterns. Form.

FIG. 11 is a diagram illustrating a configuration of a main part of the image signal processing unit 62 in this case.
As shown in the figure, the present embodiment is different in that a noise storage unit 635 and a noise folding unit 636 are provided instead of the error initial value determination unit 633 of FIG.
The noise storage unit 635 stores different noise data in the internal nonvolatile memory, and the noise folding unit 636 reads out different noise data for each of the reference patterns P1 to P3 from the noise storage unit 636, and the pattern storage unit The data is superimposed on the pattern data stored in 629 and output to the error diffusion processing unit 638.

  Specifically, for example, assuming that the pattern storage unit 629 stores 8 × 8 pixel patch data 601 indicating a gradation value of 4 as shown in FIG. 12A (however, the actual size is larger). The noise storage unit 635 stores noise data (hereinafter referred to as “noise pattern”) 602 on a matrix of 8 × 8 pixels as shown in FIG. These data are folded for each pixel in the overlapping portion 636 to obtain the folded image data 603 shown in FIG.

  Here, the noise pattern 602 is configured such that the pixel values “0”, “1”, and “−1” are appropriately distributed, and the sum of the pixel values is “0”. In this way, the sum of the pixels of the noise pattern is set to “0” because the gradation value (pseudo gradation value) to be displayed as an entire 8 × 8 pixel block is changed by overlapping the noise pattern. This is to prevent you from doing it.

The dot pattern of the toner patch formed based on the image data obtained by error diffusion processing of the patch data on which noise is superimposed in this way is the case where the patch data on which noise is not superimposed is error diffusion processed. It goes without saying that they are different.
Further, it is possible to form a toner patch having a different dot pattern by changing the noise distribution state in the noise pattern and the noise level of each pixel.

FIG. 13 is a diagram illustrating a state where toner patches having different dot patterns are formed in the case of this example. Also here, for easy understanding, each of the reference patterns P1 to P3 has five gradation values (gradation values of 30, 80, 130, 180, and 230) as in the case of FIGS. It shall consist of patches.
When forming each patch in the reference patterns P1, P2, and P3, if different noise data N1, N2, and N3 are overlapped, for example, patches P11 to P31 indicating the gradation value 30 of each reference pattern P1 to P3. When each is enlarged, it becomes like the square frame shown on the left, and the dot pattern is different.

  Circles in each enlarged view show examples of detection areas by the pattern detection sensor 18. Even in the present embodiment, even if the relative detection areas are the same in each patch showing the same gradation value. Since each dot pattern is different, different detection results are obtained. Therefore, by creating the γ correction data using the average of these detection values as the representative density of the patch indicating the gradation value, it is possible to perform highly accurate gradation correction as in the first and second embodiments. Become.

  One of the reference patterns P1 to P3 may be one that does not fold noise data. Generally, when N reference patterns are created, at least (N-1) types of different noise patterns are necessary. Each noise pattern has a noise position (“1” or “−1” position in the 8 × 8 matrix in the example of FIG. 12B), the number of noises, and the magnitude of each noise as necessary. The size may be set as appropriate.

<Modification>
Although the present invention has been described based on various embodiments, the content of the present invention is not limited to the specific examples shown in the above embodiments. For example, the following modifications are possible. An example can be considered.
(1) In each of the above-described embodiments, patches indicating each gradation value are continuously formed as shown in FIG. 5 to form one reference pattern. However, this is not always necessary, and sub-scanning is not necessarily required. Each patch may be separated and formed at a predetermined interval in the direction. Further, the number of reference patterns is not limited to three and may be more than that. For patches that show the same gradation value by increasing the number of reference patterns, if the density of as many different dot patterns as possible is detected, the representative density will more appropriately indicate the density of the toner patch. Therefore, more accurate γ correction data can be obtained. In this case, the representative density may be obtained by simply averaging the sampling values of a plurality of patches showing the same gradation as in the above embodiment, or the detection values having the minimum and maximum values are excluded, Only the remaining detection values may be averaged and used as the representative density value.

Further, if at least two reference patterns are formed and γ correction data is obtained from the detected values, gradation correction can be performed more accurately than at least the conventional case.
(2) In each of the above embodiments, the number of gradations of the input image is 256, and the gradation value changes by 10 as the reference pattern “0, 10, 20, 30,..., 230, 240, In this example, 26 types of toner patches of “250” are continuously arranged. However, if a toner patch whose gradation value changes with a finer width is formed, the accuracy of the γ correction data is increased. Further, in order to perform tone correction, a toner patch showing at least two tone values is required.

(3) In each of the above embodiments, the reference pattern formed on the intermediate transfer belt is detected to obtain the γ correction data. However, the toner image of the reference pattern formed on each photosensitive drum is transferred to each photosensitive drum. Detection may be performed by a pattern detection sensor provided in the apparatus.
In addition, it is possible to secondarily transfer the reference pattern onto the recording sheet and detect it with a toner detection sensor installed in the paper discharge path.

In the description of each of the above embodiments, “detecting the density” of the reference pattern or toner patch means not only the case where the density value is directly acquired by the detection signal of the pattern detection sensor, but also the detection signal of the sensor. Alternatively, the toner adhesion amount may be once obtained from the toner and converted into the density with reference to a table or the like showing the correlation between the toner adhesion amount and the density.
(4) In each of the above-described embodiments, the example in which the gradation value of the input image data is binarized and the error diffusion process is performed has been described. In general, when the input image data is expressed by the number of gradations L. In addition, the error diffusion process can be applied as long as the number of gradations K (= 2 ^k (k = 1, 2, 3,...)) Smaller than L is reduced. Usually, K is about 4 to 16, and this makes it possible to express gradation changes more smoothly than in the case of binarization.

(5) As a method of creating toner patches having different dot patterns based on patch data indicating the same gradation value, any two or all of the methods described in the first to third embodiments are used. It can also be used in combination.
Further, in each of the above-described embodiments, the toner patches showing the same gradation value by the pattern detection sensor 18 are relatively located at the same position (indicated by the circles in FIGS. 7, 10, and 13). (See the example of the sensor detection area). However, this is not necessarily the case, and the density of some areas at relatively different positions of each toner patch may be detected. I do not care. This is because, if the dot pattern of the entire toner patch is different, it is considered that the uneven distribution of dots is different in any part of the region.

  (6) In the above-described embodiment, an example in which the image forming apparatus of the present invention is applied to a color copying machine has been described. The present invention can be applied to a facsimile machine, an MFP (Multiple Function Peripheral), and other various image forming apparatuses. In addition to a color image forming apparatus, an image forming apparatus dedicated to monochrome may be used.

  The present invention can be applied to general image forming apparatuses capable of gradation expression by the error diffusion method.

1 is a diagram showing a schematic configuration of a tandem type color copying machine 1 according to a first embodiment of the present invention. 2 is a block diagram showing a configuration of a control unit 60 of the color copying machine 1. FIG. It is a block diagram which shows the structure of the image signal process part 62 of the said control part 60. FIG. It is a block diagram which shows the structure of the error diffusion process part 628 of the said image signal process part 62. FIG. It is a figure which shows an example of the reference pattern used at the time of gradation correction data creation processing. FIG. 4 is a diagram illustrating an example in which three reference patterns are formed on an intermediate transfer belt. FIG. 4 is a diagram illustrating a state in which dot patterns are different for each toner patch in the first embodiment. It is a flowchart which shows the procedure of the (gamma) correction data creation process of this invention. It is a block diagram which shows the structure of the principal part of the image signal process part 62 which concerns on the 2nd Embodiment of this invention. FIG. 10 is a diagram illustrating a state in which dot patterns are different for each toner patch in the second embodiment. It is a block diagram which shows the structure of the principal part of the image signal process part 62 which concerns on the 3rd Embodiment of this invention. It is a figure which shows the example of the patch data after convolution of the original patch data, noise data, and patch data and noise data. FIG. 10 is a diagram illustrating a state in which dot patterns are different for each toner patch in the third embodiment. FIG. 6 is a diagram illustrating that when a density of a low density toner patch expressed by an error diffusion method is detected, a detection result varies depending on the detection position. It is a figure which shows the gradation characteristic before correction | amendment obtained by the detection result of FIG. It is a figure which shows the (gamma) curve produced based on the gradation characteristic of FIG. It is a figure which shows the gradation characteristic by which gradation correction | amendment was carried out by (gamma) curve of FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Copy machine 10 Intermediate transfer part 15 Intermediate transfer belt 18 Pattern detection sensor 20 Image forming part
30 Paper Feed Unit 40 Secondary Transfer Unit 50 Fixing Unit 60 Control Unit 61 CPU
62 Image signal processing unit 63 Image memory 64 LED array driving unit 65 RAM
66 ROM
67 EEPROM
70 Operation Panel 100 Image Reading Unit 200 Image Forming Unit 628 Error Diffusion Processing Unit
633 Error initial value determination unit 634 Error diffusion processing start position determination unit 635 Noise storage unit 636 Noise folding unit

Claims (7)

Error diffusion processing is performed under different conditions on each of the patch forming image data for forming a plurality of toner patches having different gradation values, thereby forming a plurality of toner patches having different dot patterns for each gradation value. Toner patch forming means;
Representative density acquisition means for detecting the density of a part of each of a plurality of toner patches formed for each gradation value and acquiring the representative density of the toner patch indicating each gradation value from the detected density; ,
Gradation correction data for generating gradation correction data based on the representative density value and correcting the gradation of the input image data;
An image forming apparatus comprising: an image forming unit that forms an image by processing the gradation-corrected input image data by an error diffusion method.   The representative density acquisition unit acquires an average value of densities detected for a plurality of toner patches formed for each gradation value as a representative density of a toner patch indicating the gradation value. Item 2. The image forming apparatus according to Item 1.   When the toner patch forming unit performs error diffusion processing on the data of a specific pixel in the image data for patch formation, the toner patch forming unit changes an error between the gradation value of the pixel and a value obtained by lowering the value. The image forming apparatus according to claim 1, wherein toner patches having different dot patterns are formed for the same gradation value.   The image forming apparatus according to claim 3, wherein the specific pixel is a first pixel read from the patch image data.   When the toner patch forming unit performs error diffusion processing on the image data for patch formation, the toner patches having different dot patterns with respect to the same gradation value are obtained by changing the position of the pixel from which the error diffusion processing is started. The image forming apparatus according to claim 1, wherein the image forming apparatus is formed.   When the toner patch forming unit performs error diffusion processing on the patch image data, the toner patch forming unit generates noise that does not affect the sum of gradation values of each pixel in the predetermined pixel block unit. The image forming apparatus according to claim 1, wherein toner patches having different dot patterns are formed with respect to the same gradation value by performing folding. Each of the patch formation image data for forming a plurality of toner patches having different gradation values is subjected to error diffusion processing under different conditions to form a plurality of toner patches having different dot patterns for each gradation value. A toner patch forming step,
A representative density acquisition step of detecting the density of a part of each of the plurality of toner patches formed for each tone value and acquiring the representative density of the toner patch indicating each tone value from the detected density; ,
A gradation correction step for creating gradation correction data based on the representative density value and correcting the gradation of the input image data;
An image forming step of processing the tone-corrected input image data by an error diffusion method to form an image.

Images