JPH07273991A - Image forming device - Google Patents

Abstract

PURPOSE:To hold gradation characteristics in the best state for a long period by feeding variation in error-dispersed parameter back corresponding to variation in tonality and making automatic adjustments as image formation conditions change. CONSTITUTION:The device which makes image data from a scanner into a multi-valued signal by the multi-value-coding means 11 of an error diffusing process part 6 on the basis of a threshold value set in a control means 1 and outputs it to a printer performs an image formation optimizing process again when a specific time or a specific number of copies is reached after the last image formation optimizing process. At this time, a calculating means 13 calculates the comparison error between image data of a test pattern from an input device and detection data on a toner sticking amount or reflection factor from a detection part, i.e., the error between the corrected image signal from a correcting means 10 and the converted multi-valued signal from a multi-value-coded signal converting means 12. When the error is outside a permissible range, a correction quantity is calculated and the multi-value-coded data in a memory 2 are varied to perform the optimizing process for image formation conditions. Consequently, gradation variation due to a secular change and an environmental change is compensated.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image forming apparatus for subjecting a document image including a photographic image to multi-valued processing satisfying faithful gradation reproducibility by using an error diffusion method, and
The present invention relates to maintaining the image quality by minimizing the variation in gradation characteristics due to the influence of environmental changes and aging.

[0002]

2. Description of the Related Art An image forming apparatus for printing, copying, or displaying image data of a document original including a photographic image or a computer, image data edited, image display characteristics adjusted in an initial stage, particularly gradation reproduction or The characteristics of color reproduction vary depending on the environment (temperature, humidity, atmospheric pressure, atmosphere) inside and outside the image forming apparatus, the above environment left unattended, the history of frequency of use, and deterioration with time of image forming means.

Therefore, the maintenance of the image quality which has fluctuated by performing the replacement of parts and materials, the readjustment, etc. has been performed as the maintenance.

In recent years, attempts have been made to divide a deteriorated portion and a relatively hard-to-deteriorate portion into a unit so that the deteriorated portion can be easily replaced.

Further, a physical quantity relating to a characteristic change of image formation such as a change of environment is detected, and image forming conditions (for example, in the case of electrophotography, charge amount, developing bias, exposure amount, pulse width, etc.) are detected.
In the case of thermal transfer, it has been proposed to change the head temperature, etc.) so as to reduce image quality deterioration. It has been proposed to detect the final fluctuation of the image density and change the image forming conditions and the gradation compensation means (for example, a gamma correction table, a color correction table, etc.) for image processing. Further, in the electrophotographic apparatus, it has been proposed to detect the amount of toner adhering to the photoconductor by a reflection type detection means and change the image forming condition or the gradation compensation means of the image processing based on the detection result.

On the other hand, generally, in a document image processing apparatus capable of handling not only code information but also image information, image information having a strong contrast such as characters and diagrams on a document read by a reading means such as a scanner has a fixed threshold value. The simple binarization is carried out by the method, and the image information having gradation such as a photograph is binarized by a pseudo gradation method such as a dither method. This is because when the read image information is subjected to a simple binarization process with a fixed threshold, the resolution of the character and line image areas is preserved and the image quality does not deteriorate, but in the area of the photographic image, The image quality is degraded because the tonality is not preserved. On the other hand, if the read image information is subjected to gradation processing by the tissue dither method, the gradation of the photographic image is preserved and the image quality is not deteriorated, but the resolution of the character and line image areas is improved. Is deteriorated, the image quality is deteriorated. That is, with respect to the read image information, it is impossible to simultaneously satisfy the image quality of the areas of each image type having different characteristics with a single binarization process.

However, the gradation of the area of the photographic image is satisfied, the gradation of the area of the character / line image is satisfied, and
The error diffusion method has been proposed as a binarization method having a better resolution than the tissue dither method in the area of the line image. The error diffusion method adds a value obtained by multiplying a target pixel with a multivalued (binarized) error of a peripheral pixel that has already been multivalued (binarized) by a weighting coefficient, and then multivalued with a fixed threshold value. This is a method of binarizing (binarizing) and distributing the multi-valued (binarized) error generated at this time to peripheral pixels. That is, although the output level corresponding to each pixel is a predetermined multi-value (binary), the error related to the local area of the output image corresponding to the local area of the original image (multi-valued image data) having continuous tone is This is a method of compensating for gradation by eliminating or minimizing the total sum. Therefore, the number of input gradation steps can be represented with a smaller number of output levels than the number of gradation steps of the input pixel signal.

As described above, conventionally, when an attempt is made to maintain the gradation characteristic by changing the image forming condition, the gradation characteristic variation and the correction effect generally have a non-linear relationship, and a simple single image forming condition variable is changed. However, it is difficult to realize accurate correction for all areas from high density to low density. Also, in order to absorb this non-linear correlation, in order to change the pulse width modulation characteristics, gamma correction, or color correction table that can be corrected for each pixel forming an image, a high-speed memory is finally expressed. The number of gradation steps required is equal to or more, and the pulse width modulation requires a high number of highly accurate operation time resolution and the number of steps of the writing means that finally emits light or generates heat. Further, in the gamma correction table or the like, it is necessary to perform the calculation such that the originally nonlinear input / output characteristic becomes an inverse characteristic (inverse function) with respect to the further nonlinear variation for the number of gradation steps to be finally expressed or more.

In the error diffusion processing means as described above, the multi-valued signal means a predetermined number of levels such as the number of writing stages of the image forming system. On the other hand, the input signal may be a read signal from a scanner or the like, a created image from a computer or the like, and may be used as a reflectance or density information. Will be different. Therefore, in the error calculation, a contradiction occurs in which physically different dimensions are added and subtracted.

Further, when the gradation characteristic of the output image of the image forming apparatus which forms an image based on the multi-valued signal of the error diffusion processing changes due to environmental changes, aging, etc., the gamma correction means or color correction is provided upstream. Means, or a multi-valued signal converting means or the like is provided outside the error diffusion processing section, and a correction characteristic is calculated (for example, table contents) corresponding to a detection result of gradation variation. However, even if there is no contradiction in the calculation of the different dimensions described above when there is no change in the gradation characteristics, even if the gradation characteristics of the output image change, even if a gamma table is corrected,
The relationship between the corrected input pixel signal and the multi-valued level changes, and the accuracy of error calculation deteriorates.

Further, in the error diffusion processing for performing multi-valued processing with a fixed threshold value, substantial gradation characteristics corresponding to each level of the multi-valued signal and macro gradation with respect to variations in the number of gradations are provided. Even if the characteristics change little, it is not possible to suppress the deterioration of the image quality such as the deterioration of the resolution and the increase of the noticeability of the texture.

[0012]

SUMMARY OF THE INVENTION The present invention eliminates the above-mentioned drawbacks, and provides a method in which substantial gradation expression by an image forming system has characteristics of area gradation, density gradation, or both. Even if there is a change in gradation characteristics in the image forming system due to environmental changes or aging, and even if the number of gradation steps per pixel is substantially reduced due to the change, gradation characteristics can be accurately measured in a short time. Another object of the present invention is to provide an inexpensive image forming apparatus capable of automatically compensating for fluctuations in resolution to a minimum, labor saving or reducing maintenance, and reducing total running cost.

[0013]

According to the present invention, even if the gradation characteristic per pixel of the image forming system changes or the gradation step number decreases due to the change due to the influence of environment and aging, the gradation characteristic of the local area is reduced. Is maintained so that the initial image quality is maintained for a long period of time.

The basic concept of the present invention is that, in an image forming apparatus that performs gradation processing by error diffusion processing, an input output signal is output based on the output of a means for detecting a change in image forming condition that appears due to a change in image density. When changing the parameter for error diffusion in the unified dimension, the closed-loop control for maintaining the gradation characteristic is performed by feeding back the changed value according to the variation of the gradation characteristic with respect to the target.

According to one aspect of the present invention, an image forming apparatus which multi-values a plurality of pixel signals forming an image signal into a predetermined number of levels and compensates a multi-valued error corresponding to the pixel signals in a local region. In, a multi-valued means for inputting a corrected pixel signal obtained by correcting the input pixel signal, multi-valued the corrected pixel signal into a signal having a predetermined level, and providing a multi-valued signal; Multi-valued signal conversion means for converting the multi-valued signal by generating a predetermined signal corresponding to each level of the signal, and providing the converted multi-valued signal, the correction pixel signal and the converted multi-valued signal An error calculation means for calculating the error between the error calculation means and the multi-valued error signal, and the error quantity calculated by the error calculation means with the previously stored error quantity for each relative position with respect to the current pixel of interest. Error description stored as a total Means, a correction means for outputting the corrected pixel signal to the multivalued means by the input pixel signal and the correction amount, with the total error amount of the positions corresponding to the current pixel of interest as the correction amount, Detecting means for detecting an amount relating to the image density formed based on the digitized signal, pattern forming means for forming a pattern for density variation detection by the detecting means, and the multi-valued signal in the multi-valued signal converting means. Changing means for changing the conversion value corresponding to each level of the converted signal, and conversion corresponding to each level by the changing means based on the detection result of the detecting means corresponding to a plurality of predetermined patterns formed by the pattern forming means Control means for changing the value, forming a pattern again by the pattern forming means, and repeatedly executing the detecting operation by the detecting means. The image forming apparatus is obtained which.

According to another aspect of the present invention, the control means is provided with a determination means for determining whether or not the change processing by the change means should be executed based on the detection result of the detection means. The image forming apparatus described in the preceding paragraph is obtained.

[0017]

According to the present invention, in the above structure, the image data from the input device is subjected to error diffusion to be converted into multi-valued data, and the multi-valued data is output to the output device having a predetermined number of levels. In the image forming apparatus, the image forming condition optimizing process is started when it is detected that a predetermined time has elapsed from the last image forming condition optimizing process or the number of formed images reaches a predetermined number, and the input device is operated. For example, with respect to the target pixel and the peripheral pixels of the test pattern, the error in the pixel unit is calculated from the converted image data and the image data from the input device, and when the error is out of the predetermined range, A correction amount is calculated based on the calculated pixel-by-pixel error with respect to the target pixel and peripheral pixels, and the multi-valued data is changed based on the calculated correction amount. There are those who to again perform the image forming condition optimization process.

[0018]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A first embodiment of the present invention will be described below with reference to the drawings.

In the first embodiment, for example, a CCD scanner carries out A / D conversion of a linear analog signal having a reflectance of a pixel unit according to a reading resolution of an original image, and a standardized reflection corresponding to each pixel. An input signal from a reading device that outputs a rate signal is used as an input pixel signal, and a printing area per pixel is changed based on a multi-valued image signal as an output pixel signal. The present invention relates to an image processing device of a system which is an output device such as a photographic printer, a sublimation type thermal transfer printer or a halftone dot printing machine.

FIG. 1 is a schematic diagram of an image processing apparatus of the present invention. This image processing device is an input image signal (reflectance signal) input after being read by a reading device such as a CCD scanner.
Is input as, for example, 8-bit digital data per pixel, and this device is used to correct the image data of the pixel of interest by the error diffusion method.

That is, as shown in FIG. 1, the image processing apparatus comprises a control means 1, a non-volatile memory 2, a pattern generator 3, a selector 4, a buffer 5, and an error diffusion processing section 6.

The control means 1 is composed of, for example, a CPU and controls the entire image processing apparatus, so that an initial adjustment process and an optimization process of threshold values and reflectance values as error diffusion parameters, which will be described later, are performed. It has become. An operation panel (not shown) for instructing various operations is connected to the control device 1. The control means 1 is supplied with a detection signal from a detection means for detecting the toner adhesion amount of an area gradation type electrophotographic printer as an output device.

The non-volatile memory 2 is stored in a test pattern data memory 2a in which reflectance data of 2 or 3 test patterns is stored, a threshold value register 21a in a multilevel halftoning means 11, which will be described later. The initial value of the threshold value and the reflectance memory 42a of the multilevel signal conversion means 12 ...
An initial value memory 2b in which initial values of reflectance corresponding to the respective levels stored in are stored, and a correction amount (Δg ′) of the multilevel signal conversion value corresponding to the output level (multilevel image signal) Is stored in the multi-valued signal conversion table 2c, the threshold value set in the threshold value register 21a in the multi-valued conversion means 11 at the end, and the reflectance memory 42a of the multi-valued signal conversion means 12. , A current value memory 2d in which the reflectance (conversion multi-value signal value) corresponding to each level set in, ... Is stored, and a control standard value for pass / fail judgment (whether the change amount is within an allowable range).
2e in which is stored.

The pattern generator 3 is for generating a test pattern in response to an instruction from the control means 1.

The selector 4 outputs the test pattern from the pattern generator 3 to the error diffusion processing unit 6 at the time of printing the test pattern on the paper or at the time of drawing on the photoconductor, and at the time of normal printing, the image data from the scanner. Is output to the error diffusion processing unit 6, and the density data from the scanner in the adjustment mode is output to the buffer 5.

The buffer 5 temporarily stores the density data from the selector 4 and outputs it to the control means 1.

The error diffusion processing section 6 is, as shown in FIG.
Correction means 10, multi-value quantization means 11, multi-value quantization signal conversion means 1
2, error calculation means 13, and error correction amount calculation means 17
It is composed by.

The correction means 10 performs correction by adding the input image data of the target pixel as the reflectance data of a plurality of bits from the selector 4 with the correction amount signal supplied from the error storage means 16. The corrected image signal thus corrected is output to the multi-value quantization unit 11 and the error calculation unit 13. The correction means 10 is composed of, for example, an adder.

The multi-value quantization means 11 compares the corrected image signal of the corrected pixel of interest with the threshold for multi-value conversion to correspond to the number of output stages (the number of levels: 4) on the printer side as an output device, for example. And outputs a multi-valued image signal as multi-valued data. The multi-value quantization means 11 performs multi-value quantization based on the multi-value quantization threshold set by the control means 1. The multi-valued image signal from the multi-valued means 11 is output to the printer and is also output to the multi-valued signal conversion means 12.

The multi-value signal converting means 12 is the multi-value converting means 1.
The multi-valued image signal from 1 is converted into a converted multi-valued signal as a reflectance signal having the same dimension as the input image data, and the converted multi-valued signal is output to the error calculation means 13. .

The error calculating means 13 calculates the multi-valued error signal of the target pixel from the corrected image signal from the correcting means 10 and the multi-valued signal from the multi-valued signal converting means 12, and this calculation is performed. The multi-valued error signal thus generated is output to the error correction amount calculation means 17.

The error correction amount calculation means 17 is the error calculation means 1
The correction amount signal is calculated from the multi-valued error signal of the pixel of interest from 3 and the pre-stored surroundings of the multi-valued error, and the calculated correction amount signal is output to the correction means 10. It is supposed to be done.

The error correction amount calculation means 17 is composed of the error filter means 14 and the weighting coefficient storage means 1 shown in FIGS.
5 and error storage means 16.

The weighting coefficient storage means 15 stores four peripheral pixels for the target pixel (the target pixel * as shown in FIG. 8A).
4 peripheral pixels including a pixel on the same line and a pixel on the previous line) are stored, and the weighting factors A and 4 are stored in four registers 15a ,.
B, C, D (A = 7/16, B = 1/16, C = 5/1
6, D = 3/16) are stored.

The error filter means 14 is the error calculating means 1
The multi-valued error signal from 3 is multiplied by each of the weighting factors A, B, C, and D of the weighting count storage means 15 to calculate the weighting error of the peripheral 4 pixels with respect to the pixel of interest. The error is output to the error storage means 16. For example, it is composed of four multipliers 14a, ..., Four adders 14b ,.

The error storage means 16 is the error filter means 1
As shown in (b) of FIG. 8, the weighting errors of the four pixels calculated in step 4 are respectively e _A , e _B , and e for the target pixel *.
_C, by storing and adding the areas of e _D, and calculates a correction amount signal for the target pixel, the calculated correction amount signal are outputted to the correction means 10. For example, it is composed of a line buffer.

As shown in FIG. 2, the multi-value quantization means 11
.., 21o, fifteen comparators 22a, ..., 22o, and an encoder 23. 21o are set in the threshold registers 21a, ..., 21o according to the change target address, the read / write signal as the change information from the control means 1, and the change content information, respectively. Has become. The thresholds Th1 to Th15 of the respective threshold value registers 21a, ...
22o, the corrected image signal from the correction means 10 is supplied to each of the comparators 22a, ... The comparison result of each comparator 22a, ..., 22o is encoded by the encoder 23, converted into a multi-valued image signal, and output.

For example, when the number of gradations is 2, only the threshold Th1 of the threshold register 21a is set and only the comparator 22a is enabled. Accordingly, when the corrected image signal is larger than the threshold Th1, the comparator 2
If the corrected image signal is smaller than the threshold value Th1 when “1” is output from 2a and the encoder 23, the comparator 22
“0” is output from a and the encoder 23.

When the number of gradations is 4, the threshold value register 21
The thresholds Th1, Th2 and Th3 of a to 21c are set, and the comparators 22a to 22c are enabled. Accordingly, when the corrected image signal is smaller than the threshold Th1, the comparators 22a to 22c output "0" and the encoder 23 outputs "00". If the corrected image signal is between the threshold values Th1 and Th2, the comparator 22a
Output "1", the comparators 22b and 22c output "0", and the encoder 23 outputs "01". When the corrected image signal is between the thresholds Th2 and Th3, the comparators 22a and 22b output "1", the comparator 22c outputs "0", and the encoder 23 outputs "10". When the corrected image signal is larger than the threshold Th3, the comparators 22a to 22c output "1" and the encoder 23 outputs "11".
Also, in the case where the number of gradations is "8" and "16", the multi-valued image signal is similarly output.

One example for implementing the comparator 22a (22b, ...) Is shown in FIGS. In the figure, A0-A
Reference numeral 9 represents each bit of the corrected image signal, and B0 to B9 represent each bit of the threshold values Th1 and Th15. As shown in FIG. 3, the 10-bit corrected image signal A includes comparators 22a ,.
Is compared with each 8-bit threshold value B, and a 1-bit comparison result is output from each comparator 22a. Each bit on the input side of the comparator 22a is composed of an AND circuit 31, inverter circuits 32 and 33, and an exclusive OR circuit 34, as shown in FIG. As shown in FIG. 5, the output side of the comparator 22a is composed of an AND circuit 35, ..., An OR circuit 36, an AND circuit 37, and an inverter circuit 38.

As shown in FIG. 6, the multilevel signal conversion means 12 includes a decoder 41 and 16 reflectance memories 42a ,.
It is composed of 42p. The reflectance memory 42a, ...
The reflectance signal corresponding to each level of the multi-valued image signal is set in 42p. .. 42p are stored in the reflectance memories 42a, ... 42p based on the change target level (supplied via the decoder 41) as the change information from the control means 1, the read / write signal, and the change content information. , ~ Th15 are set.

Here, the absolute reflectance Rabs with respect to the original document;
Real number from 0 to 1, absolute reflectance of scanner white reference plate RabsW;
Example 0.91, absolute reflectance RabsB of black reference plate; Example 0.002, A /
D-converted reflectance data Rad; integer from 0 to 255, A / D converted value RadW for white reference plate; Example 240, A / D converted value RadB for black reference plate; Example 10, reflectance data after shading correction R: real number from 0 to 255, reflectance data after shading correction for the white reference plate RW; 0, reflectance data after shading correction for the black reference plate RB
255, converted absolute reflectance R '; real number from 0 to 1, absolute reflectance RW' converted to white reference plate; example 0.91 converted absolute reflectance RB 'to black reference plate; example 0.002, converted density D ′; Real number, example 0.04 to 2.7, converted density DW ′ for the white reference plate; example 0.04, converted density DB ′ for the black reference plate ′; example 2.7, Rabs by the A / D conversion is the interval [RabsB, RabsW] is converted as follows.

Rad = Aad * Rabs + Bad ... (1) However, Aad = (RadW-RadB) / (RabsW-RabsB) ... ( 2) Bad = RadW-Aad * RabsW ... (3) If shading correction is performed by the following procedure, X1 = Rad-RadB ...・ ・ ・ ・ (4) X2 = RadW-RadB ・ ・ ・ ・ (5) X3 = X1 / X2 ... (6) R = 255 * (1-X3) ... (7) The reflectance signal R output from the scanner according to the equations (4) to (7) is R = 255 * {1- (Rad-RadB) / (RadW-RadB)}. .. (8) White reference plate and black reference from equation (8) The shading-corrected data for Rad = RadW → RW = 0 ... (9) Rad = RadB → RB = 255 (10) In this embodiment, a system in which the reflectance signal R linearly standardized with respect to the reflectance of the original is input as an input pixel signal f described later is taken as an example. Hereinafter, the same dimension as the reflectance signal R expressed by the equation (8) will be simply referred to as reflectance.

That is, the multi-valued image signal from the multi-valued image forming means 11 is a signal representing an output level L representing the number of gradation steps of the output device, and the output device prints per pixel based on this output level. Area gradation is expressed by changing the area.

Here, due to changes in environment (temperature, humidity, atmospheric pressure, etc.) and aging (degradation of image forming material, writing system, etc.),
When the printing area per pixel corresponding to each output level varies, a deviation occurs between each reflectance corresponding to each output level in the above-mentioned multilevel signal conversion means 12 and the reflectance for the pixel actually printed, and this The multi-valued error signal calculated by comparing the converted multi-valued signal and the corrected pixel signal is different from the actual one, and the desired gradation of the output image cannot be reproduced with respect to the finally input document image or image signal.

Therefore, in this embodiment, the change of the image forming condition or the change of the density (reflectance) of the formed image is detected by the detecting means for detecting the toner adhesion amount of the printer, and in accordance with this change, the above-mentioned change is made. In order to reduce the gradation variation, the error calculation can be performed as accurately as possible by newly setting the reflectance close to the reflectance per pixel corresponding to each changed output level in the multilevel signal conversion means. .

Further, in this embodiment, in addition to the above correspondence, the step of detecting the image output for the predetermined pattern and the step of determining the deviation from the target are executed again by using the changed value of the converted multilevel signal. It is configured to perform the correction in a closed loop.

In FIG. 1, the control means 1 calculates the converted multi-valued signal content of the multi-valued signal conversion means 12 and the threshold value of the multi-valued means 11 to be changed based on the detection signal from the detection means. Then, the contents of the multi-value conversion unit 12 and the multi-value conversion unit 11 are changed according to the respective change information.

The decoder 41 shown in FIG. 6 receives the multi-valued image signal from the multi-valued image forming means 1 in the above-described normal processing, and when changing the contents, one element of the change information 8 from the control means 1 is used. A certain change target level 8a is input. In normal processing, the reflectance of the address corresponding to the level is read from the reflectance register 42a, which stores the reflectance corresponding to each level of the multilevel image signal, and is output as a converted multilevel signal. To do. At this time, as the read / write signal 8b, only the read signal is input. On the other hand, when the contents are changed, the memory address is specified at the change target level 8a and the contents of the change contents information 8c are read / write signal 8
It is rewritten by using b as a write signal.

In the multilevel halftoning means 11 of FIG. 2, the threshold value register 21 in which the threshold value to be compared with the correction pixel signal from the correction means 10 is stored in the above-mentioned normal processing.
With respect to a, ..., Contents of the threshold value registers 21a ,. At this time, as the read / write signal 7b, only the read signal is input. On the other hand, when the contents are changed, the target threshold value is rewritten by changing the contents of the corresponding change contents information 7c into the register designated by the change target address 7a by using the read / write signal 7b as a write signal.

FIG. 9 shows the relationship among the document density Di, the reflectance R, the output level L, and the output density Do in order to explain the change. The number of output levels of the output device is four values (L0, L1, L2, L3
) And the density Do corresponding to each level by the output device
(L). It is assumed that the converted multi-valued signal in the multi-valued signal conversion means 12 corresponding to each level is set to R (L) for the purpose of reproducing the gradation characteristic of the original density and the output image density, so-called gamma characteristic. . At this time, Do (L), Di (L) and R (L) corresponding to each level L are stored through a straight line gamma (γ) (short broken line).

Now, it is assumed that the density per pixel changes due to environmental changes, aging, and the like. In the example of FIG. 9, the fluctuation of the image density is Do (L1), Do (L2) is Do ′ (L1),
Suppose it changed like Do '(L2). Then L2, L
Since the density corresponding to 3 has changed, the gamma characteristic is γ1,
It changes to a non-linear characteristic that passes through the point of γ2.

Similar document density Di, reflectance R, and
The relationship between the output level L and the output density Do is shown. here,
In contrast to the above-mentioned concentration fluctuation, R '(L1), R' (L
By changing the converted multi-valued signal in the multi-valued signal conversion means as in 2), it becomes possible to maintain and reproduce the gradation characteristics without causing a change in the gamma characteristics (long broken line).

The relation between the absolute reflectance Rabs and the absolute reflection density D is expressed by the following well-known equation.

D = -log (Rabs) (19) Therefore, the reflectance R with respect to the density can be obtained by the equation (19) and the above equation (8).

R '(L1) and R' (L2) are output densities Do '(L1) and Do' (L2) per pixel, and input densities Di '(L1) and D'
When i '(L2), it is ideal that the value is converted by the time-variant Di · R characteristic. Therefore, the value closest to the above value with the allowable resolution is set as the corresponding R value.

FIGS. 11A to 11C show examples of the converted multilevel signal value per pixel and the reflectance of each pixel actually printed. In this example, the set conversion multilevel signal value and the conversion value of the density actually printed are as shown in FIG.
Then, when they match according to the initial setting, (b) of FIG. 11 shows the case where the printed density fluctuates and does not match the converted multi-valued signal value, and (c) of FIG. 11 shows the multi-valued signal. In this case, the converted multi-valued signal corresponding to each level is changed to the corresponding converted density conversion value. To simplify the explanation, the converted multilevel signal is all over the wide area.
Assuming that the input is a value of 100 ', the initial converted multilevel signal value in FIG.
Suppose it was 0 '.

In FIG. 11A, the converted value of the density printed at the selected level of the multi-valued signal with respect to the input pixel signal becomes the same "100" and the difference between the input pixel signal and the converted multi-valued signal. That is, the multi-value quantization error is “0”. Therefore, the error is not diffused and the density conversion value of "100" is printed uniformly, and the average value per pixel becomes "100" which is equal to the input pixel signal.

Here, the conversion value of the density actually printed is'
In the example of (b) of FIG. 11 which changes from 100 ′ to “80”, the converted multilevel signal value is the same as (a) of FIG. Top is error '
It becomes 0 '. Therefore, the error is not diffused and the same multilevel signal is output uniformly. However, since the actual density per pixel varies, the converted value of the density of each pixel and its average value become '80'. Therefore, in this embodiment, the converted multilevel signal value is changed based on the fluctuation.

In FIG. 11 (c), the value of "80", which is equal to the converted value of the density actually printed, which is changed to "80", is changed with respect to the level of "100". Similarly, the adjacent level (in this case, the level one level higher) is also assumed to be the converted density conversion value “120”, and the setting is changed. As a result, if the input pixel signal '100' is multi-valued and the converted multi-valued signal corresponding to the multi-valued signal is '120',
The multi-value quantization error becomes "-20", which is diffused to the peripheral pixels. Now, for the sake of simplification, if it is assumed that the pixels on the right side and the pixels on the lower side are divided by half, the previous error component is “−10” for the next input pixel signal, and the correction error signal is “90”.
Becomes If the threshold value is' 100 = (120 + 80) / 2 ', the level of' 80 'is selected, the corresponding converted multilevel signal is'80', and the multilevel error is' It becomes -10 'and is propagated. By repeating this, the probability of hitting with '120' and '80' in this example is halved, and the average value per pixel approaches the input pixel signal '100'.

That is, even if the density variation per pixel occurs, the average density (reflectance) in the local region is compensated by changing the converted multilevel signal value according to the state.

On the other hand, regarding the threshold value, in this embodiment, the average value of the converted multilevel signals corresponding to the adjacent levels is used for the comparison of the multilevel signals corresponding to the respective levels.

Th (l) = (Ro (l) + Ro (l + 1)) / 2 (20) However, l = 0, 1, 2, ..., Lmax-1 Lmax; Number of signal levels This is an addition calculation in a unified dimension in which the input pixel signal is the dimension of the reflectance and is converted by the multi-valued signal conversion means 12 into the reflectance corresponding to the output density per pixel of the output device. This is because, if the error is calculated by, the error between the input reflectance and the output reflectance corresponding to the multivalued level is minimized.

FIGS. 12A and 12B show the difference in reflectance corresponding to the output pixel according to the threshold value. FIG. 12A shows an example in which the threshold value is the average value of the above-mentioned adjacent levels, and FIG. 12B is an appropriate selection. The input pixel signals are all set to 200 as values that can be output per pixel.

In FIG. 12A, the threshold value is 200
The threshold values 160 and 220 averaged between the output values of 240 and 120 adjacent thereto are set. Therefore, if the input is 160 or more and 220 or less, it is multi-valued as the output level 2 and the area printing corresponding to the output value 200 is performed. Therefore, the input of 200 results in the output of 200, and becomes 200 for all the pixels with zero error. On the other hand, in FIG. 12B, the output value is 200 as in FIG.
, But the threshold value is selected appropriately, the error pixel feedback causes the correction pixel signal to be 200
Fluctuates up and down. Therefore, output values corresponding to a plurality of levels are selected.

Therefore, in the case of uniform density, there is no influence of the average density (reflectance) in a wide range, but the area where the desired density can be reproduced increases, and in the case of an image in which continuous gradation occurs, gradation is improved. Will decrease, and the resolution will also deteriorate. In addition, a texture or the like may occur when the density difference between the levels adjacent to each other is large.

Therefore, by interlocking with the converted multi-valued signal value to be changed, the average value of the converted multi-valued signals corresponding to the adjacent levels is set as the threshold value, so that the gradation and resolution can be improved. The decrease can be suppressed.

Next, the operation of this embodiment will be described with reference to FIGS.
This will be described with reference to the flowchart shown in FIG.

FIG. 13 shows a flow chart of the image forming apparatus of this embodiment after the power is turned on.

First, a control panel (not shown)
When the power is turned on while pressing the above predetermined button, the control means 1 determines that it is in the adjustment mode, performs the initial adjustment processing, and enters the standby state.

Here, the initial adjustment process will be described with reference to the flowchart shown in FIG.

That is, the control means 1 outputs the reflectance data of the test patterns corresponding to the two densities stored in the test pattern data memory 2a of the nonvolatile memory 2 to the pattern generator 3 and the input of the selector 4. To the pattern generator 3 side. As a result, the test pattern pixel signal from the pattern generator 3 is supplied to the correction means 10 in the error diffusion processing unit 6 via the selector. At this time, the initial values read from the initial value memory 2b are stored in the threshold value registers 21a, ... In the multi-value quantization means 11 and the reflectance memories 42a ,.

A test pattern is printed on the basis of the data converted into the multilevel signal through the error diffusion processing similar to the above. Then, under the control of the control means 1, the density from the detection means such as the output image density sensor 50c and the toner adhesion amount sensor 50d, or the detection value (target value QT) such as the toner adhesion amount is stored in the non-volatile memory 2. It

When the power is turned on without touching the above-mentioned predetermined buttons on the control panel, the control means 1 causes the threshold value register to store the respective threshold values previously stored in the present value memory 2d of the non-volatile memory 2. 21a, ..., The reflectance for each level is set in the reflectance memory 42a, ..., Each parameter is set in a predetermined register, and other peripheral devices are initialized.

After the warm-up operation, the image forming condition, that is, the error diffusion parameter optimizing process is performed, and the system enters the standby state.

FIG. 14 shows a flowchart of the process when a copy button from the control panel or a print start signal from the outside is detected in the standby state after the power is turned on as described with reference to FIG.

That is, when the copy button from the control panel for instructing the print start by the user or the print start signal from the external printer or facsimile controller is detected, the control means 1 executes the image processing with the set parameters. Then, the printing operation is performed. At this time,
The above-mentioned error diffusion processing is performed in the error diffusion processing unit 6, the document read data or the input image data from the outside is processed, the processed data is output to the printer, and the printer prints out based on the data.

Here, when the predetermined time or the predetermined number of prints has been reached since the image forming condition optimizing process was finally completed, the image forming condition optimizing process is performed again. By this operation, optimization is performed not only immediately after the power is turned on, but also for a change in the characteristics of the image forming system due to the passage of a long time or the printing of many sheets.

That is, when the copy button from the control panel and the print start signal from the outside are not detected in the above standby state, the error diffusion parameter optimizing process of the previous time is completed, and then a predetermined interval is given in units of minutes or hours. When a time elapses, a calendar timer, or a copy counter after printing a predetermined number of sheets, such as 100 to 1000 sheets, or when a start command signal is input from the outside, the respective information is transmitted via I / O to control means. The error diffusion parameter, which is an image forming condition, is optimized. When the normal printing operation or the error diffusion parameter optimizing process is completed, the standby state is resumed.

This error diffusion parameter optimization process is shown in FIG.
5, the output image density detecting means 50c and the toner adhesion amount detecting means 5 for detecting the adhesion amount of the toner image after development
The output of the detection means such as 0d is output to the control means 1 via the A / D converter 51.

A calendar for detecting the passage of a predetermined time
The output of the timer 50g and the copy counter 50f for detecting the formation of a predetermined number of sheets is output to the control means 1 via the I / O interface 52.

Next, the error diffusion parameter optimizing process will be described with reference to the flowchart shown in FIG.

After the warm-up is completed and after the above-mentioned predetermined time has elapsed and the predetermined number of prints has been reached, a weight status display indicating that the user is prohibited is displayed on the control panel and the apparatus status is displayed to the outside (printer / FAX controller, etc.). After setting the signal, the following processing is executed.

First, printing processing of a plurality of test patterns is performed. At this time, the donor adhesion amount or reflection density (reflectance) of each test pattern printed on the drum or paper is measured. The measurement result is the detection means 50c, 5
It is obtained as a detection signal from 0d (image forming condition variation signal, image density variation signal, toner adhesion amount variation signal, or the like). Each of these detection signals is compared with a target value, and a variation value (deviation) is extracted and stored in a current value memory 2d of the nonvolatile memory 2 based on the variation value, that is, within a predetermined allowable value or outside the range. The control means 1 determines whether or not the current converted multilevel signal value is changed. When the deviations for all the test patterns are within the allowable values, the current image forming conditions are stored in the non-volatile memory 2 and the optimization process is ended. At this time, the weight display and the busy signal are released and the operation returns to the standby state. When the deviation is outside the allowable range, the correction of the image forming condition is calculated based on the above-mentioned deviation,
Set the calculated image forming condition change amount.

After that, the test pattern is printed again, the fluctuation is detected, and the quality is judged. Based on the judgment result, the change amount is stored, and whether the process is finished or the correction process is further executed is selected and executed.

FIG. 17 is a graph showing the relationship between the density and the reflectance when the toner adhesion amount on the photosensitive drum is used. In the first quadrant of the figure, the vertical axis shows the original absolute reflectance or the absolute reflectance conversion value R ′ of reflectance data, and the horizontal axis shows the normalized reflectance data R.

In the third quadrant, the vertical axis indicates the toner adhesion amount Q on the photosensitive drum, and the horizontal axis indicates the printed image density conversion value D '. In the figure, R'1 to R'corresponding to three test pattern reflectance data Rtp1 to Rtp3, respectively.
3, Dtp1 to Dtp3, Q1 to Q3, and the respective relationships are shown by dotted lines. When the gamma characteristic is 1 ideally, the dotted line becomes a rectangle as described with reference to FIGS. 9 and 10.
Further, the first quadrant and the second quadrant can be formulated. Therefore, if the third quadrant or the fourth quadrant is known, the true value of the reflectance data can be uniquely converted by detecting the deviation of the toner adhesion amounts Q1 to Q3 when the image forming characteristic variation occurs. . In the same figure, Q and D'are linearly approximated for simplification. Therefore, each toner adhesion amount detecting means, the correlation between the print density and the toner adhesion amount is non-linear, and the difference between the solids causes It may not be the ideal system of.

FIG. 18 shows the toner adhesion amount Q and the density conversion value D '.
Two examples are shown in which the relationship of is non-linear. By detecting the toner adhesion amount for each predetermined test pattern in the initial adjustment processing and setting the result as the toner adhesion amount target value, the solid-to-solid difference with respect to the target value and the non-linearity of the system can be considered (see FIG. 13). Regarding the detecting means or the image forming system having different characteristics depending on the difference between the solids as in the examples shown in Example 1 and Example 2, the calculation of the correction amount is uniquely determined with respect to the deviation from the target value of the toner adhesion amount. Can not. Therefore, the calculation of the correction amount of the reflectance data with respect to the deviation of the toner adhesion amount is only an estimation, and the method of cyclically correcting the correction amount a plurality of times as shown in FIG. 16 is effective.

19 and 20 show graphs similar to those in FIG. 9 or FIG. Further consideration is needed to make multiple corrections cyclically. As described above, when only the toner adhesion amount target value at the time of initial adjustment is quantitatively reliable, the density deviation or reflectance deviation estimated from the toner adhesion amount deviation is an estimated value. For example, assuming that the output density deviations dD1 to dD3 (black circles) are estimated from the deviation of the toner adhesion amount, and the optimum value reflectance data change amounts dR1 to dR3 for correcting this are assumed, a test pattern corresponding to this reflectance data is printed. Even if it does, it becomes a value outside the target toner adhesion amount, and it is not possible to judge whether the correction is correct or not. This is because the multilevel L referred to here is a number representing each of a predetermined number of output stages and is represented by using a discrete natural number. For this reason, when a uniform density pattern is printed by assigning a predetermined level to all the pixels of one test pattern, it is natural that if the characteristics of the image forming system fluctuate, they are deviated from their target values and detected. is there.

Therefore, in this embodiment, as shown in FIG. 20, a test composed of a plurality of levels selected through gradation processing as in the case of setting a virtual output level and printing a test pattern. It uses a pattern. A virtual output level conversion value corresponding to reflectance data based on a predetermined reflectance is set, and a test pattern is printed through gradation processing based on the reflectance data. If it is the same as the density at the time of initial adjustment, the deviation of the toner adhesion amount detection value also matches the initial adjustment value, that is, the target value. From the reflectance data corresponding to the estimated virtual output level, the change amount corresponding to the actual output level (for example, L = 1, 2, ...) Is linearly interpolated and set. Even if the detection, estimation, and correction are cyclically performed based on this way of thinking, the target value can always be the detection value stored in the initial adjustment.
Alternatively, it is possible to absorb the non-linearity of the imaging system and the difference between the solids.

FIG. 21 shows a flowchart for calculating the correction amount. 22 to 27 are graphs showing the manner of calculation in the correction amount calculation process. Hereinafter, description will be made with reference to FIGS. 21 to 27.

The toner adhesion amount on the photosensitive drum is detected for a plurality of test patterns, compared with the target value, and the deviation is calculated (see FIG. 22). The deviation dQ is converted into a density deviation. At this time, the correlation of the toner adhesion amount / density deviation with respect to each of the converted reference patterns is obtained by referring to the data in the look-up table format stored in advance in the memory. FIG. 23 shows the relationship between the adhered amount deviation and the density deviation for three different patterns. If the reference is assumed to be linear as shown in FIG. 17, the three curves are expressed as the same straight line in the expression of FIG.

On the other hand, a multilevel signal value (output level) for the test pattern data (test pattern reflectance data) is calculated. FIG. 24 shows a graph showing the correspondence between the reflectance data and the output level. A virtual output level LTP corresponding to the test pattern reflectance data RTP is linearly interpolated and calculated from the current multilevel signal conversion value R0 corresponding to the actual output level.

The density deviation corresponding to each actual output level is calculated from the density deviations dD and LTP with respect to the previously obtained test pattern. FIG. 25 shows an example of this situation. At this time, 0 and the maximum output level are fixed and are not subject to correction.

FIG. 26 shows the absolute reflectance RABS for each output level. A new reflectance RABSnew is obtained from the previously obtained density deviation and the current reflectance data.

FIG. 27 shows a graph showing the correlation between the absolute reflectance and the normalized reflectance data.

A new multi-valued signal conversion value is calculated through the above steps and can be converged to the target value by cyclically repeating detection, estimation and correction.

As described above, according to the first embodiment,
Maintaining the gradation characteristics in the local area even if the gradation characteristics per pixel of the image forming system change due to the environment and time, and even if the number of gradation steps per pixel actually decreases due to the fluctuation. In addition, the output section resolution and the number of stages for maintaining the gradation are small, and the high-speed memory for maintaining the gradation is small.

Further, when the gradation expression by the image forming system is dominated by the density gradation characteristics, it is possible to accurately compensate the gradation characteristics by unifying the error in the density dimension.

Further, since the threshold value is also changed in conjunction, multi-valued processing corresponding to the changed output value is possible, so that the error occurrence for each pixel can be reduced.
Gradation and resolution can be kept in good condition.

Further, by setting the average of adjacent output values as the threshold value, the error calculation and the threshold value for multilevel conversion have the same dimension. Therefore, when the threshold value is set to the average value, the error per pixel is generated. Is minimized, and the gradation and resolution can be maintained in good condition.

Next, the second embodiment will be described.
In any of the following embodiments, after setting the image forming condition change amount, the test pattern is printed again, the change detection and the pass / fail judgment are performed, and the change amount is stored, the end, or the feedback control is executed based on the determination result. The difference between the respective embodiments will be described, and the details will not be described.

In the second embodiment, the scanner of the input device uses a standardized density signal (hereinafter referred to simply as a density signal) whose reflectance / density is converted by the same means as in the first embodiment or the logarithmic conversion means. When the pixel signal to be provided is input and the above-mentioned input device is input in the reflectance dimension, the predetermined reflectance / density conversion means performs multi-value conversion as it is in the density dimension. The multi-valued signal is transferred to the output device, and the output device performs gradation expression by changing the density per pixel based on the multi-valued signal.

FIG. 29 shows a block diagram of an image processing apparatus according to the second embodiment. Similar to the scanner of the first embodiment, the scanner is assumed to be input with the reflectance dimension represented by the equation (8). Further, the data stored in the non-volatile memory 2, the registers 21a, ... And the memories 42a, ... Are data corresponding to the density instead of the reflectance. For example, density data is stored as the converted multilevel signal value. Further, the reflectance memories 42a, ... Are density value memories.

By the way, the relationship between the reflectance and the density can be defined by the following concept.

The input pixel signal R (= f) is standardized to [0,1].

Here, in order to distinguish the normalized reflectance from R, let Rref be Rref = 1-R / 255 (11) The values normalized to the white reference plate and the black reference plate by the equations (9) to (11) are RW = 0 → RrefW = 1. ) RB = 255 → RrefB = 0 (13) Therefore, Rref expressed by the equation (11) is the reflectance of the white and black reference plates. Means the converted relative reflectance. The original data R after shading correction is R
It can be interpreted that ref is a value normalized to the interval [0,255]. When the densities of the white reference plate and the black reference plate are measurable, assuming that they are DW and DB, respectively, the converted absolute reflectances are RW '= 10 ^ (-DW) .... ····························································································································· (15) ′ = Aabs * Rref + Babs ・ ・ ・ ・ ・ ・ ・ ・ ・ ・ ・ ・ ・ ・ ・ (16) However, Aabs = (RW′−RB ′) / (RrefW−RrefB) = (RW′−RB ′) / (1-0) = RW'-RB '... (17) Babs = RB'-Aabs * RrefB = RB'-Aabs * 0 = RB' ... (18) From equations (14) to (18), R '= {10 ^ (-DW) -10 ^ (-DB)} * Rref + 10 ^ (- DB)・ ・ ・ ・ ・ ・ ・ ・ (19) Therefore, if DW and DB are known, the concentration conversion value D '
Is calculated by the following equation.

D ′ = − log (R ′) ... (20) Here, regarding the white and black reference plates according to the equation (20), The relationship is established.

DW '=-log (RW') (21) DB '=-log (RB') (22) If D'is standardized by [0, 255] and is newly set as D, D = AD * D '+ BD ... (23) However, AD = (DB'-DW ') / 255 (24) BD = -AD DB' (25) Therefore, in order to standardize the pixel signal R with the densities of the white reference and the black reference, any conversion means satisfying the equations (11) to (23) may be used. In the second embodiment, the density D corresponding to the pixel signal R is recorded in advance in the read-only high speed memory, and the density conversion means 18 is used as a look-up table (hereinafter referred to as LUT).
Is building.

In accordance with this, the multilevel signal conversion means 12
Then, the dimension of the converted multi-valued signal to be converted must be the same as that of the density converting means 18 described above. That is, the printer output density per pixel corresponding to each level of the multilevel signal is set. As a result, the error can be calculated with the unified dimension of the density. In the second embodiment, the reflectance, which is the contents of the reflectance registers 42a, ... Of FIG. 6 described in the first embodiment, is the density,
Similarly, the threshold value which is the content of the threshold value registers 21a, ... Of FIG. 2 is also set so as to change from the reflectance dimension to the density dimension.

Here, when the gradation characteristic variation of the output device occurs, the above-mentioned density converting means is not changed, and the multi-valued signal converted by the multi-valued signal converting means is changed as in the first embodiment. By changing the output density per pixel corresponding to each level, it is possible to maintain the gradation in the local area, and it is possible to perform correction with a small rewritable high-speed memory and in a short calculation time.

Next, the third embodiment will be described.

The third embodiment inputs a relatively small number of input level signals from a data generating means such as a computer for image creation, editing, recording, transfer, etc. An image processing system in which a level signal is directly transferred to an output device such as a printer and an output image is output from a printer that prints based on a pixel signal. The present invention relates to an image forming apparatus that compensates gradation characteristics in a local area of a corresponding output image. The system has the same number of input and output stages.

FIG. 30 shows a schematic block diagram of the image processing apparatus of the third embodiment. The number of levels of the input level signal from the data means and the output level signal to the printer is n (for example, four values). Pixel signal converting means 19 is provided for changing the difference in the allocation of the output level signal to the input level signal, and the converted multilevel signal value of the multilevel signal converting means 12 is also the pixel signal converting means 19 described above.
The reflectance is set to be the same as the dimension of the value converted by. For example, when the input level defines a predetermined reflectance standardized with respect to the level number, the multilevel signal conversion means 1 defined as the reflectance.
Since the converted value of 2 is not changed, the reflectance memories 42a, ... Of the multi-valued signal converting means 12 are configured by a look-up table (LUT). Further, in the nonvolatile memory 2, the test pattern data memory 2f in which the reflectance data of the 16-level test pattern is stored, the threshold value register 21a, ... An initial value memory 2g for storing the initial value of the threshold value and an initial value of the reflectance corresponding to each level stored in the reflectance memory 42a of the multi-valued signal converting means 12, ... Corresponding to each output level. The converted value characteristic table 2h in which the converted multi-valued signal value is stored as a converted value, the threshold value set in the threshold value register 21a in the multi-valued means 11, ... And the reflection of the multi-valued signal conversion means 12 The reflectance (converted multilevel signal value) corresponding to each level set in the rate memory 42a, ...
And a current value memory 2i in which is stored, and a standard value memory 2j in which a control standard value (whether the amount of change is within a permissible range) for quality determination is stored.

Therefore, immediately after the initial adjustment in the state where there is no gradation change, the defined reflectance conversion value of the printer output density corresponding to each level becomes equal to the above reflectance conversion value. Further, the content of the multilevel signal conversion means 12 at this time is also written with a value equal to the reflectance conversion value. Therefore, when the output level equal to the input level is output when the gradation change of the printer does not occur, the error in the dimension of the reflectance becomes 0, and the error diffusion processing unit 6 of the third embodiment substantially. No effect on (through pass).

On the other hand, when the gradation change of the printer occurs, the density change per pixel which has changed is detected as in the first embodiment, and the multi-value signal conversion corresponding to each output level converted into the reflectance is detected. By rewriting the converted multi-valued signal value of the means 12, the image processing apparatus of this embodiment diffuses the error to the peripheral pixels, thereby maintaining the gradation characteristic defined by the input level in the local area.

FIG. 31 is a block diagram relating to the charging, exposing and developing means of the color laser printer according to the fourth embodiment and its control means. In the figure, the photosensitive drum 71
Rotates counterclockwise (in the direction of the arrow in the drawing) with respect to the drawing. The charger 72 is mainly composed of a charging wire 73, a conductive case 74, and a grid electrode 75. The charging wire 73 is connected to a high voltage power supply 76 for corona and charges the surface of the photoconductor drum 71 by corona discharge. The grid electrode 75 is connected to a high voltage power supply 77 for grid bias, and controls the amount of charge on the surface of the photoconductor drum 71 by the grid bias voltage.

An electrostatic latent image is formed on the surface of the photosensitive drum 71 uniformly charged by the charger 72 by the exposure of the modulated laser beam light 79 from the optical system 78. The gradation data buffer 80 stores gradation data from an external device or controller (not shown), corrects the gradation characteristics of the printer, and converts it into laser exposure time (pulse width) data.

The laser drive circuit 81 uses the laser beam light 7
The grayscale data buffer 80 so as to be synchronized with the scanning position 9
The laser drive current (light emission time) is modulated according to the laser exposure time data from. Then, the semiconductor laser oscillator (not shown) in the optical system 78 is driven by the modulated laser drive current. As a result, the semiconductor laser oscillator emits light according to the exposure time data.

Further, the laser drive circuit 81 is provided in the optical system 7
The output of a monitor light receiving element (not shown) in 8 is compared with a set value, and control is performed to keep the output light amount of the semiconductor laser oscillator at the set value by the drive current.

On the other hand, the pattern generation circuit 82 generates gradation data of two test patterns having different densities, low density and high density, for measuring the toner adhesion amount, and sends them to the laser drive circuit 81. . The test pattern above is
It may be stored in the storage unit 61 described later.

Of the test patterns for the two gradation data, the one having a darker density is the high density test pattern and the one having a lighter density is the low density test pattern.

Now, the photosensitive drum 71 on which the electrostatic latent image is formed is developed by the developing device 83. The developing device 83
For example, in a two-component developing system, a developer containing toner and a carrier is stored, and the weight ratio of the toner to the developer (hereinafter, referred to as toner concentration) is measured by the toner concentration measuring unit 84. Then, the toner concentration measuring unit 84
The toner replenishing motor 91 for driving the toner replenishing roller 85 is controlled in accordance with the output of the toner replenishment roller 138 so that the toner in the toner hopper 86 is replenished in the developing device 83.

The developing roller 87 of the developing unit is formed of a conductive member, is connected to a high voltage power source for the developing bias, and rotates under the condition that the developing bias voltage is applied. The toner is attached to the image corresponding to the electrostatic latent image of. The toner image in the image area thus developed is transferred to the transfer sheet supported and conveyed by the transfer drum.

Further, the control circuit 88 generates gradation data from the pattern generation circuit 82 at the end of the warm-up process after the power is turned on, so that the photosensitive drum 71 is
Two high and low gradation patterns for measuring the toner adhesion amount are exposed on the top.

Then, the positions where the high and low gradation patterns on the photosensitive drum 71 are exposed are developed and synchronized with the arrival of the toner adhesion amount measuring unit 89 at the toner adhesion amount measuring unit 89. Measures the toner adhesion amount. The output of the toner adhesion amount measuring unit 89 is digitized by the A / D converter 90 and input to the control circuit 88.

The control circuit 88 has a toner adhesion amount measuring section 89.
Output (measured value) is compared with a preset reference value, and two of the grid bias voltage of the charger 72 and the developing bias voltage of the developing device 83, which are image forming conditions, are changed according to the comparison result. Perform the process.

Further, the control circuit 88 controls switching of gradation data from an external device or controller (not shown) and gradation data of a test pattern of the printer alone and a pattern for measuring the toner adhesion amount, and measuring units 89 and 84. Each output of the output, control of the output amount of the high voltage power supplies 76, 77, 92, setting of the target value of the laser drive current, setting of the target value of the toner density, toner replenishment control, correction processing of the gradation characteristics of the printer for gradation data. And so on.

The high voltage power supplies 77 and 92 are controlled by output voltage control signals supplied from the control circuit 88 via the D / A converters 93 and 94, respectively.

The control circuit 88 measures a rewritable storage unit 61 composed of an EEPROM or the like which is not erased even when the power is turned off, a storage unit 62 composed of a RAM or the like for storing data, a standby time and the like. It is composed of a timer 63 and a CPU 64 that controls the entire control circuit 88.

Various set values are stored in advance in the storage unit 61. For example, the initial grid bias voltage value and the initial development bias voltage value corresponding to the bias condition that provides the standard gradation characteristic of normal temperature and normal humidity. , Test pattern gradation data (high density part, low density part), predetermined target value for toner adhesion amount in high density part (used when calculating deviation), predetermined value for toner adhesion amount in low density part Target value (used when calculating deviation),
Control standard value for deviation of high density part, control standard value for deviation of low density part, coefficient representing surface potential characteristic, predetermined number of prints, predetermined elapsed time, maximum control count, bias condition value,
The upper limit and the lower limit (predetermined range) of the abnormal range of the toner adhesion amount measuring unit 89, the reflected light amount other than the test pattern region, the reflected light amount of the high density portion, and the reflected light amount of the low density portion are stored.

The bias condition values are the upper limit value and the lower limit value (predetermined range) of the grid bias and the developing bias, respectively, and whether the difference voltage between the grid bias and the developing bias is within the predetermined range.

The target value of the high density portion and the target value of the low density portion can be changed and input and displayed by the control panel 95.

Further, the storage unit 61 also stores a table regarding the change amount of the contrast voltage and a table regarding the change amount of the background voltage.

The storage unit 62 has a toner adhesion amount measuring unit 89.
Of the bias value (stored when the bias change mode is set) that was set before the abnormal state of, the counter that counts the number of times of control, the counter that counts the number of printed sheets, and the sensor that is turned on when the toner adhesion amount measuring unit 89 is abnormal. An error flag and a toner empty flag that is turned on when the toner is empty are provided.

FIG. 32 shows the grid electrode 75 of the charger 72.
With respect to the absolute value VG of bias voltage (hereinafter, simply referred to as grid bias voltage) with respect to V.sub.0, the surface potential of the photosensitive drum 71 uniformly charged by the charger 72 (hereinafter, referred to as unexposed portion potential) Vo, and The surface potential (hereinafter, exposure portion potential) VL of the photosensitive drum 71, which has been entirely exposed to a constant amount of light by the system 78, is attenuated and the developing bias voltage VD.
(Dashed-dotted line) is shown.

In this embodiment, the polarity of the voltage is negative due to the reversal development. When the grid bias voltage VG increases, the absolute values of the unexposed portion potential VO and the exposed portion potential VL respectively decrease. A linear approximation of the exposed portion potential VL and the unexposed portion potential VO with respect to the grid bias voltage VG can be expressed as the following equation.

VO (VG) = K1.VG + K2 (31) VL (VG) = K3.VG + K4 (32) where K1 to K4 are constants, VO, VG and VL are absolute values, and VO ( VG) and VL (VG) represent the magnitudes of VO and VL with respect to an arbitrary VG, where the absolute value VD of the developing bias voltage, the exposed portion potential VL and the unexposed portion potential VO.
Therefore, the development density changes. Now, the contrast voltage VC and the background voltage VBG are defined as follows.

VC = VD (VG) -VL (VG) ... (33) VBG = VO (VG) -VG (VG) ... (34) However, VD (VG) is the magnitude of VD with respect to an arbitrary VG. The contrast voltage VC is particularly related to the density of solid areas (see FIG. 33). The background voltage VBG mainly contributes to the density of the low density portion in the multi-gradation method using pulse width modulation (see FIG. 34).

Here, the following equation is obtained from equations (31) to (34).

VG (VC, VBG) = (VC + VBG-K2 + K4) / (K1-K3) (35) VD (VBG, VG) = K1.VG + K2-VBG (36) Formula (35), From (36), when the relationship (K1 to K4) between the exposed portion potential VL and the unexposed portion potential VO with respect to the grid bias voltage VG is known, the grid bias voltage VG, by determining the contrast voltage VC and the background voltage VBG, The developing bias voltage VD can be uniquely determined.

After measuring the surface potential of the photosensitive drum 71 in advance and obtaining the relationship (K1 to K4) between the exposed portion potential VL and the unexposed portion potential VO with respect to the grid bias voltage VG, the contrast voltage VC and the background voltage VBG are calculated. Set.
From the equations (35) and (36), the grid bias voltage V
G and the developing bias voltage VD are uniquely determined, a plurality of density patterns are imaged under these conditions, the toner adhesion amount Q after development of these is measured, and this measured value and a preset reference value are set. And the correction values ΔVC and ΔVBG of the contrast voltage VC and the background voltage VBG for achieving the proper development density are inferred from the deviation ΔQ. Based on this inference result, the grid bias voltage VG and the developing bias voltage VD are set again, the toner adhesion amount of the density pattern is measured, and the measurement is repeated until it is within the allowable range of goodness.

The bias changing step is mainly divided into three small steps. (1) A step of determining the amount of change in the potential relationship represented by two parameters from the relationship between both deviations, (2) The changed potential relationship and a coefficient representing the surface potential characteristic of the photosensitive drum 71 prepared in advance. Calculating a bias value to be changed from a function including
(3) Then, it is a step of setting change values calculated for the grid bias and the developing bias at respective predetermined timings.

This causes a problem in the method of directly selecting the developing bias voltage value and the grid bias voltage value from the table prepared in advance from the deviation of the high density portion and the deviation of the low density portion, respectively. In addition to the influence of the environment, the photosensitive drum 1,
The appropriate amount of change in bias varies depending on the use of developer, leaving history, and individual differences, and it also changes with time.Therefore, the convergence value when repeated detection and operation is changed from the target value over time. There is a possibility that it will come off.

In this case, as the contents of the table, the position type control data for correlating the contrast voltage value / background voltage value and the grid bias value / developing bias value are used to correspond to the contrast voltage change amount and the background voltage change amount. The method of velocity type control data is preferable.

Further, the effect of potential change acting on the high-concentration portion and the low-concentration portion is not necessarily independent, but has an interaction.
There is a contradiction in determining each bias value from each deviation.

(1) Therefore, the change amount of the potential relationship represented by two parameters is selected from the table prepared in advance based on the relationship between the deviation of the high density portion and the deviation of the low density portion.

One parameter is a contrast voltage representing a voltage between an exposure portion potential, which is the surface potential of the developing position when the entire surface is exposed with a predetermined exposure amount, and a developing bias voltage.
The other parameter is the background voltage which is the voltage between the unexposed portion potential, which is the surface potential of the developing position that is not exposed after charging, and the development bias voltage. Has a greater effect in the lower concentration portion.

In FIG. 35, the horizontal axis represents the gradation data and the vertical axis represents the output image density, showing the change in gradation characteristics when the contrast voltage is changed. Similarly, FIG. 36 shows a change in gradation characteristic when the background voltage is changed. However, the changes in the contrast voltage and the background voltage act on the high-density portion and the low-density portion, respectively, and there is an interaction in how they act.

Therefore, from the relationship between the deviation of the high density portion and the deviation of the low density portion, the table of the contrast voltage change amount,
A table of the background voltage change amount is prepared in the storage unit 61 based on the relationship between the deviation of the high density portion and the deviation of the low density portion, whereby the deviation amount of the contrast voltage from the deviation of the high density portion and the deviation of the low density portion, The amount of change in the background voltage is derived.

The contents of each table take into consideration the interaction between the contrast voltage and the background voltage, and the effective voltage change can be appropriately changed from the relationship between both deviations, and when both deviations are 0, each change amount is 0. Therefore, the steady-state deviation after convergence approaches 0.

(2) The obtained change amount of the contrast voltage, the change amount of the background voltage, the contrast voltage when the test pattern is formed, and the new contrast voltage and the background voltage to be changed from the background voltage are obtained.

Since these are parameters indicating the voltage relationship, the grid bias voltage value and the developing bias voltage value to be set for realizing these voltage relationships are calculated.

This calculation is uniquely obtained by a function (described in the above equations (35) and (36)) prepared in advance in the storage unit 61 that includes a coefficient representing the surface potential characteristic of the photosensitive drum 71. You can

(3) The obtained new grid bias voltage value and developing bias voltage value are applied to the respective high voltage power supplies 76, 9
Change the output control value to 2.

When the settings are changed and the test pattern is imaged again, the grid bias voltage value and the development bias voltage value are changed at predetermined timings.

Next, the qualitative algorithm regarding the contents of the above table will be described.

In this embodiment, in the step of deriving the change amount of the two potential relationships from the deviation of the high density portion and the deviation of the low density portion in the bias changing step, the contrast voltage is mainly decreased when both deviations are positive. When both deviations are negative, the contrast voltage is mainly increased. When the deviation in the high density area is within a predetermined value near 0, the background voltage is decreased when the deviation in the low density area is negative, and the deviation in the high density area is around 0. The background voltage is increased when the deviation of the low density portion is positive within a predetermined value of. This is considered so that the voltage relationship which is highly effective due to the action of the contrast voltage and the background voltage is mainly used.

FIG. 35 shows the effect of changing the contrast voltage on the gradation characteristics.

The horizontal axis shows the gradation data, and the vertical axis shows the output image density. It can be seen that when the contrast voltage is increased, the density on the high density side is increased and the gradient is increased.

FIG. 36 shows the effect of changing the background voltage on the gradation characteristic.

It can be seen that when the background voltage is increased, the development start of the low density portion is shifted to the higher gradation data, and the gradient is increased.

Also, from FIGS. 35 and 36, it can be seen that the effect on the gradation characteristics is greater even if the change amount of the background voltage is smaller than the change amount of the contrast voltage. further,
Since there is a risk of fogging of the photosensitive drum 71, adhesion of the oppositely charged toner, and carrier adhesion when the developer is a two-component developer, the background voltage is not largely changed, and the high-density portion is emphasized mainly by the contrast voltage. Coarse adjustment is performed, and fine adjustment is made according to the contrast voltage and the background voltage including the low density portion.

A table for deriving an amount of change for changing the potential relation as described above is prepared in the storage unit 61 in the control circuit 88 from the qualitative rule considering these.

FIG. 37 shows the contents of the table (stored in the storage unit 61) regarding the amount of change in the contrast voltage. The horizontal axis represents the deviation of the high density portion, the deviation of the low density portion in the depth direction, and the contrast voltage in the height direction. Both the deviation of the high density portion and the deviation of the low density portion are 0 at the center of the frame in the plane formed by the deviation axis of the high density portion and the deviation axis of the low density portion, that is, the toner adhesion amount of the high density portion and the toner of the low density portion This is the point where the adhered amount matches each target value. In this example, the change amount of the contrast voltage hardly depends on the deviation of the low density portion.

FIG. 38 shows the contents of the table (stored in the storage unit 61) relating to the change amount of the background voltage. With the same expression as in FIG. 37, when the deviation of the high density portion is largely deviated from 0, the change amount of the background voltage is 0, that is, it is not changed. The background voltage is changed only when the deviation in the high density portion is near zero.

When the change amount of the contrast voltage and the change amount of the background voltage are determined from the relationship between the deviation of the low density portion and the deviation of the high density portion, the manipulated variable change amount is independently determined for each deviation. In particular, the change amount of the background voltage may be erroneously determined. On the other hand, even when one deviation has the same value but the other deviation is different, an appropriate operation amount can be determined by the parameter change amount suitable for the effect.

39 and 40 show examples of changes in two different gradation characteristics, respectively. 39 and 40, it is assumed that the deviation in the low density portion is detected as the same value, and the deviation in the high density portion is very low in FIG. 39 and close to 0 in FIG. At this time, in the case of FIG. 39 in which the deviation of the high density portion is extremely low from the effect of the contrast voltage shown in FIG. 35 and the effect of the background voltage shown in FIG. 36, it is effective to mainly increase the contrast voltage. Then, in the case of the example of FIG. 40 in which the deviation in the high-density portion is close to 0, it can be inferred that the change in which the background voltage is slightly lowered is effective.

Instead of individually determining the manipulated variable from the deviation of the high density portion and the deviation of the low density portion, by considering the relationship between the deviation of the high density portion and the deviation of the low density portion as in the above example. It is possible to derive an appropriate amount of operation according to it.

In the first adhesion amount measuring step, when the deviation in the high density portion is slightly negative and the deviation in the high density portion is largely negative, the change amount of the contrast voltage is increased in the positive direction. The amount of background voltage change is 0 (no change)
(See FIGS. 41 and 42).

After using this result to calculate and change the bias value, the adhesion amount of the test pattern is measured again. As the effect of changing the bias, as expected from FIG. 35, both the deviation in the high density portion and the deviation in the low density portion should shift in the positive direction.

If it is within the standard value, the control is terminated.
The deviation in the high-density area is within the standard value, but if the deviation in the low-density area is slightly out of the standard value in the negative direction, as shown in FIGS. The amount of change becomes slightly negative, and the amount of change in the background voltage becomes slightly negative.

When the background voltage is lowered, the image density becomes higher on the low density side. The high-density portion should be slightly larger, but at the same time, the contrast voltage is slightly lowered, so that the high-density portion hardly changes.

By repeating the measurement of the adhered amount and the change of the bias as in the above example, the contents of the table of the storage unit 61 are changed according to the relationship between the deviation of the high density portion and the deviation of the low density portion.
Rough adjustment mainly for high density areas by changing the contrast voltage,
After that, sequential control such as fine adjustment including the background voltage and the contrast voltage even in the low density portion can be executed at the same time.

Next, with reference to FIGS. 45 to 48, changes in the toner adhesion amount and the bias value, which are inputs to the measurement system in the control process, will be described.

45 and 46 show the high density toner adhesion amount QH and the low density toner adhesion amount Q in a low temperature and low humidity environment, for example.
This is an example when both L are lower than the respective target values QHT and QLT. The horizontal axis of each of FIGS. 45 and 46 is the number of times of control, the vertical axis of FIG. 45 is the toner adhesion amount detection value, and the vertical axis of FIG. 46 is the bias value.

When the control count is 0, the grid bias voltage value VG and the developing bias voltage value VD are set to predetermined initial values to form high density and low density test patterns. The toner adhesion amount value QH in the high-density portion and the toner adhesion amount value QL in the low-density portion detected for the test pattern are lower than the target values QHT and QLT, respectively, and are outside the control standard values QHP and QLP. Therefore, the change amount is calculated in the bias changing step.

In this case, as in FIGS. 43 and 44, the high density portion is very small (the deviation of the high density portion is negatively large), so that the grid bias voltage value VG and the development bias are increased so that the contrast voltage is increased. The voltage value VD is changed (control count 1).

Then, the test pattern is formed and the toner adhesion amount is detected with the changed bias voltage value. Fig. 35
As can be seen from the above, by increasing the contrast voltage, both the toner adhesion amount values QH and QL increase and approach the respective target values (control count 1).

The toner adhesion amount value QH in the high density portion is lower than the target value QHT, and the toner adhesion amount value QL in the low density portion is larger than the target value QLT.

At this time, from the tables of FIGS. 41 and 42,
The amount of change that increases the contrast voltage a little and the background voltage is extracted, and the grid bias voltage value VG and the developing bias voltage value VD are changed according to the change amount of these voltages.
Is calculated and changed (control count 2).

Again, the test pattern is formed and the toner adhesion amount is detected with the changed bias voltage value. On this occasion,
The toner adhesion amount values QH and QL are control standard values Q, respectively.
Since it does not reach HP and QLP (control count 2), the bias change similar to the above is repeated (control count 3). As a result,
The toner adhesion amount values QH and QL are both control standard values QHP and Q.
Enter the LP and end the control. In this example, the maximum number of times of control is set to 5, but the number of times of control is set to 3 and converges and ends normally.

45 and 46 show, for example, a high density toner adhesion amount QH and a low density toner adhesion amount Q in a high temperature and high humidity environment.
This is an example when both L are higher than the respective target values QHT and QLT. The horizontal axis of each of FIGS. 45 and 46 is the number of times of control, the vertical axis of FIG. 45 is the toner adhesion amount detection value, and the vertical axis of FIG. 46 is the bias value.

In this example, the toner adhesion amount value QH in the high-density portion and the toner adhesion amount value QL in the low-density portion are higher than the target values QHT and QLT (the number of control times 0) at the initial bias value, and the contrast voltage is decreased. As a result, the grid bias voltage value VG and the developing bias voltage value VD are changed (control count 1). The toner adhesion amount value QH and the toner adhesion amount value QL of the low density portion are the target values QHT and QLT respectively.
Approach. After that, the background voltage and the contrast voltage are mainly changed to make them converge within the respective control standard values. In this example, the control requires four times for convergence.

However, if only the bias is changed,
It is not possible to completely deal with the nonlinear fluctuations in the highlight part and the intermediate density part. Further, as a side effect of changing the bias, there is a possibility of image defects such as missing of a solid portion due to carrier adhesion, adhesion of reversely charged toner, and chipping at the rear end of the solid portion in the transport direction. Further, when the maximum density of the output device is lowered, it is impossible to express the density equal to or higher than the lowered maximum density even if the converted multilevel signal value is changed.

Therefore, in this embodiment, the bias is changed within the range where no image defect appears, the raw adjustment is performed, and the fine adjustment is performed by changing the converted multilevel signal value and the threshold.

FIG. 49 shows a flowchart of the points of difference between the fifth embodiment and the fourth embodiment. In contrast to the fourth embodiment, the system from the pass / fail judgment to the setting value change is one system, whereas in the fifth embodiment there are two systems, and regarding the above-mentioned bias change, the subscript "-a", The subscript "-b" is added to the converted multilevel signal value and the threshold value change similar to those in the fourth embodiment.

In FIG. 49, in the control original mode setting, the control means is Correct = a (bias change). As a result, when the fluctuation is large, steps 26-a to 28
-Execute a. When the pass / fail judgment a is cleared by changing the bias, the control means b is selected (similar to the fourth embodiment), and steps 24-b to 28-b are executed. As a result, the target document mode is cleared and the process of changing the target document mode ends.

FIG. 50 shows a graph showing the control effect. The horizontal axis represents the number of corrections (changes), and the vertical axis represents the toner adhesion amount. In this example, the low concentration, the medium concentration, and the high concentration are started in a lowered state. High density and low density are detected to change the bias, and the bias is changed to the control standard values QH3-a, QL3-a, QH1-a, and QL1-a.
Cleared the second time. (The middle density of white circles is not used for changing the bias, so a test pattern is not normally printed; reference value). Here, the above-mentioned step 17-b is cleared to move to the conversion multilevel signal value change. 3 in this example
Adhesion amount detection (low concentration, intermediate concentration, high concentration) for one concentration, and control standards QH1-b, QL1-b, QH2-b, QL2-b, QH3-b, which are stricter than the above bias changes. QL3-b is set, and the converted multilevel signal values are changed to converge to the respective target values QT1, QT2, QT3, and in this example, the values reach the standard value for the fourth time.

The bias change will be specifically described.
Since the basic sequence is the same as that of the above-described fourth embodiment, only different points will be described.

Variable definitions are added as follows.

VG grid bias voltage real number 0 to 1000V VB development bias voltage real number 0 to 1000V VGnew grid bias voltage change value real number 0 to 1000V VBnew development bias voltage change value real number 0 to 1000V DG grid bias D / A value = 255 * ( VG-VGMIN) / (VGMAX-VGMIN) 0-255 VGMIN Grid bias lower limit voltage VGMAX Grid bias upper limit voltage VBMIN Development bias lower limit voltage VBMAX Development bias upper limit voltage DB Development bias D / A value = 255 * (VG-VGMIN) / ( VGMAX-VGMIN) 0-255 DGI initial grid bias D / A value 0-255 DBI initial development bias D / A value 0-255 VGI initial grid bias voltage = (VGMAX-VGMIN) * DGI / 255 + VGMIN real number 0-1000V VBI initial development bias voltage = (VBMAX-VBMIN) * DBI / 255 + VBMIN real number 0-1000 AVG unexposed part potential gradient real number B VG unexposed part potential cut side real number C VG exposed part potential gradient real number D VG exposed part potential cut edge real number VC contrast voltage real number VCnew contrast voltage change value real number VCI initial contrast voltage = VBI-C * VGI-D real number VBG background voltage real number VBGnew background voltage change value real number VBGI initial background voltage = A * VGI + B-VBI real number dVC contrast voltage correction amount real number dVBG background voltage correction amount real number bias change LUT VCTABLE (ΔQ (1 ), ΔQ (2)) Example FIG. 37 VGBTABLE (ΔQ (1), ΔQ (2)) Example FIG. 38, and the default value (voltage value is an absolute value) is tentatively defined as follows.

DGI = initial grid bias A / D value DBI = initial development bias A / D value VGMIN = 300V VGMAX = 1000V VBMIN = 100V VBMAX = 800V VGI = (VGMAX-VGMIN) * DGI / 255 + VGMIN VBI = (VBMAX -VBMIN) * DBI / 255 + VBMIN A = 0.9 B = 50 C = 0.1 D = 40 VC = VBI-C * VGI-D VBG = A * VGI + B-VBI VG = VGI VB = VBI.

Now, when the toner adhesion amount deviations corresponding to the two patterns are dQ [1] and dQ [2], respectively, the grid bias D / A value and the developing bias D / A value to be changed are determined by the following equations. it can.

Step 24-a. dVC = VCTABLE [dQ [1], dQ [2]] dVBG = VBGTABLE [dQ [1], dQ [2]] Step 25-a. VCnew = VC + dVC VBGnew = VBG + dVBG VGnew = (VCnew + VBG-B-G) / (A-C) VBnew = A * VGnew + B-VBGnew Step 26-a, 27-a IF VGnew <VGMINV (G) new> Provisional) IF VBnew <VBMIN (provisional) IF VBnew> VBMAX (provisional) IF VGnew <VBnew +50 (provisional) IF VGnew> VBnew +200 (provisional) THEN VG, VBG, VC, VBG are not changed D = 28-a INT (255 * (VG-VGMIN) / (VGMAX-VGMIN) +0.5) DB = INT (255 * (VB-VBMIN) / (VBMAX-VBMIN) +0.5) VC = VCnew VBG = VBGnew Grid bias D above After setting / A and the DB to the developing bias D / A, the fluctuation detecting step is executed again.

The sixth embodiment of the present invention will be described below.

FIG. 51 shows in detail the flow of image data (including gradation information) and the functional blocks relating to the exposure system according to this embodiment.

In normal printing, image data including gradation information (hereinafter referred to as gradation data) from an external device or the document reading unit and the image processing unit is processed in accordance with the image transfer clock and command / status information. , To the interface (I / F) 161 of this apparatus. The control unit 136 manages the transmission / reception of commands / statuses including data requests and printer busy.

The control unit 136 sets the selector 162 to select the grayscale data from the interface 161. The selected gradation data is sent to the data conversion unit 163, where it is converted into pulse width data of the laser beam light according to the conversion data given from the control unit 136 and stored in the line buffer 164. The data transfer up to this point is performed in synchronization with the image transfer clock.

[0200] Here, the selector 162 and the data converter 1
The gradation data buffer 137 is constituted by 63 and the line buffer 164.

On the other hand, the laser beam light emitted from the laser diode 165 in the optical system 113 passes through a pre-deflection optical system (not shown) and is deflected and scanned by a polygon mirror 167 rotated by a motor 166 as a deflecting means. The motor 166 is driven by the motor driver 168. Here, the laser diode 165 and the motor 16
6, polygon mirror 167, and motor driver 1
The optical system 113 is constituted by 68.

The horizontal sync detector 169 detects the position of the laser beam light 114 deflected by the polygon mirror 167 to generate a horizontal sync signal at the writing scanning position, and the horizontal sync signal is used to generate the horizontal sync signal.
Send to 70. The synchronization clock generation circuit 170 generates a write synchronization clock signal for each pixel based on the input horizontal synchronization signal, and the horizontal synchronization signal and the write synchronization clock signal are managed by the line buffer 164 and the control unit 136. , PWM (pulse width modulation) circuit 172 and laser driver 173, respectively.

The writing synchronization signal and the horizontal synchronization signal signal are synchronized with the scanning of the laser beam.

Under the control of the control unit 136, the counter 171 generates a timing signal for the writing area (top, bottom, right, left margin) according to the horizontal synchronizing signal, the writing synchronizing signal, and the print start position information.
Based on the timing signal and the synchronous clock signal, the read / transfer from the line buffer 164, the PWM circuit 1
The processing of 72 and the exposure writing of the laser driver 173 are executed in synchronization.

That is, the image transfer clock is used as a reference until the writing of the gradation data into the line buffer 164.
From the reading from the line buffer 164 to the exposure with the laser beam, the laser scanning position is used as a reference. This is because of the difference between the external image transfer rate and the writing speed due to the effective angle of view of the laser optical system, and this speed is absorbed by being temporarily stored in the line buffer 164.

The PWM circuit 172 generates a gate pulse corresponding to the light emission time of the laser diode 165 per unit pixel based on the pulse width data per unit pixel. The laser driver 173 supplies a drive current according to the gate pulse to the laser diode 165 to control the laser diode 165 to be turned on and off.

Therefore, when the pulse width data per unit pixel is large, the light emission time of the laser diode 165 per unit pixel is long, the energy exposed to the photosensitive drum is large, and the area is large. Conversely, when the pulse width data per unit pixel is small, the light emission time of the laser diode 165 per unit pixel is short, the energy with which the photosensitive drum is exposed is small, and the area is also small. As a result, the latent image pattern after exposure forms a latent image of a gradation pattern that is expressed in gradation by the exposure of the laser beam light 114 based on the pulse width data.

The light emission amount of the laser diode 165, that is, the exposure amount is controlled by the light amount control circuit 174. The laser diode 165 has main light emission (front surface) used for exposure and monitor light emission on the back surface, and includes a monitor diode (not shown) for detecting the monitor light emission amount. The light amount control circuit 174 is
The output of this monitor diode is detected and compared with a target value, and a signal for correcting the amount of laser drive current is generated for the laser driver 173 so as to reduce the deviation.

The horizontal sync detector 169, the sync clock generation circuit 170, the PWM circuit 172, the laser driver 173, and the light quantity control circuit 174 constitute the laser drive circuit 138.

When the pulse width correction characteristic is changed, it is changed by transferring the conversion data from the control unit 136 to the data conversion unit 163. This makes it possible to change the conversion characteristic (pulse width correction characteristic) of gradation data to pulse width data.

When the exposure amount, that is, the light emission amount of the laser diode 165 is changed, it is changed by transferring the target value of the light amount control circuit 174 from the control unit 136. Thereby, the exposure amount can be changed.

On the other hand, when the test pattern is created, the control of the control unit 136 switches the selection of the selector 162 to select the image data of the test pattern generated from the pattern generation circuit 139 and the internal image transfer clock. The pattern generation circuit 139 transfers the test pattern image data to the selector 162 in synchronization with the data including the gradation information according to the data from the control unit 136 in synchronization with the internal image clock. The data flow is the same as the image data from the interface 161.

Further, in order to set the image writing area to a predetermined test pattern size, the control unit 136 causes the counter 171 to operate.
Image writing area information (top, bottom, right, left margin) is rewritten to test pattern data, and test pattern exposure is performed according to this. Therefore, the type, size, and print position of the test pattern including gradation can be specified by the setting from the control unit 136.

FIG. 52 shows the effect of changing the light amount on the gradation characteristics. The horizontal axis represents the gradation data, the vertical axis represents the output image density ID, P0 is the reference light quantity, and P1 to P4 are the changed light quantities, and the gradation characteristics at that time are shown. P1 <P2 <
The relationship is Po <P3 <P4. It can be seen that the higher the concentration, the greater the change with respect to the change in the light amount, and the greater the light amount, the greater the gradation gradient, and the smaller the light amount, the smaller the gradient.

Therefore, by changing the light amount with respect to the gradation characteristics that have changed as the density increases, it is possible to perform correction so that the gradation characteristics become closer to the reference gradation characteristics. However, it is difficult to correct the non-linear change on the low density side only by changing the light amount.

Therefore, an exposure amount change amount table corresponding to the high density portion deviation is prepared, and the exposure amount change amount is derived from the high density portion deviation. When the deviation is "0",
Since the change amount is "0", the steady-state deviation after convergence is "0"
Approach.

In the present embodiment, the light quantity is calculated by the following formula for the voltage value given to the target value of the light quantity control means (light quantity control circuit 174) of the exposure optical system.

Pnew = P + ΔP where Pnew is the new light amount, P is the current light amount, and ΔP
Is the light amount change amount obtained by the table.

53 and 54, the toner adhesion amount measurement value Q in the control process and the light emission amount of the laser diode 165 (hereinafter,
An example of a graph showing the state of P) (hereinafter simply referred to as the laser emission amount) is shown. FIG. 53 is a plot of the high density toner adhesion amount measurement value QH with respect to the number of times of control. In the figure, the broken line QHT is the target value of the toner adhesion amount in the high density portion,
QHP is the control standard value for the high density portion.

FIG. 54 is a plot of the laser emission amount P with respect to the number of times of control. The number of control times "0" is QH when it is first measured in the toner adhesion amount measuring step.
Then, the condition when the test pattern is formed becomes the value of P. The QH is lower than the QHT, that is, the deviation is large in the negative. Therefore, a change is made to increase the laser emission amount P in the image forming condition changing step described above.

However, although the maximum density can be secured by changing the exposure intensity, it is difficult to correct the non-linear fluctuation of the intermediate density and low density areas.

Therefore, the maximum intensity is ensured by changing the exposure intensity, and the gradation characteristic is changed by changing the conversion multilevel signal.

FIG. 55 shows a flowchart of the process similarly to FIG.

Now, the exposure intensity correction variable will be defined.

P init exposure intensity (light intensity control target value) 0-255, integer P exposure intensity (light intensity control target value) 0-63, integer dP exposure intensity correction amount 0 ± 255, integer LUT (look・ Up table) PTABLE (ΔQ (N)) size 255B Test pattern level LTP [M, N] ... Large changeability pattern number N = 0 to NMAX Default value P = PINIT Step 24-a. Reference correction amount LUT dP = PTABLE [dQ [N]] Step 25-a. Calculation of new set value P = P + dP Step 26-a. Is the value normal? And step 27. Correction process IF P <128 THEN P = 128 IF P> 255 THEN P = 255 Step 28-a. Set value change Light amount control circuit Target value register set to P value FIG. 57 shows an example of the control effect of the sixth embodiment. The exposure intensity is increased by the first correction, and the high density area has the standard value QH3−.
a, QL3-a, and control standards QH1-b, QL1-b, QH2-b, QL2-b, QH by changing the conversion multilevel signal value.
Cleared the third time against 3-b and QL3-b.

[0226]

As described above in detail, according to the present invention,
Even if the substantial gradation expression by the image forming system has the characteristics of area gradation, density gradation, or both, even if the gradation characteristic variation in the image forming system occurs due to environmental changes / aging, Furthermore, even if the number of gradation steps per pixel is reduced due to the fluctuation, it is possible to accurately and automatically correct the fluctuations in gradation and resolution to a minimum in a short time, thus saving maintenance work.
Alternatively, it is possible to provide an inexpensive image processing apparatus and image forming apparatus that can reduce the total running cost.

[Brief description of drawings]

FIG. 1 is a block diagram showing the overall configuration of an image processing apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram showing the configuration of multi-value quantization means.

FIG. 3 is a diagram showing a configuration of a comparator.

FIG. 4 is a diagram showing a configuration of a comparator.

FIG. 5 is a diagram showing a configuration of a comparator.

FIG. 6 is a diagram showing a configuration of a multilevel signal conversion means.

FIG. 7 is a diagram showing a configuration of a weight count storage unit, an error filter unit, and an error storage unit.

FIG. 8 is a diagram for explaining stored contents of a weight count storage unit and stored contents of an error storage unit.

FIG. 9 is a diagram showing a relationship among a document density Di, a reflectance R, an output level L, and an output density Do for explaining a change state.

FIG. 10 is a document density Di for explaining a change state,
The figure which shows the relationship of reflectance R, output level L, and output density Do.

FIG. 11 is a diagram showing an example of a converted multilevel signal value per pixel and a reflectance of each pixel actually printed.

FIG. 12 is a diagram showing a difference in reflectance corresponding to an output pixel depending on a threshold value.

FIG. 13 is a flowchart for explaining a process when the power is turned on.

FIG. 14 is a flowchart for explaining processing after ready.

FIG. 15 is a diagram showing various sensors connected to control means.

FIG. 16 is a flowchart illustrating an image forming condition optimizing process.

FIG. 17 is a diagram showing a relationship between reflectance data, a toner adhesion amount, and a density conversion value.

FIG. 18 is a diagram showing a relationship between reflectance data, a toner adhesion amount, and a density conversion value.

FIG. 19 is a document density Di for explaining the state of fluctuation,
The figure which shows the relationship of reflectance R, output level L, and output density Do.

FIG. 20 is a document density Di for explaining the state of fluctuation,
The figure which shows the relationship of reflectance R, output level L, and output density Do.

FIG. 21 is a diagram showing a flowchart for calculating a correction amount.

FIG. 22 is a diagram showing a standardized state of test pattern reflectance data.

FIG. 23 is a diagram showing the relationship between the density deviation estimated value and the test pattern toner adhesion amount deviation.

FIG. 24 is a diagram showing a relationship between reflectance data and an output level.

FIG. 25 is a diagram showing a relationship between an estimated density deviation value and a multi-valued image signal.

FIG. 26 is a diagram showing a relationship between an absolute density value and a multi-valued image signal.

FIG. 27 is a diagram showing absolute reflectance with respect to output level.

FIG. 28 is a flowchart showing an initial adjustment process.

FIG. 29 is a block diagram showing the overall configuration of an image processing apparatus according to another embodiment.

FIG. 30 is a block diagram showing the overall configuration of an image processing apparatus in yet another embodiment.

FIG. 31 is a block diagram showing the overall configuration of an image processing apparatus in yet another embodiment.

FIG. 32 is a diagram showing an unexposed portion potential, an exposed portion potential, and a developing bias voltage of a photosensitive drum with respect to a grid bias voltage of a charger.

FIG. 33 is a diagram showing an image density of a solid portion with respect to a contrast voltage.

FIG. 34 is a diagram showing the relationship between the potential of an unexposed portion on the surface of the photosensitive drum, the voltage of a low-density pattern, and the developing bias voltage.

FIG. 35 is a diagram showing a toner adhesion amount with respect to gradation data when the background voltage is increased. 6 is a flowchart for explaining a processing operation in a bias changing mode.

FIG. 36 is a diagram showing a change in gradation characteristic when the contrast voltage is changed.

FIG. 37 is a diagram showing the contents of a table relating to the amount of change in contrast voltage.

FIG. 38 is a diagram showing the contents of a table related to the amount of change in background voltage.

FIG. 39 is a diagram showing an example of variation in gradation characteristics.

FIG. 40 is a diagram showing an example of variation in gradation characteristics.

FIG. 41 is a diagram showing the change amount of the contrast voltage when the deviation in the high density portion is slightly negative and the deviation in the high density portion is largely negative.

FIG. 42 is a diagram showing the amount of change in the background voltage when the deviation in the high-density portion is slightly negative and the deviation in the high-density portion is significantly negative.

FIG. 43 is a diagram showing the amount of change in the contrast voltage when the deviation in the high density portion is within the standard value and the deviation in the high density portion is slightly negative.

FIG. 44 is a diagram showing the amount of change in the background voltage when the deviation in the high-density portion is within the standard value and the deviation in the high-density portion is slightly negative.

FIG. 45 is a diagram illustrating a change in the toner adhesion amount, which is an input of the measurement system in the control process.

FIG. 46 is a diagram for explaining changes in the bias value that is an input of the measurement system in the control process.

FIG. 47 is a diagram illustrating a change in the toner adhesion amount which is an input of the measurement system in the control process.

FIG. 48 is a diagram illustrating a change in a bias value that is an input of the measurement system in the control process.

FIG. 49 is a flowchart for explaining an image forming condition optimizing process of another embodiment.

FIG. 50 is a diagram showing the relationship between the toner adhesion amount and the number of corrections.

FIG. 51 is a block diagram showing the overall configuration of an image processing apparatus in yet another embodiment.

FIG. 52 is a diagram showing the relationship between light intensity and gradation data.

FIG. 53 is a diagram showing the relationship between the toner adhesion amount and the number of times of control.

FIG. 54 is a diagram showing a relationship between a bias value, a light emission amount, and the number of times of control.

FIG. 55 is a flowchart for explaining an image forming condition optimizing process according to another embodiment.

FIG. 56 is a diagram showing the relationship between the number of corrections and image output characteristics.

[Explanation of symbols]

1 ... Control means, 2 ... Non-volatile memory, 2a ... Test pattern data memory, 2b ... Initial value memory, 2c ... Multi-valued signal conversion table, 2d ... Present value memory, 2e ... Standard value memory, 3 ... Pattern generator, 4 ... Selector, 5
... buffer, 6 ... error diffusion processing section, 10 ... correction means, 1
DESCRIPTION OF SYMBOLS 1 ... Multi-value conversion means, 12 ... Multi-value conversion signal conversion means, 13 ... Error calculation means, 14 ... Error correction amount calculation means, 15 ... Error filter means, 16 ... Weight count storage means, 17 ... Error storage means, 21a , ~ 21o ... Threshold register, 42a,
~ 42p ... Reflectance register

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI technical display location G06F 15/68 310 J

Claims (2)

[Claims] 1. An image forming apparatus that multi-values a plurality of pixel signals forming an image signal into a predetermined number of levels and compensates a multi-valued error corresponding to the pixel signal in a local region Input the corrected pixel signal
Multivalued means for providing the multileveled signal by multileveling the corrected pixel signal into a signal having a predetermined level, and the multileveled value by generating a predetermined signal corresponding to each level of the multileveled signal. A multilevel signal conversion means for converting the multilevel signal to provide a multilevel signal, and an error calculation for calculating an error between the corrected pixel signal and the multilevel signal and outputting a multilevel error signal. Means, an error storage means for storing the error amount calculated by the error calculating means as a sum of the error amount previously stored for each relative position with respect to the current pixel of interest, and a position of the position corresponding to the current pixel of interest. A correction unit that outputs the corrected pixel signal to the multi-value quantization unit based on the input pixel signal and the correction amount, using a total error amount as a correction amount, and relates to an image density formed based on the multi-valued signal. To detect the amount Means, pattern forming means for forming a pattern for density fluctuation detection by the detecting means, and changing means for changing the conversion value corresponding to each level of the multilevel signal in the multilevel signal converting means, The conversion value corresponding to each level is changed by the changing means based on the detection result of the detecting means corresponding to the plurality of predetermined patterns formed by the pattern forming means, and the pattern forming means forms a pattern again, An image forming apparatus comprising: a control unit that repeatedly executes a detection operation by the detection unit. 2. The control means comprises a determination means for determining whether or not the change processing by the change means is required to be executed based on the detection result of the detection means.
The image forming apparatus according to item 1.