JP2013148882A - Diffused reflected light output conversion method, powder sticking amount conversion method, and image forming device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a diffused reflected light output conversion method capable of reducing approximate error in calculating a correction coefficient in a sticking amount conversion algorithm, a powder sticking amount conversion method, and an image forming device.SOLUTION: A gradation pattern 70 formed on a detection object surface 18 is detected by optical detection means 40. A sticking amount conversion algorithm utilizing regular reflected light and diffused reflected light and consisting of the following flow: (1) output conversion (regular reflection output → calculation of a regular reflected light output conversion value (normalized value); (2) output conversion (diffused reflected light output → calculation of a diffused reflected light output conversion value); (3) approximation; (4) correction of the diffused reflected light output conversion value (calculation of a correction coefficient and conversion of the diffused reflected light output), recording means, and arithmetic means, utilizes also stored data in addition to data newly acquired when obtaining the correction coefficient in the sticking amount conversion algorithm when obtaining the correction coefficient in (3) and (4) of the sticking amount conversion algorithm, and calculates the correction coefficient by the arithmetic means.

Description

  The present invention relates to a diffuse reflected light output conversion method in powder adhesion detection such as toner, a powder adhesion amount conversion method using this method, and an electronic device such as a copying machine or a laser beam printer that can implement the method. The present invention relates to an image forming apparatus using a photographic system.

Conventionally, in an image forming apparatus such as an electrophotographic copying machine or a laser beam printer, a density detection toner patch is provided on an image carrier such as a photoconductor in order to always obtain a stable image density. (Tone pattern) is created, the patch density is detected by optical detection means (hereinafter also simply referred to as a sensor), and the development potential is changed based on the detection result (specifically, LD power, charging bias) , Change of development bias).
As such a density detection patch detection means, a reflection type sensor in which an LED as a light emitting element (light emitting means) and a PD (photodiode) or PTr (phototransistor) as a light receiving element (light receiving means) is generally combined. Are known.

As shown in FIG. 2, the sensor configuration includes a type that detects only specularly reflected light (see Patent Document 1), and a type that detects only diffusely reflected light as illustrated in FIG. 3 (see Patent Documents 2 and 3). 4) As shown in FIG. 4, there are three types of types that detect both simultaneously (see Patent Document 4).
2, 3, and 4, reference numerals 50 </ b> A, 50 </ b> B, and 50 </ b> C denote element holders, 51 denotes an LED, 52 denotes a regular reflection light receiving element, 53 denotes a detection target surface, and 54 denotes a toner patch on the detection target surface. , 55 denotes a diffuse reflection light receiving element.

  In recent years, as shown in FIG. 5, a type in which a beam splitter is provided in the light path on the light emitting side and the light receiving side (see Patent Documents 5, 6, 7, etc.) has come to be frequently used. In FIG. 5, 56 is an LED, 57 and 58 are beam splitters, 59 is a photodiode as a light receiving means for P wave light (regular reflection light), and 60 is a light receiving means for S wave light (diffuse reflected light). Each photodiode is shown.

By the way, it has been found that the conventional optical detection means (sensor) has an output variation of slightly less than twice on the light emitting element side and slightly less than four times on the light receiving element side. The size of the element variation seems to vary depending on the type of element (top view type, side view type) and manufacturer, but at least the variation at the level that requires adjustment should be no matter which element is used. It is.
From this point, in order to accurately detect the amount of adhesion by the method described in the prior art, it is necessary to strictly adjust the output at the shipping inspection stage of the sensor (element).

Therefore, what happens when there is no adjustment will be described below based on the prediction result based on experimental data.
FIG. 6 shows the result of measuring the amount of color toner adhering to the transfer belt using the sensor shown in FIG. 4. The horizontal axis: the amount of adhesion, the vertical axis represents the specular reflection light output voltage, and the diffuse reflection light output. The voltage is plotted.

  Here, even when there is element variation in each of the light emitting element, the specular reflection light receiving element, and the diffuse reflection light element, at least for the specular reflection light output, since it has the characteristic that the output is maximum at the belt background, If the LED current is adjusted so that the output at the belt background becomes a certain value (3.0 V in this case), output variations due to variations in light emitting elements and regular reflection light receiving elements can be absorbed. An almost unique output characteristic can be obtained as the sensor output with respect to the quantity.

  The large squares in FIG. 6 are the points where the diffused light output after LED adjustment is plotted. If the variation in the light receiving element is doubled, the light receiving element with the diffuse reflected light output has a light receiving sensitivity of 1/2. Since the diffused light output at that time is an output (Vd / 2) represented by a small square, the difference between the diffused light output and the specularly reflected light (Vr) in each case gives the output relationship to the adhesion amount. It is not decided uniquely. This is the same even when the ratio is taken.

  Based on the above, the relationship between “regular reflection light output” and “diffuse reflection light output” when the amount of adhesion is converted based on the difference between “regular reflection light output” and “diffuse reflection light output” or ratio data. It is necessary to always satisfy a certain relationship, and for this purpose, for example, at the sensor shipping inspection stage, it is necessary to correct variations such as strictly adjusting the relationship between the specular reflection light output and the diffuse reflection light output with respect to a certain reference plate. Become.

Even if the method described in the above prior art was adjusted as described above, it is still necessary to simply take the difference or take the ratio. Because of belt fluctuations), accurate adhesion amount cannot be detected.
Further, in a color image forming apparatus, since fluctuations in image density lead to fluctuations in color, it is important to accurately detect the amount of density detection pattern adhesion and control the density in order to stabilize the image density.

Since the image density to be stabilized here is “the image density of the output image”, it is desirable that the color image forming apparatus perform it on the transfer belt immediately before being transferred onto the paper, and the aim of the image density control is Since control is performed so that the maximum target adhesion amount becomes a target value, it is desirable that accurate detection is possible up to a high adhesion amount region.
Therefore, in Patent Document 8, in the detection of the adhesion amount of powder such as toner, a regular reflection light output conversion method and a diffuse reflection light output conversion method that can always stably and accurately detect the adhesion amount over the entire adhesion amount. , Powder adhesion amount conversion method, image forming apparatus capable of performing these methods, powder adhesion amount detection apparatus capable of performing the above method, and adhesion amount conversion algorithm for providing a gradation pattern used in the above method Is shown about.

  Further, in Patent Document 9, in this adhesion amount conversion algorithm, approximation accuracy is increased by performing polynomial approximation on the nonlinear relationship between the normalized value and the diffuse reflection light output conversion value.

  However, in the conventional example described above, the linear approximation or polynomial approximation performed in the adhesion amount conversion algorithm is not based on the physical relationship between the normalized value and the diffuse reflected light output, and is therefore established locally. Even if you do it, it may fail globally. If the diffuse reflected light output conversion value with the reference normalization value is calculated under the situation where the approximation fails, this becomes an approximation error. Due to the effect of this approximation error, diffused reflected light output (hereinafter referred to as “diffuse reflected light”) after correcting for fluctuations in the background in the normalized value (hereinafter referred to as “normalized value”) of the regular reflected light (regular reflection component) as a reference. The light output conversion value ”is also varied. For this reason, there is a problem that it is impossible to calculate the sensitivity correction coefficient (hereinafter referred to as “correction coefficient”) with high accuracy. Here, calculating the correction coefficient is generally synonymous with calibration.

Therefore, the error in the correction coefficient is directly connected to the adhesion amount conversion error, and adversely affects a series of control systems related to image density stabilization.
The present invention relates to a diffuse reflected light output conversion method that reduces and corrects an approximation error in calculation of a correction coefficient in an adhesion amount conversion algorithm, a powder adhesion amount conversion method using the same, and image formation that implements the method The purpose is to provide a device.

In order to achieve the above object, according to the first aspect of the present invention, a plurality of continuous gradation patterns of powders having different adhesion amounts formed on a detection target surface are arranged at positions facing the detection target surface. It consists of the following flow using regular reflection light and diffuse reflection light of a detected gradation pattern, which is detected by an optical detection means that has a light emitting means and a light reception means and can detect regular reflection light and diffuse reflection light simultaneously. Adhesion amount conversion algorithm “1” Output conversion (regular reflection output ⇒ calculation of regular reflection light output conversion value (normalized value))
Obtain the value obtained by multiplying the minimum value of the ratio of the regular reflection light output and diffuse reflection light output of the gradation pattern obtained by the diffuse reflection light output, and subtracting this value from the regular reflection light output. A regular reflected light output conversion value (= normalized value) that is a ratio to the regular reflected light output of the target surface is obtained.
“2” Output conversion (diffuse reflected light output ⇒ diffuse reflected light output conversion value calculation)
The diffuse reflection light output conversion value is obtained by subtracting from the diffuse reflection light output a value obtained by multiplying the regular reflection light output conversion value by the background diffuse reflection light output directly reflected from the background portion of the detection target surface.
[3] Function approximation is performed on the relationship between the regular reflection light output conversion value of the approximate gradation pattern and the diffuse reflection light output conversion value.
"4" Calibration of diffuse reflection output conversion value (calculation of correction coefficient and conversion of diffuse reflection output)
The diffuse reflection light output conversion value at the reference point, which is the reference regular reflection light output conversion value, is calculated from the function obtained in [3], and is multiplied by a correction coefficient so that this becomes the reference value. The light output conversion value is converted to a value that is uniquely determined in relation to the amount of adhesion.
And a regular reflection light output conversion value and a diffuse reflection light obtained from the gradation pattern when obtaining the correction coefficient in the adhesion amount conversion algorithms “3” and “4”. The combination data of the output conversion values is stored in the recording means, the gradation pattern is newly detected, and the stored data is also used in addition to the newly obtained data when the correction coefficient is obtained by the adhesion amount conversion algorithm. The correction coefficient is calculated by the calculation means.

  According to the present invention, the diffuse reflected light output conversion can be calibrated in a state where the influence of the approximation error is reduced, and the approximation error can be reduced using past data.

1 is a side view showing an overall configuration of a color image forming apparatus to which a diffuse reflection light output conversion method according to a first embodiment of the present invention is applied. It is a block diagram of the optical detection means of the type which detects only regular reflection light. It is a block diagram of the optical detection means of the type which detects only diffuse reflection light. It is a block diagram of the optical detection means of the type which detects a regular reflection light and diffuse reflection light simultaneously. FIG. 5 is a configuration diagram of an optical detection means using a beam splitter, which is a type that detects specular reflection light and diffuse reflection light simultaneously. It is a graph which shows the detection result of the regular reflection light output with respect to color toner adhesion amount, and a diffuse reflection light output. It is a graph which shows the regular reflection light output characteristic with respect to the black toner adhesion amount. It is a graph which shows the regular reflection light output characteristic with respect to the color toner adhesion amount. It is a graph which shows the diffuse reflected light output characteristic with respect to black toner adhesion amount. It is a graph which shows the diffuse reflected light output characteristic with respect to the color toner adhesion amount. It is a top view which shows a gradation pattern. Indicates that the light received by the specular reflection light receiving element as specular reflection light includes a diffuse reflection light component from the detection target surface and a diffuse reflection light component from the toner layer in addition to a pure specular reflection light component. It is a schematic diagram. It is a block diagram which shows the relationship between the reflected light component which should be actually detected by the optical detection means, and the reflected light component which should be removed. It is a graph which shows the relationship between the adhesion amount at the time of data sampling, and a detection output. It is a graph which shows the relationship between the sensitivity correction coefficient multiplied by a diffuse reflection light output, the adhesion amount, and a detection output. It is a graph which shows component decomposition | disassembly of regular reflection light. It is a graph which shows normalization of the regular reflection component of a regular reflection light output. It is a graph which shows the relationship between the surface part fluctuation | variation correction amount of a diffuse reflected light output, the adhesion amount, and a detection output. It is a schematic diagram which shows that a some component exists also in the component reflected from a belt background part. It is a graph which shows the relationship between the normalization value of a regular reflection component, and a diffuse reflection light output conversion value. It is a figure which adds the newly acquired data in a diffuse reflection light output conversion value-normalized value graph. It is a graph explaining that the function approximation accuracy can be maintained by specifying the data range in the diffuse reflected light output conversion value-normalized value. It is a graph explaining that the new data obtained in the diffuse reflected light output conversion value-normalized value is an abnormal value. It is a graph explaining the state where data exist too much in the diffuse reflection light output conversion value-normalized value. It is a graph explaining the correction coefficient value calculated by the adhesion amount conversion algorithm of the latest several times which is the object to be averaged. It is a graph in the case of adding new data to the diffuse reflected light output conversion value-normalized value. It is a graph explaining the correction process performed after new data addition to a diffuse reflected light output conversion value-normalization value. The flowchart of the calculation process of a diffuse reflected light output conversion value is shown.

The present invention has been made to solve the above-mentioned problems hidden in the prior art, and is achieved by an adhesion amount conversion algorithm according to the present invention described below and an image forming apparatus using the same. .
Hereinafter, a first embodiment of the present invention will be described with reference to FIGS. 1 to 28 (excluding FIGS. 2, 3, and 6).
As shown in FIG. 1, a schematic configuration of a color laser printer A of a quadruple tandem direct transfer system as an image forming apparatus and a powder adhesion amount detection apparatus in this embodiment will be described.

The color laser printer A has three paper feed trays, one manual feed tray 36, two paper feed cassettes 34 (first paper feed tray), and 34 (second paper feed tray). Transfer sheets (not shown) as sheet-like recording media fed from the manual feed tray 36 are separated one by one from the uppermost one by the manual feed roller 37 and conveyed toward the registration roller pair 23. The transfer sheets fed from the first sheet feed tray 34 or the second sheet feed tray 34 are separated one by one from the uppermost one by the sheet feed roller 35 and are transferred to the registration roller pair 23 via the transport roller pair 39. It is conveyed toward.
The fed transfer paper is temporarily stopped by the pair of registration rollers 23, the skew is corrected, and then the leading edge of the image formed on the photosensitive drum 14Y positioned at the uppermost stream described later and the transfer paper in the transport direction. The sheet is conveyed toward the transfer belt 18 at a timing that coincides with the predetermined position. At this time, the transfer sheet is conveyed toward the transfer belt 18 by the rotation operation of the registration roller pair 23 by ON control of a registration clutch (not shown).

When the transfer paper passes through a paper suction nip composed of the transfer belt 18 and a paper suction roller 41 in contact with the transfer belt 18, the transfer paper is attracted to the transfer belt 18 by an electrostatic force by a bias applied to the paper suction roller 41. It is conveyed at a process linear velocity of 125 mm / sec.
The transfer paper adsorbed to the transfer belt 18 is moved by the transfer brushes 21B, 21C, 21M, and 21Y disposed at positions facing the photosensitive drums 14B, 14C, 14M, and 14Y of the respective colors with the transfer belt 18 interposed therebetween. Transcribed. That is, a transfer bias (plus) having a polarity opposite to the charging polarity (minus) of the toner is applied to each of the transfer brushes. As a result, the toner images of the respective colors formed on the photosensitive drums 14B, 14C, 14M, and 14Y are transferred in the order of yellow (Y), magenta (M), cyan (C), and black (Bk). .

  The transfer paper that has undergone the transfer process of each color is separated from the transfer belt 18 by the curvature of the drive roller 19 on the downstream side, and conveyed to the fixing device 24. By passing through a fixing nip formed by the fixing belt 25 and the pressure roller 26 in the fixing device 24, the toner image is fixed on the transfer paper by heat and pressure. In the single-sided printing mode, the fixed transfer paper is discharged to an FD (face-down) tray 30 formed on the upper surface of the apparatus main body.

  When the double-sided printing mode is selected in advance, the transfer paper that has exited the fixing device 24 is sent to a reversing unit (not shown), and the front and back sides are reversed by the unit, and then the double-sided transport unit positioned below the transfer unit. It is conveyed to 33. The transfer paper is fed again from the double-sided conveyance unit 33 and conveyed to the registration roller pair 23 through the conveyance roller pair 39. Thereafter, it passes through the fixing device 24 through the same operation as in the single-sided printing mode, and is discharged to the FD tray 30.

Next, the configuration and image forming operation in the image forming unit of the color laser printer A will be described in detail.
Since the image forming unit has the same configuration and operation for each color, the configuration and operation for forming a yellow image will be described as a representative, and the other components are denoted by reference numerals corresponding to the respective colors and description thereof is omitted.
An image forming unit 12Y having a charging roller 42Y and a cleaning unit 43Y, a developing unit 13Y, an optical writing unit 16, and the like are provided around the photosensitive drum 14Y positioned on the most upstream side in the transfer paper conveyance direction.

At the time of image formation, the photosensitive drum 14Y is driven to rotate clockwise by a main motor (not shown), is neutralized by an AC bias (DC component is zero) applied to the charging roller 42Y, and the surface potential is a reference of approximately −50v. It becomes a potential.
Next, the photosensitive drum 14Y is uniformly charged to a potential substantially equal to the DC component by applying a DC bias with an AC bias superimposed on the charging roller 42Y. At this time, the surface potential is charged to approximately −500 v to −700 v (the target charging potential is determined by the process control unit).

Digital image information is sent as a print image from a controller unit (not shown). This digital image information is converted into a binarized LD light emission signal for each color, and a photoconductor by an optical writing unit 16 having a cylinder lens, a polygon motor, an fθ lens, first to third mirrors, a WTL lens, and the like. The exposure light 16Y is irradiated onto the drum 14Y. The drum surface potential of the irradiated portion becomes approximately −50 v, and an electrostatic latent image corresponding to the image information is formed.
The electrostatic latent image corresponding to the yellow image information on the photosensitive drum 14Y is visualized by the developing unit 13Y. By applying DC (-300 to -500 v) with an AC bias superimposed on the developing sleeve 44Y of the developing unit 13Y, toner (Q / M: -20 to -30 μC / g) is developed to form a toner image.

The formed toner images on the photosensitive drums 14B, 14C, 14M, and 14Y of the respective colors are transferred onto the transfer paper adsorbed on the transfer belt 18 by the transfer bias.
In the color laser printer A according to the present embodiment, in addition to the image forming mode as described above, a process control operation operation (hereinafter referred to as “process control operation operation”) is performed in order to optimize the image density of each color when the power is turned on or after a predetermined number of sheets have passed. Abbreviated as “procone operation”).

  In this process control operation, a gradation pattern (hereinafter abbreviated as P pattern), which is a density detection patch for each color, is formed on the transfer belt by sequentially switching between a charging bias and a developing bias at an appropriate timing. Next, the output voltage of these P patterns is detected by a density detection sensor (hereinafter abbreviated as P sensor) 40 which is an optical detection sensor disposed outside the transfer belt 18 in the vicinity of the driving roller 19. Further, the output voltage is converted by the adhesion amount conversion algorithm (powder adhesion amount conversion method) of the present invention to calculate (development γ, Vk) representing the current developing ability, and based on this calculated value Control for changing the developing bias value and the toner density control target value is performed.

The configuration of the P sensor 40 is as shown in FIG. 4, and the specifications thereof are as described above.
Here, PTr (phototransistor) is used as the light receiving element, but a light receiving element such as PD (photodiode) may be used.
Next, an adhesion amount conversion algorithm that is a main part in the embodiment of the present invention will be described.
Prior to the description of this algorithm, a basic adhesion amount conversion algorithm according to an embodiment of the present invention will be described with reference to FIGS. In this algorithm, a powder adhesion amount conversion method using the diffuse reflection light output conversion method is shown in which the diffuse light output is converted into an adhesion amount value according to the following procedure.
(1) Sampling the regular reflection light output and diffuse reflection light output of the gradation pattern. (2) By dividing the specular reflection light output into [regular reflection light component] (see FIGS. 7 and 8) and [diffuse reflection light component] (see FIGS. 9 and 10), only [specular reflection light component] is obtained. To extract. (3) [Diffusion light component from toner] is extracted by removing [diffuse reflection light component from belt background] (substantially sero) from the diffuse reflection output (see FIGS. 9 and 10).

  (4) The linear relationship with respect to the adhesion amount of two output conversion values that are independent (intersect) obtained from the above (2) and (3) is used. Further, in the adhesion amount range (low adhesion amount region) in which the adhesion amount detection by specular reflection light is possible, a diffuse reflection light output conversion value of a certain specular reflection light output conversion value (or adhesion amount) becomes a certain value. The sensitivity of the diffuse reflection light output conversion value is corrected. Thereby, the diffuse reflected light output (correction value) with respect to the adhesion amount can be uniquely determined. (5) The adhesion amount conversion process is performed from the relationship between the “adhesion amount” and the “diffuse reflected light output correction value” obtained in advance.

(1) to (5) will be described below in order.
Description of (1) In FIGS. 8 and 10, the density detection P pattern 70 shown in FIG. 11 formed on the transfer belt 18 (considered to be equivalent to the paper surface) is detected by the P sensor 40 shown in FIG. . The “regular reflected light output voltage” and “diffuse reflected light output voltage” detected thereon are plotted against the color toner adhesion amount [mg / cm ² ] precisely measured by the electronic balance. In the density detection P pattern (gradation pattern) 70, the toner adhesion amount increases on the upstream side in the belt movement direction. As the transfer belt 18 as a detection target surface in the conventional example, a 16-gradation pattern is formed on three types of transfer belts having different “specular gloss (Gs)” and “lightness (L *)”. The detection target surfaces of the three types of transfer belts are black belts: specular gloss: Gs (60) = 57, brightness: L * = 10
Brown belt: Specular gloss: Gs (60) = 27, Brightness: L * = 25
Gray belt: Specular gloss: Gs (60) = 5, Brightness: L * = 18
Was adopted.

Explanation for (2) Here, the regular reflection light output characteristic with respect to the black toner adhesion amount shown in FIG. 7 and the regular reflection light output characteristic with respect to the color toner adhesion amount shown in FIG. 8 will be compared. In FIG. 8, it can be seen that when the specular reflection light output exceeds a certain adhesion amount (in this case, 0.2 to 0.4 mg / cm ² ), the monotonous decrease has changed to the monotonous increase. This is because, as shown in FIGS. 12 and 13, the light received by the regular reflection light receiving element 52 as regular reflection light includes [diffuse reflection light component from the belt surface] in addition to pure [regular reflection light component]. This is because [diffuse reflected light component from toner layer] is included. In FIG. 12B, reference numeral 64 indicates a cyan solid portion, and reference numeral 53 indicates a belt surface that is a detection target surface.

  Considering that the irradiation light from the LED 51 is evenly diffused on the detection target surface as shown in FIG. 12, the diffuse reflected light component received by the regular reflection light receiving element 52 and the diffusion entering the diffuse light receiving element 55. There should be an n-fold relationship with the reflected light. The value of n times used here is a value determined by the optical layout such as the light receiving diameter and arrangement of the light receiving elements 52 and 55.

  The actual output is output as a voltage after the reflected light entering each of the light receiving elements 52 and 55 is IV-converted in the circuit by the OP amplifier. The difference between these is also accumulated, and the α-fold relationship should be established. If such a coefficient α can be obtained, it is considered that the specular reflection light output can be decomposed into “regular reflection component” and “diffuse reflection component”. Here, considering how to obtain the coefficient α, since the diffuse reflection component of Bk toner is so small that it is almost equal to zero, the Bk regular reflection light output characteristic shown in FIG. This is considered to be almost equal to the regular reflection component output characteristic from which the component is removed. As shown in FIG. 7, the specular reflection light output characteristic of Bk toner becomes almost zero or slightly positive as the adhesion amount increases, and never becomes negative. From this, the minimum value of the ratio between the specular reflection light output (see FIG. 8) and the diffuse reflection light output (see FIG. 10) is determined for each P pattern (see FIG. 11) of the color toner. Further, if the diffuse reflection light output is multiplied by the minimum value of this ratio and subtracted from the regular reflection light output, the output characteristics of only the regular reflection component as intended should be extracted.

The processing flow will be described below based on the output result of the brown belt (Gs = 27, L * = 25) shown in FIG. In addition, the meaning of the symbol (abbreviation) in the following description is as follows.
Vsg: Output voltage at the background of the transfer belt 18
Vsp ・ ・ ・ Output voltage of each pattern
Voffset ・ ・ ・ Offset voltage (Output voltage when LED51 is off)
_reg .... Specular reflection light output (Regular Reflection)
_dif .... Diffuse reflected light output (Diffuse Reflection, see JISZ8105 color terminology)
[n]: Number of elements: n array variable (STEP 1): data sampling: calculation of ΔVsp and ΔVsg (see FIGS. 14 and 15)
First, the difference from the offset voltage is calculated for all points [n] for both the regular reflection light output and the diffuse reflection light output. This is because, in the end, it is desired to express only “the increment of the sensor output by the change of the amount of color toner attached”.

However, when using an OP amplifier in which each offset voltage value (Voffset_reg .: 0.0621 V, Voffset_dif .: 0.0635 V) when the LED 51 is off is sufficiently small to a negligible level as in this embodiment. Such difference processing is not necessary.
(STEP2): Calculation of sensitivity correction coefficient α (see FIG. 15)
ΔVsp_reg. Determined in STEP 1 [N], ΔVsp_dif. [N], ΔVsp_reg. [N] / ΔVsp_dif. [N] is calculated, and the coefficient α to be multiplied by the diffuse reflected light output (ΔVsp_dif. [N]) is calculated when the component decomposition of the regular reflected light output is performed in STEP3.

The reason why α is obtained from the minimum value of the ratio is that it is known in advance that the minimum value of the regular reflection component of the regular reflection light output is almost zero and is a positive value. Here, the gradation pattern has at least one, preferably three or more adhesion amount patterns in the vicinity of the adhesion amount that provides the minimum value of the ratio between the specular reflection light output and the diffuse reflection light output. . At least one or more, preferably three or more, in the vicinity of the amount of adhesion at which the minimum value of the ratio between the specular reflection light output increment and the diffuse reflection light output increment obtained from the difference from each output value when the light emitting means is off is obtained. You may make it have an adhesion amount pattern. In addition, the regular reflection light output conversion value may have at least one, preferably three or more adhesion amount patterns within an adhesion amount range in which the linear reflection relationship with the adhesion amount is linear.
(STEP3): Component decomposition of specularly reflected light (see FIG. 16)
The component decomposition of the regular reflection light output is performed by the following equation.

When the components are decomposed in this way, the regular reflection component of the regular reflection light output becomes zero in the pattern portion where the sensitivity correction coefficient α is obtained.
By this processing, as shown in FIG. 16, the regular reflection light output is decomposed into [regular reflection light component] and [diffuse reflection light component].
(STEP 4): Regular reflection light output_Normalization of regular reflection component (see FIG. 17)
Next, in order to correct the difference in the specular reflection light output of the background portion of the three belts, the ratio of each pattern portion output to the belt background output is set to a normal value of 0 to 1. Convert to a digitized value.

FIG. 17 shows the results of conversion into normalized values obtained by performing the same processing for all three types of belts shown in FIG.
In this way, by analyzing the components of the specularly reflected light, only the specularly reflected light component is extracted and converted into a normalized value, thereby uniquely determining the relationship between the specularly reflected light component and the amount of adhesion. Can do. This value represents the exposure rate of the belt background portion, and in the adhesion amount range from zero to one layer formation, this normalized value (= belt background exposure rate) is relative to the adhesion amount. There is a linear relationship.

If it is desired to obtain the toner adhesion amount in the low adhesion amount region from {M / A} = 0 to 0.4 mg / cm ² , the relationship between the adhesion amount and the normalized value as shown in FIG. It is experimentally obtained in advance as mathematical formulas or table data. In this way, the amount of adhesion can be converted by inversely converting this or by referring to the table.
Here, compare with the prior art. In claim 4 of Patent Document 10, regular reflection light + (diffuse reflection light−diffuse reflection light output min) × predetermined coefficient is shown. In the examples in the specification, the corrected output has a primary correlation. Thus, there is a description that the predetermined coefficient is “−6”. However, it can be said that it is practically meaningless to multiply the predetermined coefficient having such a shape in that the characteristic variation of the optical detection means is not taken into consideration as described above.
On the other hand, in the basic technology of the embodiment of the present invention, since the coefficient calculated based on the sensor output of the regular reflection light and the diffuse reflection light is multiplied as the predetermined coefficient, the characteristic variation of the optical detection means is different. It is possible to perform highly accurate detection in consideration.

Explanation for (3) Next, processing for removing [diffuse reflected light output component from the belt background portion] from [diffuse reflected light output voltage] will be described.
What is finally desired in the basic adhesion amount conversion algorithm of the embodiment of the present invention is a unique relationship between the diffuse reflection light output and the toner adhesion amount.
However, as shown in FIG. 13, the light that enters the diffuse reflected light receiving element 55 includes diffuse reflected light (noise component) from the belt background in addition to the diffuse reflected light from the toner layer. This component needs to be removed from the output.

In FIG. 15, the ratio between the “background portion output” and the “pattern portion output” of the regular reflection component is uniquely determined with respect to the adhesion amount (adhesion amount detectable range: 0 to 0.4 mg / cm ² ).
In addition, in the diffuse reflection component from the toner layer, if the irradiation light to the detection target surface is constant, the relationship to the adhesion amount is uniquely determined (adhesion amount detectable range: 0 to 1.0 mg / cm ² ).
As a continuation of STEP 4, the processing flow will be described based on the output result of the brown belt (Gs = 27, L * = 25) shown in FIG.

As shown in the results of FIG. 14, the diffuse reflected light output from the belt background portion becomes maximum at the belt background portion where the toner is not attached, and the component gradually decreases as the toner adheres.
The relationship with the amount of adhesion of the diffuse reflection light output voltage increment due to the light that directly enters the diffuse reflection light receiving element 55 from the belt background is the exposure ratio of the transfer belt 18, that is, the specular reflection component of the specular reflection light output obtained previously. Is proportional to the normalized value (see FIG. 17). For this reason, the processing for removing [diffuse reflected light output component from the belt background portion] from [diffuse reflected light output voltage] is as follows.
(STEP5): Correction of background variation of diffused light output (see FIG. 18)

The results are shown in FIG. By performing such correction processing, the influence of the background portion of the transfer belt 18 can be eliminated. Accordingly, the [diffuse reflected light output directly reflected from the belt background portion (diffuse reflected light component)] can be removed from the [diffuse reflected light output] in the low adhesion amount range where the regular reflected light output has sensitivity. it can.
By performing such processing, the corrected diffuse reflection light output in the adhesion amount range from zero adhesion to one layer formation is converted to a value having a linear relationship with the adhesion amount through the origin. .

  Here, a supplementary explanation of the diffuse reflected light will be given. Since the specularly reflected light is light reflected from the surface of the detection target surface, as shown in FIG. 17, when the detection target surface is 100% covered with toner, the output changes substantially in the adhesion amount region beyond that. The normalized conversion value becomes almost zero. On the other hand, the diffuse reflected light is light that is reflected from the LED 51 and enters the toner layer, and thus is reflected multiple times. Therefore, as shown in FIG. 10, the high adhesion amount region in which the toner layer is covered 100% or more. However, the sensor output has a monotonically increasing characteristic.

  Therefore, as shown in FIG. 19, the light reflected from the belt background portion also includes a primary component reflected directly from the belt background portion and secondary and tertiary components reflected through the toner layer. is there. In the present embodiment, only the primary component is corrected in STEP5. Even with such correction alone, at least in the low adhesion amount region where sensitivity correction is performed, the influence of the belt background can be removed almost accurately, and the secondary and tertiary components are sufficiently smaller than the primary component. Therefore, a practically sufficient accuracy can be obtained even by correcting only the primary component.

Explanation of (4) By the above processing, in the low adhesion amount region where the specular reflection light output has sensitivity, the relationship between the toner adhesion amount and the specular reflection light can be uniquely expressed in (2) [regular reflection light component ] Only,
In (3), the [diffuse reflected light component directly reflected from the belt background] could be removed from the diffuse reflected light. Therefore, the sensitivity correction of the diffuse reflected light output is performed based on these.
Here, the reason for performing the sensitivity correction is to perform the following correction as described above.
(A) Correction for lot variation of light emitting element output and light receiving element output (b) Correction for temperature characteristic and aging deterioration characteristic of light emitting element output and light receiving element output The biggest point in this processing is that only one toner layer is formed. In the low adhesion amount area,
<1> The normalized value of the regular reflection light output (regular reflection component), that is, the exposure rate of the transfer belt background portion, has a linear relationship with the toner adhesion amount.
<2> [diffuse reflection component from toner layer] has a linear relationship that passes through the origin with respect to the toner adhesion amount.
That is, the sensitivity of the diffuse reflected light output is corrected by utilizing the fact that the two corrected outputs of the specular reflected light and the diffuse reflected light are both in a linear relationship with the toner adhesion amount. Several methods can be considered for this sensitivity correction. Here, the first method will be described as an example.

(STEP 6): Sensitivity correction of diffuse reflected light output (see FIG. 18)
<Processing formula by the first method>
As shown in FIG. 20, the diffuse reflected light output conversion value, which is the diffuse reflected light output after the background portion variation correction, is plotted against “normalized value of regular reflected light (regular reflected component)”. Further, the sensitivity of the diffuse reflected light output is obtained from the linear relationship in the low adhesion amount region, and correction is performed so that this sensitivity becomes a predetermined target sensitivity.

Here, the sensitivity of the diffuse reflected light output is the slope of the straight line shown in FIG. 20, and the diffuse reflected light output conversion value of the reference normalized value is a certain value (here, 0.3). The correction coefficient by which the current inclination is multiplied is calculated and corrected so as to satisfy 1.2).
(4-1) The slope of the straight line is obtained by the least square method.

In the present embodiment, the lower limit of the range of x used for the calculation is 0.06, but this lower limit is a value that can be arbitrarily determined within a range where x and y are in a linear relationship. The upper limit value is set to 1 because the normalized value is a value from 0 to 1.
(4-2) A sensitivity correction coefficient γ is set such that the diffuse reflection light output conversion value corresponding to the reference normalized value a calculated from the sensitivity thus obtained becomes the reference diffuse reflection light output conversion value b. Ask.

  (4-3) The diffuse reflected light output conversion value obtained in STEP 5 is corrected by multiplying by this sensitivity correction coefficient γ. A reference point for sensitivity correction (hereinafter abbreviated as “reference point”) (normalized value a serving as a reference) is in an area where the amount of adhesion can be detected by specularly reflected light.

  In Patent Document 9, in this adhesion amount conversion algorithm, approximation accuracy is increased by performing polynomial approximation on the nonlinear relationship between the normalized value and the diffuse reflection light output conversion value.

  The linear approximation and polynomial approximation performed in the conventional adhesion amount conversion algorithm as described above are not based on the physical relationship between the normalized value and the diffuse reflected light output conversion value, and thus are established locally. Even if you do it, it may fail globally. If the diffuse reflection light output conversion value at the reference point is calculated under the situation where the approximation fails, this becomes an approximation error. Due to the effect of the approximation error, the diffuse reflection light output conversion value at the reference point also has variations. For this reason, there is a problem that it is impossible to calculate the sensitivity correction coefficient (hereinafter referred to as “correction coefficient”) with high accuracy. Here, calculating the correction coefficient is generally synonymous with calibration.

Therefore, the error in the correction coefficient is directly connected to the adhesion amount conversion error, and adversely affects a series of control systems related to image density stabilization.
Therefore, the adhesion amount conversion algorithm in the embodiment of the present invention performs the adhesion amount conversion by correcting the basic algorithm configuration described above.
In the calculation of the correction coefficient by the adhesion amount conversion algorithm in the present invention, the approximation error can be reduced, calibration can be performed, and the powder adhesion amount calculation accuracy can be improved.
The diffuse reflected light output conversion method in the present invention (this embodiment) will be described below. This method is characterized in that, if there is no change in the light emission amount of the light emitting means, calibration is performed using the characteristic that the calculation result of the correction coefficient is theoretically different from the past result.

In this method, the diffuse reflected light output is calibrated according to the following procedures << 1 >> to << 5 >>.
<< 1 >> Sampling of regular reflection light output and diffuse reflection light output of a gradation pattern (corresponding to the control function in the preceding stage of claim 1).
Using the basic algorithm configuration (1), here, the density detection P pattern 70 shown in FIG. 11 formed on the transfer belt 18 is detected by the P sensor 40 shown in FIG. Output voltage Vsp_reg. [N] "and" diffuse reflected light output voltage Vsp_dif. [N] "are obtained simultaneously. The control of << 1 >> is shown in steps s1 and s2 of the diffuse reflection light output conversion flow in FIG.

<< 2 >> Acquisition and storage of normalized values and diffuse reflection conversion values (corresponding to the control functions “1” and “2” of claim 1 of the present application).
Using the basic algorithm configurations (2) and (3), the specular reflection output is changed to “regular reflection output ⇒ specular reflection output conversion value” (normalized value β [n]), and the diffuse reflection output is Conversion to “diffuse reflection light output → diffuse reflection light output conversion value” (conversion value ΔVsp_dif ′). If there is no change in the light emission amount of the light emitting means 51, the relationship between the normalized value and the diffuse reflection light output conversion value does not change. Therefore, every time this method is implemented, in addition to the normalized value and diffused light output conversion value obtained by << 1 >>, the normalized value and diffused light output conversion value obtained in the past << 1 >> are also called. It can be used. As described above, the combination data (hereinafter referred to as “data”) of the obtained normalized value and diffuse reflection output conversion value is stored in the recording means.
The control of << 2 >> is shown in steps s3 to s7 of the diffuse reflected light output conversion flow of FIG.

<< 3 >> Approximation of normalized value and diffuse reflected light output conversion value characteristics (corresponding to the control function of “3” in claim 1 of the present application).
Here, the basic algorithm configuration (4) is used to calculate the value of the diffuse reflected light output conversion value at the reference point a (reference normalized value). Here, approximation is performed using the fact that both the regular reflection light output conversion value (normalized value) and the diffuse reflection light output conversion value in the low adhesion amount region have a linear relationship with the toner adhesion amount.
In calculating the approximate curve 75, in addition to the data acquired this time (□ mark) converted by << 2 >> as shown in FIG. 21, data obtained in the past (× mark) stored in the recording means. Can be used by calling.
As described above, the approximate curve 75 is calculated by using the data obtained in the past together. However, since the number of data at this time can be increased, the approximation error at the reference point a can be reduced. It becomes possible. The configuration using the past data here corresponds to the control function in the latter stage of claim 1 of the present application. The control of << 3 >> is shown in step s8 of the diffuse reflection light output conversion flow in FIG.

<< 4 >> Calculation of diffuse reflection light output conversion value at the reference point (corresponding to the front part of the control function of “4” in claim 1 of the present application).
Using the basic algorithm configuration (4), the reference point a (a value that is an area in which the amount of adhesion can be detected by specular reflection light in this embodiment) with respect to the approximate curve obtained by << 3 >>. As a result, the diffuse reflected light output conversion value at 0.15 is calculated.
The control of << 4 >> is shown in step s9 of the diffuse reflection light output conversion flow of FIG.

<< 5 >> Calculation of correction coefficient (corresponding to the rear part of the control function of “4” in claim 1 of the present application)
Using the basic algorithm configuration (4), the ratio of the diffuse reflection output conversion value obtained in << 4 >> and the diffuse reflection output conversion value b, which is the reference value obtained in advance, is taken and corrected. The coefficient is γ. That is, the correction coefficient γ is multiplied so that the obtained diffuse reflection output conversion value becomes the reference value. With such processing, approximation errors can be reduced using past data.
Here, the configuration using the past data corresponds to the control function in the latter stage of claim 1 of the present application.
Control of << 5 >> is shown in step s10 of the diffuse reflection light output conversion flow in FIG.

Here, it is utilized that the two corrected outputs of regular reflection light and diffuse reflection light are in a first-order relationship with the toner adhesion amount. That is, the diffuse reflection light output conversion value is corrected for sensitivity so that the diffuse reflection light output conversion value at a certain regular reflection light output conversion value (or adhesion amount) becomes a certain value (0.15) at the reference point.
Here, it is necessary to be careful when calculating an output with a normalized value as a reference, which is a reference point a, by approximating with a non-linear function.
◎ Do not add a reference point to the data used for approximation. ◎ There are two ways that function approximation is performed in a range that does not fail.

  In order to prevent the reference point from becoming an additional point as in the former case, when performing the function approximation in << 3 >>, the data on the side larger than the reference point and the data on the smaller side (marked with □) are both nearest one point at a time. It is necessary to select and adopt the data in the vicinity of the reference point including. Further, in order to prevent the function approximation from failing as in the latter case, it is necessary to preferentially adopt data in the vicinity of a reference normalized value (indicated by a broken line in FIG. 22) that is the reference point a. In the present embodiment, data of each nearest point is secured on the larger side and the smaller side than the reference point as neighboring regions where these two are compatible. In addition, 0.05 <normalized value <0 as shown in the approximate curve 80 of FIG. 22 as a range in which the correlation coefficient between the measured value and the quadratic function used for approximation can be sufficiently recognized to be 0.9 or more. A range of .4 was set. By performing this setting, the approximation 81 can be performed with high accuracy.

Next, when performing measurement, an abnormal value may be mixed for some reason. When storing past data in << 2 >>, it is desirable to store these data excluding these abnormal values.
In order to determine the abnormal value, it is appropriate to obtain the difference 91 between the value when the data to be added to the function used to obtain the approximate curve 90 and the measured value are input as shown in FIG.
This difference has a threshold value (this time, the electrical noise of the diffuse reflected light sensor used, and the influence of fluctuation at the time of sampling of the detection pattern was within 5% of the diffuse reflected light output value conversion value. If the threshold is exceeded, it is determined that the data is an abnormal value. By not storing the data determined in this way, the approximation error can be further reduced and the abnormal value can be removed.

In addition, when the stored data is enormous as shown in FIG. 24, the normalized value and the diffuse reflection output conversion value cannot be stored, and the calculation of << 3 >> may take a very long time. Concerned. When this is a concern, the correction coefficient calculated in << 5 >> is stored in addition to the normalized value and the diffuse reflection light output conversion value. When the storage capacity specified in advance is exceeded, the combination of the normalized value and diffuse reflection output obtained by the adhesion amount conversion algorithm executed in the most recent several times is left, and the previous normalized value and diffuse reflection output are left. Delete the combination.
These are used for the calculation of << 3 >>. Further, as shown in FIG. 25, a weighted average of the approximation 92 which is the calculation result of << 5 >> and the approximation 93 which is the correction coefficient calculated immediately before several times is taken. As a result, the amount of memory used can be reduced in two senses, namely, the first is to reduce the amount of stored data and the second is to reduce the calculation load. In addition, it is possible to reduce the amount of calculation when performing function approximation using the least squares method, to cope with a memory shortage, and to cope with calculation load. It has been confirmed that the calculation accuracy of the correction coefficient by this method is almost equivalent to the approximation 80 in FIG.

  When the light emission amount of the LED 51 varies, the light reception amounts of the light receiving elements 52 and 55 also vary. At this time, the relationship between the normalized value and the diffuse reflected light output conversion value is different from that before the LED light emission amount change, and the data stored before the LED light emission amount change can be used in (2). Can not. For this reason, when there is a change in the amount of light emitted from the LED, the stored data becomes unnecessary, and the memory amount can be reduced by deleting this data.

Representative numerical examples in the present embodiment are as shown in FIGS. First, a gradation pattern of powders having different adhesion amounts formed continuously on a detection target surface is detected, and calibration is performed using these.
As shown in FIG. 26, the reference point 100 is set to 0.15 as a value that is an area in which the amount of adhesion can be detected by regular reflection light.
Also, it is assumed that combination data 101 of a total of 10 normalized values and diffuse reflected light output conversion values with reference point = 0.15 sandwiched in advance is stored. Next, when the gradation pattern is newly detected and the adhesion amount conversion algorithm is executed, the combination data 102 of the normalized value and the diffuse reflected light output conversion value is obtained. A quadratic approximation is performed from the combination data 101 and 102 by the least square method to obtain an approximate curve 103. In the obtained approximate curve 103, the diffuse reflection output conversion value 104 at the reference point 100 is obtained. As shown in FIG. 27, the ratio between the obtained diffuse reflection output conversion value 104 and the reference diffuse reflection output conversion value 105 is taken as a correction coefficient. In this way, approximation errors can be reduced using past data.

  After the correction coefficient is calculated, it can be converted into an adhesion amount based on a relational expression between the adhesion amount obtained in advance and the post-calibration diffuse reflection light output value or table data.

  This diffuse reflection light output conversion method can be used for an image forming apparatus and a powder adhesion amount detection apparatus using toner.

18 Detection target surface 40 Optical detection means 70 Gradation pattern A Color laser printer (image forming apparatus)

JP 2001-324840 A JP-A-5-249787 Japanese Patent No. 3155555 JP 2001-194443 A Japanese Patent No. 2729976 JP-A-10-221902 JP 2002-72612 A Japanese Patent No. 4456828 Japanese Patent No. 4533262 JP 2001-215850 A

Claims (7)

A plurality of powder gradation patterns with different adhesion amounts formed continuously on the surface to be detected are arranged at positions facing the surface to be detected and have a light emitting means and a light receiving means, and regular reflection light and diffuse reflection. Adhesion amount conversion algorithm “1” consisting of the following flow using regular reflection light and diffuse reflection light of the detected gradation pattern, detected by optical detection means that can detect light simultaneously Output conversion (regular reflection output ⇒ positive Calculation of reflected light output conversion value (normalized value)
Obtain the value obtained by multiplying the minimum value of the ratio of the regular reflection light output and diffuse reflection light output of the gradation pattern obtained by the diffuse reflection light output, and subtracting this value from the regular reflection light output. A regular reflected light output conversion value (= normalized value) that is a ratio to the regular reflected light output of the target surface is obtained.
“2” Output conversion (diffuse reflected light output ⇒ diffuse reflected light output conversion value calculation)
The diffuse reflection light output conversion value is obtained by subtracting from the diffuse reflection light output a value obtained by multiplying the regular reflection light output conversion value by the background diffuse reflection light output directly reflected from the background portion of the detection target surface.
[3] Approximation Function approximation is performed on the relationship between the regular reflection light output conversion value of the gradation pattern and the diffuse reflection light output conversion value.
"4" Calibration of diffuse reflection output conversion value (calculation of correction coefficient and conversion of diffuse reflection output)
The diffuse reflection light output conversion value at the reference point, which is the reference regular reflection light output conversion value, is calculated from the function obtained in [3], and is multiplied by a correction coefficient so that this becomes the reference value. The light output conversion value is converted to a value that is uniquely determined in relation to the amount of adhesion.
And a regular reflection light output conversion value and a diffuse reflection light obtained from the gradation pattern when obtaining the correction coefficient in the adhesion amount conversion algorithms “3” and “4”. The combination data of the output conversion values is stored in the recording means, the gradation pattern is newly detected, and the stored data is also used in addition to the newly obtained data when the correction coefficient is obtained by the adhesion amount conversion algorithm. And a diffused reflected light output conversion method, wherein a correction coefficient is calculated by a calculation means.   2. The diffuse reflection light output conversion method according to claim 1, wherein among the stored data, the specular reflection light output conversion value is in the vicinity of the reference point including one point closest to each of the larger side and the smaller side of the reference point. A diffuse reflected light output conversion method, wherein data is selected and used to calculate a correction coefficient.   3. The diffuse reflection output conversion method according to claim 1, wherein the diffuse reflection obtained by inputting the regular reflection light output conversion value of the pattern into the function used in the adhesion amount conversion algorithm when storing in the recording means. A diffuse reflected light output conversion method characterized by not including combination data in which an error between the light output converted value and the diffuse reflected light output converted value of the obtained gradation pattern exceeds a threshold value.   4. The diffuse reflected light output conversion method according to claim 1, wherein the calculation result of the correction coefficient is first stored in the recording means, and the data stored each time the adhesion amount conversion algorithm is executed exceeds the storage capacity. Delete as much old data as you can. Next, when the gradation pattern is newly detected and the correction coefficient is obtained by the above adhesion amount conversion algorithm, the stored data is also used in addition to the newly obtained data, and the correction coefficient is calculated by the calculation means. A diffuse reflected light output conversion method characterized in that a correction coefficient is obtained by a calculation means using a calculation result of a calculated correction coefficient and a stored correction coefficient.   5. The diffuse reflection light output conversion method according to claim 1, wherein the stored data and the correction coefficient are deleted when a change occurs in the light emission amount of the light emitting means. The diffuse reflected light output conversion method.   A powder adhesion amount conversion method using the diffuse reflection light output conversion method of claim 1.   An image forming apparatus comprising the method according to claim 1.

Images