JP2013257187A - Movement information detection device and multicolor image forming device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a movement information detection device suitable for real-time detection while eliminating an estimated error of a peak position, and a multicolor image forming device using the same.SOLUTION: A movement information detection device comprises: a light source; a coupling lens for guiding a beam emitted from the light source to irradiate a moving member; an image sensor having a plurality of pixels and acquiring one- or two-dimensional image; an imaging lens for imaging a scattered beam scattered by the moving member on the image sensor; image pattern acquiring means for acquiring an image pattern on the image sensor in time series; and movement information calculation means for calculating a cross-correlation function between two image patterns P1, P2 adjacent to each other in acquisition time, and calculating a speed or displacement of the moving member. The movement information calculation means for calculating the cross-correlation function includes a step of executing discrete Fourier transform and a step of removing background in a frequency space.

Description

The present invention relates to a movement information detection apparatus and a multicolor image forming apparatus.
The “multicolor image forming apparatus” of the present invention can be implemented as a digital copying machine, a laser printer, a laser plotter, a laser facsimile, or a complex machine thereof.

  Furthermore, it can be implemented as a multicolor image forming apparatus using an inkjet method.

  In addition, the “movement information detection apparatus” may be implemented as an apparatus that detects, as movement information, the speed and displacement of an intermediate transfer belt, a conveyance belt, and a sheet-like recording medium, which are movement members in these multicolor image forming apparatuses. it can.

Recent multi-color image forming apparatuses are so-called four photoconductors (image carriers) corresponding to toners of four colors (black, cyan, magenta, yellow) arranged in parallel to meet the demand for higher speed. The “tandem method” has become mainstream.
In the tandem system, each color toner image developed on each photoconductor must be finally superimposed on a sheet-like recording medium such as paper (standard paper, postcard, cardboard, OHP sheet, etc.).

As a superimposing method, there is a “direct transfer method” in which superimposing directly on a sheet-like recording medium.
Further, there is an “intermediate transfer belt method” in which each color toner image is superimposed on the intermediate transfer belt using the intermediate transfer belt, and the superimposed color toner images are collectively transferred to a sheet-like recording medium.
In these methods, unless the speed and displacement of the transport belt and the intermediate transfer belt are driven with high accuracy, “color shift” occurs in the obtained color image.
Further, in a multicolor image forming apparatus using an ink jet method, ink is directly ejected to a sheet-like recording medium that is driven to be displaced, so that the displacement of the sheet-like recording medium must be maintained with high accuracy.

  It is known that an object (moving member) such as an intermediate transfer belt or a sheet-like recording medium is irradiated with laser light, the scattered light is visualized as an image pattern by an image sensor, and the amount of displacement of the object is calculated. (Patent Documents 1 and 2).

  In order to drive the above intermediate transfer belt with high accuracy, for example, Patent Document 1 discloses the following method.

  That is, the belt is irradiated with laser light, and the pattern of the particle image generated on the belt surface is read. Then, the read pattern is analyzed, and the movement of the position of each particle in the particle image pattern accompanying the movement of the belt is detected as a velocity vector.

Then, based on the detection result, the position or operation of the belt is corrected to realize high-accuracy driving.
Patent Document 2 discloses an embodiment in which the paper conveyance amount of an ink jet printer is detected by a two-dimensional image sensor by detecting laser speckles from the paper, and drive control is performed to perform high-precision paper conveyance. ing.

  As a method for measuring displacement using an image pattern, the “evaluation value regarding the degree of coincidence of image patterns before and after movement” is calculated by shifting one pixel at a time, and the position having the highest degree of coincidence (this is referred to as “peak position”). ) Is known as an image displacement (Patent Documents 3 and 4).

The “evaluation value” includes “sum of squares of difference”, “absolute value of difference”, “cross-correlation function”, etc., and all of them are calculated in real space.
The “general calculation procedure” for the peak position is in the order of image pattern acquisition, evaluation value calculation, and peak position calculation.

In Patent Document 3, in order to remove the influence of shot noise and analog line noise caused by an image sensor and accurately perform “estimation of displacement”, after obtaining an image pattern, before calculating an evaluation value (absolute value of difference), Image improvement processing by image averaging (recursive filter) is performed.
That is, in the method disclosed in Patent Document 3, the peak position calculation procedure is image pattern acquisition, image quality improvement processing, evaluation value calculation, and peak position calculation as shown in FIG. The “image quality improvement process” is added to the “standard calculation procedure” before the evaluation value calculation.

  In Patent Document 4, as shown in FIG. 23 (b), after obtaining an image pattern, the obtained image pattern is subjected to “one-time differential processing” and then “evaluation value (cross-correlation function) calculation” is calculated.

Then, “peak value calculation” is performed thereafter, or, as shown in FIG. 23C, an evaluation value is calculated based on the acquired image pattern, and “double differential processing” is performed on the calculated evaluation value. Then, “peak value calculation” is performed to obtain the peak position.
In these peak position calculation procedures, "image improvement processing", "image pattern one-time differentiation processing" or "evaluation value (cross-correlation function) two-time differentiation processing" is performed in contrast to the general procedure described above. It has been added.

That is, in “peak value calculation” described in Patent Documents 3 and 4, “evaluation value calculation is executed in real space”.
Then, “calculation processing in real space (image improvement processing, one-time differentiation processing, two-time differentiation processing)” is added before or after evaluation value calculation, and the estimation error of the peak position is removed.

  Here, an example in which the inventors experimentally obtained a “cross-correlation function as an evaluation value” will be described.

While the white paper was being transported at 100 mm / second in a “direction parallel to the paper surface”, the moving paper was irradiated with “a beam from a red semiconductor laser light source (wavelength: 658 nm)”.
Then, the scattered beam from the paper is continuously imaged by an image sensor (128 × 32 pixels) of 504 fps (frames per second) through an imaging lens (focal length: f = 12.5 mm, F value: 6). The “image pattern” was obtained.

  The image patterns are two image patterns acquired adjacent in time.

  The “cross-correlation function as an evaluation value” was performed by a method using a discrete Fourier transform described later.

The correlation mutual function shown in FIG. 1 is composed of “peak part PCP and background part BGP”.
In this example, “the background portion BGP is larger than the peak portion PCP”.
Then, the “tilted background portion BGP” is superimposed on the peak portion PCP, and an error occurs in the estimation of the peak position.
For this reason, in order to eliminate an error, it is necessary to perform peak position calculation in a state where the “background portion to be superimposed” is subtracted from the correlation mutual function.

FIG. 2 shows a cross-sectional view of the image sensor in the longitudinal direction (128 pixel direction) through the peak portion PCP in FIG.
As shown in FIG. 2, the peak position is generally detected by “interpolating the background portion BGP in the vicinity of the peak portion PCP (the portion indicated by the broken line) and estimating and subtracting the background portion obtained by linear interpolation”. It is done by.
In this case, a specific calculation process is a three-stage calculation process (linear interpolation based on a certain algorithm, determining two points for linear interpolation, performing linear interpolation at the two points, and subtracting the background portion). Calculation for determining two points to be performed, calculation for linear interpolation, and calculation for subtracting the background portion).

In the method disclosed in Patent Document 4, the peak position is obtained after differentiating the obtained cross-correlation function twice as a method for removing the influence of the background portion.
That is, a “two-stage arithmetic process” is required in which the obtained cross-correlation function is differentiated once and further differentiated once more.
As another method, there is a method of “differentiating the image pattern once and then calculating the cross-correlation function”, but this method also differentiates the image pattern before movement once and the image pattern after movement once. A “two-stage arithmetic process” of differentiation is required.

  In the method described in Patent Document 4, in any of the above cases, in addition to “general calculation procedure of peak position”, “two-stage differential calculation processing in real space before and after the cross-correlation function calculation” is “ "It is necessary extra."

  Such “extra calculation processing” increases the calculation processing time, and is a serious detriment to real-time detection of the peak position, ie, displacement.

  It is known to use discrete Fourier transform for “calculation of cross-correlation function” (Patent Document 5). When the discrete Fourier transform is used, calculation in the frequency space becomes possible.

  The present invention has been made in view of the above-described circumstances, adopts a cross-correlation function using a discrete Fourier transform as an evaluation value, and adds “calculation processing in real space” before and after the calculation of the cross-correlation function. Therefore, it is possible to eliminate a peak position estimation error only by adding a minimum calculation process in a frequency space, and to realize a movement information detection apparatus suitable for real-time detection, and a multicolor image forming apparatus using the same. And

  The movement information detection apparatus according to the present invention is a movement information detection apparatus that detects the speed and displacement of a movement member as movement information, guides a light source and a beam emitted from the light source, and irradiates the movement member. Coupling lens, an image sensor having a plurality of pixels and capable of acquiring a one-dimensional or two-dimensional image, an imaging lens that forms an image of the scattered beam scattered by the moving member on the image sensor, and an image Image pattern acquisition means for acquiring image patterns on the sensor in time series, and movement information calculation means for calculating the cross-correlation function of two image patterns P1 and P2 adjacent to each other in acquisition time to calculate the speed and displacement of the moving member The calculation of the cross-correlation function in the movement information calculation means includes a step of executing a discrete Fourier transform and a step of executing a background removal process in the frequency space. Characterized in that it comprises a flop, and performing an inverse discrete Fourier transform, a.

  In the movement information detecting apparatus of the present invention, the calculation of the cross-correlation function includes a step of executing “forward / reverse discrete Fourier transform” and a step of executing background removal processing in the frequency space. Since the processing is easier and can be realized in a shorter time than "calculation processing in real space", it is possible to realize a movement information detection apparatus suitable for real-time detection and a multicolor image forming apparatus using the movement information detection apparatus. .

It is a figure which shows one example of a correlation mutual function. It is sectional drawing of the correlation mutual function of FIG. It is a figure for demonstrating movement information detection. The flowchart of the procedure which calculates a peak position is shown. It is a figure which shows the example of two image patterns acquired continuously in time. It is a flowchart which shows the procedure which calculates the deformed cross correlation function. It is a figure which shows one example of the low frequency removal pattern LC. It is a figure for demonstrating the effect of a background removal process. It is a figure which shows the example of two image patterns acquired as preliminary operation | movement before detecting a speed and a displacement. It is a figure which shows three examples of the low frequency removal pattern LC. It is a figure which shows the example of the two image patterns acquired. FIG. 11 is a diagram illustrating correlation patterns PS1, PS2, and PS3 calculated and calculated using the low frequency removal patterns LC1, LC2, and LC3 illustrated in FIG. It is a figure which shows another two examples of a low frequency removal pattern. It is a figure explaining that this invention is effective with respect to a subpixel process. It is a figure which shows the graph of the correlation peak position PP which performed the sub pixel process with respect to two adjacent image patterns, when a background removal process is performed. It is a figure which shows the graph of correlation peak position PP when not performing a background removal process. It is a figure for demonstrating one Embodiment of a multicolor image forming apparatus. It is a figure for demonstrating another form of implementation of a multi-color image forming apparatus. It is a figure for demonstrating an example of a structure of an optical scanning device. It is a figure which shows the schematic diagram of the structure of a liquid-crystal deflection | deviation element. FIG. 6 is a diagram illustrating an embodiment when the movement information detection device is applied to the detection of the speed fluctuation of the intermediate transfer belt of the multicolor image forming apparatus. 1 is a diagram illustrating an embodiment of a multicolor image forming apparatus that prints and records pixels with an ink discharge head. FIG. It is a figure for demonstrating a prior art.

Hereinafter, embodiments will be described.
First, the movement information detection of the moving member will be described with reference to FIG.
The movement information is “the displacement and speed of the moving member” as described above.
That is, only one of the displacement / velocity of the moving member can be detected as movement information, or both can be detected as movement information.

In FIG. 3, symbol LS indicates “linear stage”, symbol VCP indicates “air suction plate”, symbol S indicates “paper”, symbol LTS indicates “semiconductor laser”, and symbol CL indicates “coupling lens”.
Reference sign LN indicates “imaging lens”, reference sign IS indicates “image sensor”, reference sign PSN indicates “image pattern acquisition means”, and reference sign CON indicates “speed / displacement calculation means”.

  The sheet S is a sheet-like recording medium used in the multicolor image forming apparatus, specifically the above-described “standard sheet, postcard, cardboard, OHP sheet, etc.”, and “moving member whose movement information is to be detected” It is.

  The portions excluding the paper S, the linear stage LS, and the air suction plate VCP constitute one embodiment of the “movement information detection device”.

  The sheet S as the moving member is sucked and fixed in a plane on the linear stage LS by the air sucking plate VCP, and moved to the right (arrow direction) in the figure by the linear stage LS.

“Displacement and speed” of the paper S by the linear stage LS is movement information.
The semiconductor laser LTS is a “light source”, specifically a “red semiconductor laser” having an emission wavelength of 658 nm.
The coupling lens CL is a positive lens having a focal length of 4 mm, collects the beam from the semiconductor laser LTS, and irradiates the paper S.
The “irradiation angle” is inclined 45 degrees with respect to the paper surface.

  The imaging lens LN is a positive lens having a focal length of 12.5 mm and an F value of 6, and forms an image of “scattered beam” that is diffused light diffusely reflected by the paper S on the light receiving surface of the image sensor IS.

  The image sensor IS is a two-dimensional CMOS sensor in which 128 pixels are arranged in the moving direction of the paper S (left-right direction in the drawing) and 32 pixels are arranged in a direction orthogonal to the drawing at an arrangement pitch of 7.4 μm.

The imaging lens LN has a conjugate relationship between the surface of the paper S and the light receiving surface of the image sensor IS, and the imaging magnification is 0.82. Since the scattered beam from the paper S is coherent, a “speckle pattern” is generated on the light receiving surface of the image sensor IS.
The configuration of FIG. 3 is a configuration in which the scattered beam diffusely reflected by the moving member (paper S) is imaged through the imaging lens LN, and so-called “speckle pattern motion in the imaging region” is detected. It is.

  As the “light source”, it is preferable to use a laser light source capable of producing a high-contrast speckle pattern, particularly a “small and low-cost semiconductor laser LTS” as in this example.

Of course, the present invention is not limited to this, but the contrast of the speckle pattern is reduced, but a “lower cost LED light source” can also be used.
The linear stage LS performs “moving operation at a speed of 100 mm / second in the moving direction”.
Since the speckle pattern on the light receiving surface of the image sensor IS forms an inverted image, as the sheet S moves, “the moving speed of the sheet multiplied by the imaging magnification” is opposite to the sheet S. Moving.
During this movement, the image sensor IS continuously captures a speckle pattern at a constant time interval (frame rate 504 fps), and acquires the speckle pattern by the image pattern acquisition means PSN.
The speckle pattern acquired by imaging in this way is an “image pattern”.

  The velocity / displacement calculating means CON calculates the movement amount of the speckle pattern between temporally adjacent frames by “cross-correlation calculation between adjacent frames” of the acquired image pattern, and accumulates the movement amount between the adjacent frames. Then, the “cumulative displacement amount due to the movement operation” is calculated.

The displacement amount indicates a “position accompanying movement” as an accumulated value.
“The quotient obtained by dividing the amount of speckle pattern movement between each adjacent image-captured frame by the time between frames is the speed at each point during movement”.
By this method, “both speed and displacement” can be detected as movement information.

Next, a procedure for calculating “a cross-correlation function as an evaluation value” from the acquired image pattern and calculating its peak position (also referred to as “correlation peak position”) will be described.
The correlation peak position can be converted into a displacement amount using the imaging magnification and the pixel pitch of the image sensor IS.
That is, as described above, in the process of calculating the displacement amount from two image patterns acquired sequentially in time, the two image patterns are updated as needed to accumulate the displacement amount, and the “cumulative displacement amount” of the moving object Can be calculated.

Hereinafter, the embodiment of the invention will be described by taking the case of FIG. 3 as an example.
FIG. 4 shows a flowchart of the “procedure for calculating the peak position”.
In “image pattern acquisition 201”, two temporally continuous image patterns (speckle patterns for two frames) at a certain time are acquired.

In the “evaluation value calculation 202” using the two acquired image patterns, a part of the calculation of the “evaluation value” which is a cross-correlation function is performed.
As a method of calculating a cross correlation function, a method of calculating using a “discrete Fourier transform” according to the following equation (2) is conventionally known.

f * g ^* = F ⁻¹ [F [f] · F [g] ^* ] (2)
Here, f and g are two image patterns acquired successively.
In Equation (2), “*” represents “complex conjugate”, F [] represents “discrete Fourier transform operation”, and F ⁻¹ [] represents “inverse discrete Fourier transform”.

  “★” represents “cross-correlation calculation”.

That is, “f * g ^* ” on the left side of the equation (2) is a “cross-correlation function” or “evaluation value” calculated by a cross-correlation operation between the image pattern f and “complex conjugate of the image pattern g”.

This cross-correlation operation is performed on the product of the discrete Fourier transform “F [f]” of the image pattern f and the complex conjugate “F [f] ^* ” of the discrete Fourier transform “F [f]” of the image pattern g. It is obtained by performing an inverse discrete Fourier transform operation.

  The “cross-correlation function” calculated by the “calculation using discrete Fourier transform / inverse discrete Fourier transform” in the equation (2) described above is not subjected to “background removal processing”.

In the present invention, the calculation step of the cross correlation function includes a calculation step of “background removal processing”.
That is, a step of executing a discrete Fourier transform and a calculation step of a background removal process are performed on “two image patterns acquired in succession”, and an inverse Fourier transform is performed on the result to obtain a “reciprocal” A “correlation function” is calculated.

  Therefore, in order to distinguish the “cross-correlation function” calculated in the present invention from the “cross-correlation function” obtained without performing the background removal process, the “cross correlation function obtained by performing the background removal process” will be described below. The “correlation function” will be referred to as a “modified cross-correlation function”.

  In the present invention, in order to remove the influence of the background portion superimposed on the peak portion, the “background removal processing in the frequency space” is added and the “deformed cross-correlation function” is calculated.

  This calculation will be described later.

  In “peak position calculation 203” in FIG. 4, the correlation peak position is calculated.

  In the movement information detecting apparatus of the present invention, unlike the prior art, there is no “addition of arithmetic processing in real space before and after evaluation value calculation”. Therefore, the “calculation processing time” hardly increases.

Hereinafter, each procedure will be described.
FIG. 5 shows two image patterns acquired successively in time at a certain point in the step of “image pattern acquisition 201” in FIG.
5A shows the image pattern P1 at time t1 (before movement), and FIG. 5B shows the image pattern P2 at time t2 subsequent to time t1 (after movement).
In the example in the description, the time difference between time: t1 and t2 is “time between adjacent frames (1 second / 504): 1.98 ms” according to the frame rate (504 fps).
A “deformed cross-correlation function” is calculated using the two image patterns P1 and P2.

FIG. 6 is a flowchart showing the procedure for calculating the “modified cross-correlation function”.
Image patterns P1 and P2 are acquired in real space. Then, a Fourier pattern F1 is calculated by discrete Fourier transform of the image pattern P1.
Similarly, a Fourier pattern F2 is calculated by a discrete Fourier transform of the image pattern P2.

A complex conjugate F2 ^* of the Fourier pattern F2 is calculated, and a product of the complex conjugate F2 ^* and the Fourier pattern F1 is executed to obtain a synthesized Fourier pattern F3.
The synthesized Fourier pattern F3 is subjected to “background removal processing in frequency space (background processing in frequency space)” to calculate a BG removal Fourier pattern F4.
As described above, the “deformed cross-correlation function” is obtained by performing the “inverse discrete Fourier transform” on the “BG-removed Fourier pattern F4 subjected to background removal processing in the frequency space”.

  In the example of FIG. 6, product operation of “low frequency removal pattern LC” is performed as “background removal processing in frequency space”. This will be described below.

  FIG. 7 is a diagram illustrating an example of the low frequency removal pattern LC.

  As shown in FIG. 7, the low frequency removal pattern LC is 128 × 32 pixels, which is a simple pattern in which “3 × 3 pixels near the center pixel” is 0 and other pixels are 1.

The BG removal Fourier pattern F4 is obtained by calculating the “product for each pixel” between the low frequency removal pattern LC and the synthesized Fourier pattern F3.
A “deformed cross-correlation function” is obtained by performing inverse discrete Fourier transform on the BG-removed Fourier pattern F4 thus obtained.
This “deformed cross-correlation function” is referred to as a correlation pattern PS. The “deformed cross-correlation function” referred to here is the “correlation cross-function” referred to in the present invention.

The correlation pattern PS obtained in this way is shown in FIG.
As apparent from comparing this correlation pattern PS with the “cross-correlation function without background removal processing” shown in FIG. 8B, the “background portion size” is effectively reduced in the correlation pattern PS, The peak part stands out.

  When the peak position is calculated in this state, the influence of the background portion (background portion) is removed, the estimation error of the peak position can be removed, and the peak position can be specified more accurately.

The “region where the vicinity of the center pixel is 0” in the low-frequency removal pattern LC can be obtained and set in advance so that “the background portion can be removed satisfactorily” by experiments or the like.
The “background removal process” is performed in the frequency space.
As shown in FIG. 8B, the “background portion” has a “gradual mountain” shape. In the frequency space obtained by discrete Fourier transform, the low-frequency portion is the origin position in the frequency space (the above-mentioned “128 × 32 pixels”). In the vicinity of the central pixel at “.
Therefore, it can be effectively removed by “setting the data in the vicinity of the center pixel to 0”.

The “deformed cross-correlation function” can be expressed as the following equation (3) by modifying the above equation (2).
f * g ^* = F− ¹ [B [F [f] · F [g] ^* ]] (3)
Here, “B []” is a calculation of the background removal processing in the frequency space, and if the “low frequency removal pattern LC” described above is used,
f * g ^* = F ⁻¹ [LC · [F [f] · F [g] ^* ]] (3A)
It can be expressed as.

In the above procedure for obtaining a deformed cross-correlation function, “calculation process of product of low-frequency removal pattern LC in frequency space” is added once to the calculation calculation of the cross-correlation function according to the conventional equation (1). Only. That is, the number of operations added is only once.
Since the specific calculation process is only the process of “replace the vicinity of the center pixel of the synthetic Fourier pattern F3 with 0”, the calculation processing time is hardly increased.

  The case where the difference of the background removal pattern FB is performed will be described as “another example of calculating the deformed cross-correlation function (background removal processing in the frequency space)”.

The “background removal pattern FB” can be obtained in advance by an experiment under the condition that the peak portion of the cross-correlation function does not occur, that is, “there is no correlation between the two image patterns”.
For example, in FIG. 1, a portion excluding the peak portion PCP can be used as a background portion, and this can be subjected to a discrete Fourier transform operation to obtain a “background removal pattern FB”.

  When the background removal pattern FB is applied to the image patterns P1 and P2, the difference between the “synthetic Fourier pattern F3” and the background removal pattern FB is calculated for each pixel by calculating “BG”. A removed Fourier pattern F4 "can be obtained.

  By applying inverse discrete Fourier transform to the “BG removal Fourier pattern F4”, a correlation pattern PS subjected to background removal processing can be obtained.

  Even in this case, compared to the case of “calculation calculation of cross-correlation function” according to the conventional equation (1), the number of calculation processes added is one time of “difference in background removal pattern FB in frequency space”. Only.

The specific calculation processing is also “difference between the two patterns F4 and FB”, and “the calculation processing time is slightly longer than in the case of using the low frequency removal pattern LC”, but the increase in the processing time is slight. .
In addition, since the background removal pattern FB can be generated under the actual detection conditions, the peak position estimation error can be reliably eliminated.
In the case of using the conventional equation (1), in order to remove the background, it is necessary to add “multiple arithmetic processes in real space” before and after the calculation of the cross-correlation function.
In the present invention, it is only necessary to “add the arithmetic processing in the frequency space once”, and the increase in the arithmetic processing time can be effectively suppressed.

  Another embodiment will be described.

  In this embodiment, “a plurality of low frequency removal patterns LC and background removal patterns FB are prepared in advance” instead of “specifying and setting one low frequency removal pattern LC and background removal pattern FB in advance”. .

  Then, an optimum one is selected from “a plurality of prepared low-frequency removal patterns FC and background removal patterns FB” according to the acquired image patterns P1 and P2.

  A case where a plurality of low-frequency removal patterns are selected and used will be described.

  A plurality of low frequency removal patterns LC1, LC2,..., LCn,... Are stored in advance in the “speed / displacement calculating means CON (see FIG. 3)”.

As shown in FIG. 9, two image patterns (referred to as Q1 and Q2) are acquired (image pattern acquisition 301) as “preliminary operation before detecting speed and displacement”.
.., LCn,... Are sequentially applied to the acquired image patterns Q1 and Q2 to calculate correlation patterns PS1, PS2,..., PSn,. To do.
Of the correlation patterns PSn calculated in this way, the “correlation pattern PS (best) from which the background portion is well removed” is selected.
Then, the low frequency removal pattern LC (best) used for the calculation of the correlation pattern PS (best) is selected and used when “actually detecting speed and displacement”.

This preliminary operation is performed whenever the speckle size change is considered, such as when the moving member is changed, or when the speed, the operating condition of the displacement detection device CON, etc. are changed. It is preferable to update the frequency removal pattern.
In this way, the background removal process suitable for the actual detection conditions can be performed, and the “peak position estimation error” can be eliminated.

  Actually, an example is shown in which correlation patterns PS1, PS2, and PS3 are calculated for image patterns Q1 and Q2 using three types of low frequency removal patterns LC1, LC2, and LC3.

  As the three types of low frequency removal patterns LC1, LC2, and LC3, those shown in FIGS. 10A to 10C were used.

In the low frequency removal pattern LC1 shown in (a), “only 3 × 3 pixels near the center are 0, others are 1”.
In the low frequency removal pattern LC2 shown in (b), “only the 7 × 7 pixels in the vicinity of the center are 0, and the others are 1”.
In the low frequency removal pattern LC3 shown in (c), “only 19 × 19 pixels near the center are 0, and others are 1”.

  The image patterns Q1 and Q2 are as shown in FIGS. 11 (a) and 11 (b).

  FIGS. 12A to 12C show correlation patterns PS1, PS2, and PS3 calculated and calculated for the image patterns Q1 and Q2 using the low-frequency removal patterns LC1, LC2, and LC3.

Comparison of the three calculated correlation patterns PS1, PS2, and PS3 is as follows.
When the correlation patterns PS1 and PS2 are compared, the “peak part” in these is substantially the same, but in the correlation pattern PS2, the background part is better removed.

  When the correlation patterns PS2 and PS3 are compared, in these, “the background portion is removed as well in a similar manner”, but in the correlation pattern PS2, the peak portion is larger. This is because as the number of 0 pixels near the center of the low-frequency removal pattern increases, the waveform becomes distorted.

  From this result, it is possible to “select a correlation pattern PS2 in which the background portion is well removed and the peak portion is large (that is, the contrast is high) and select the low frequency removal pattern LC2 that gives this correlation pattern PS2”. .

  In this method, since the low frequency removal pattern can be selected in accordance with the speckle size, the estimation error of the peak position can be eliminated without being influenced by “background effect” according to the detection condition.

  As an example of the low-frequency removal pattern, the above example shows “only the square pixel in the vicinity of the center is set to 0”. However, the present invention is not limited to this and may not be a “square pixel”.

For example, in the example shown in FIG. 13A, “7 × 3 pixels near the center is 0”.
In the example described above, “binarization is performed by setting the vicinity of the center as 0 and the others as 1.” However, the present invention is not limited to this, and a “real number between 0 and 1 is used. It can also be changed gently from 0 to 1.
For example, in the example shown in FIG. 13B, “3 × 3 pixels in the vicinity of the center are set to 0 tp, 5 × 5 pixels in the vicinity thereof are set to 0.5, and the others are set to 1”.

  As another embodiment, “performing background removal processing in a frequency space” is effective for so-called sub-pixel processing for calculating the correlation peak position PP on the correlation pattern PS with a resolution equal to or less than the pixel unit of the image sensor. Explain that it works.

With respect to the correlation pattern PS of FIG. 8A described above, a peak position in the pixel unit is obtained, and a one-dimensional distribution in a cross section in the longitudinal direction of the image sensor (moving direction of the moving member) at the peak position is obtained. “Sub-pixel processing” is applied to the distribution.
As the “subpixel processing”, a known “conformal function fitting (Non-Patent Document 1)” was used.

FIG. 14A shows the “one-dimensional distribution of cross section” obtained from FIG.
The correlation peak position PP in “below the pixel unit” obtained by performing the sub-pixel processing on this one-dimensional distribution is 22.081 pixels.

Similarly, FIG. 14B shows a one-dimensional distribution of a cross section obtained from the correlation pattern PS of FIG. 8B without performing background removal processing in the frequency space.
The correlation peak position PP obtained by subjecting this one-dimensional distribution to the sub-pixel processing is 21.947 pixels.
In these examples, the correlation peak position PP is larger than 22 pixels when the background removal process is performed, and smaller than 22 pixels when the background removal process is not performed.

  Although the error between the two is a small value of 0.134 pixels, this is a displacement amount for two adjacent image patterns. When actually obtaining the cumulative displacement amount with respect to the movement of the moving member, Small errors accumulate and become “a large value as the cumulative displacement amount”.

Therefore, “removal of background portion” in the present invention is a necessary means for high-accuracy measurement.
For example, in the case of “moving 50 mm at a constant speed of 100 mm / second”, it takes 0.5 seconds to move 50 mm. If the frame rate is 504 fps, “252 image patterns” are captured and acquired during this movement time.
That is, if the accumulation is simply performed 251 times, the accumulation error is “0.134 × 251 = 33.634 pixels”, and “an error of about 249 μm” occurs with respect to the accumulated displacement.

  Such movement at a constant speed corresponds to, for example, movement of an intermediate transfer belt or a conveyance belt in a multicolor image forming apparatus described later.

Actually, under the conditions described with reference to FIG. 3, the sheet S is accumulated in “movement distance: operation of 50 mm after decelerating and stopping after reaching the maximum speed of 100 mm / second after being accelerated from the stopped state”. The results of obtaining the displacement amount are shown below.
Such “intermittent movement” corresponds to, for example, movement of a recording sheet in an image forming apparatus using an ink discharge head described later.

FIG. 15A shows a graph of the correlation peak position PP obtained by performing the sub-pixel process on two adjacent image patterns when the background removal process is performed.
Further, FIG. 15B shows a graph of the accumulated displacement amount obtained by converting the correlation peak position into the displacement amount and sequentially accumulating the image patterns.

As a comparative example, FIGS. 16A and 16B show “results similar to the above” when the background removal process is not performed.
The cumulative displacement amount when the background removal process is performed is 50.020 mm, and the cumulative displacement amount when it is not performed is 49.736 mm.
The error with respect to the moving distance of 50 mm is 20 μm for the former and 264 μm for the latter. The error in the former is 0.8% of the error in the latter, and the effect of 10 times or more is seen for highly accurate detection. .

  Hereinafter, embodiments of the multicolor image forming apparatus will be described.

FIG. 17 is a diagram for explaining one embodiment of a “tandem multicolor image forming apparatus”.
Reference numerals 1 </ b> Y, 1 </ b> M, 1 </ b> C, and 1 </ b> K in the figure indicate image carriers arranged side by side along the intermediate transfer belt 105. These image carriers are “drum-shaped photoreceptors”.
In FIG. 17 and the subsequent drawings, “Y, M, C, K” in the reference numerals correspond to “yellow, magenta, cyan, black”, respectively.

  The photoreceptors 1Y, 1M, 1C, and 1K are driven to rotate in the direction of the arrows in the figure.

Around the photoreceptors 1Y, 1M, 1C, and 1K, chargers 2Y, 2M, 2C, and 2K as charging means (in the drawing, a contact type using a charging roller is shown. , Non-contact type corona charger etc. can also be used.) Developers 4Y, 4M, 4C, 4K as developing means, primary transfer means (transfer charger, transfer roller, transfer brush, etc.) 6Y, 6M, 6C , 6K, and photoconductor cleaning means 5Y, 5M, 5C, 5K, and the like.
In the figure, reference numeral 30 denotes a fixing unit, 40 denotes a secondary transfer unit, and 41 denotes a conveying unit.

The photoreceptors 1Y, 1M, 1C, and 1K are uniformly charged by the chargers 2Y, 2M, 2C, and 2K, and then light beams (intensity modulated in accordance with image information) by the optical scanning device 20 that is a latent image forming unit. For example, an electrostatic latent image is formed by exposure with a laser beam.
The basic configuration of the optical scanning device 20 that performs exposure will be described later.

  The electrostatic latent images formed on the photosensitive drums 1Y, 1M, 1C, and 1K are developed by the developing devices 4Y, 4M, 4C, and 4K, respectively, and are visualized as toner images of yellow, magenta, cyan, and black. The

The multicolor image forming apparatus shown in FIG. 17 is an “intermediate transfer belt type”, and each color toner image visualized on the photoreceptors 1Y, 1M, 1C, and 1K is sequentially superimposed on the intermediate transfer belt 105. Primary transfer.
The toner images of the respective colors superimposed on the intermediate transfer belt 105 are fed to a sheet-like recording medium such as paper fed from a paper feeding unit (not shown) and conveyed to the position of the secondary transfer unit 40 via a conveyance unit (not shown). Secondary transfer is performed at once.
The sheet-like recording medium to which the toner image is transferred is conveyed to the fixing unit 30 by a conveying unit 41 such as a conveying belt, and the toner image is fixed by the fixing unit 30.
In this way, a multicolor image or a full color image is obtained. The sheet-like recording medium after fixing is discharged to a paper discharge unit, a post-processing device, or the like (not shown).

The photoreceptors 1Y, 1M, 1C, and 1K after the toner image transfer are cleaned by cleaning members (blades, brushes, and the like) of the cleaning units 5Y, 5M, 5C, and 5K, and residual toner and paper dust are removed.
Further, the intermediate transfer belt 105 after the toner image transfer is also cleaned by a belt cleaning unit (not shown) to remove residual toner.

  In the multicolor image forming apparatus shown in FIG. 17, “a single color mode for forming an image of any one color of yellow, magenta, cyan, or black” or “an image of any two colors of yellow, magenta, cyan, or black is overlaid. Two-color mode to be formed ”,“ three-color mode to form an image of three colors of yellow, magenta, cyan, and black ”and“ full-color mode to form a full-color image by superimposing the above four colors ” By specifying and executing these modes in an operation unit (not shown), it is possible to form a single color, multicolor, or full color image.

  The multi-color image forming apparatus of FIG. 17 uses an intermediate transfer belt 105 to primarily transfer images from the photoreceptors 1Y, 1M, 1C, and 1K to the intermediate transfer belt 105 to form superimposed images of the respective colors, and then the intermediate transfer belt. This is an “intermediate transfer system” in which secondary transfer is performed collectively from 105 to a sheet-like recording medium such as paper.

  In addition to this, as in the multi-color image forming apparatus having the configuration shown in FIG. 18, instead of the intermediate transfer belt, a conveyance belt 106 that carries and conveys a sheet-like recording medium such as paper is used. A direct transfer system that directly transfers from 1C and 1K to a sheet-like recording medium may be used.

  In the multi-color image forming apparatus of the direct transfer method, as shown in FIG. 18, the approach path of the sheet-like recording medium is different from that of the intermediate transfer method shown in FIG. The photosensitive members 1Y, 1M, 1C, and 1K are conveyed.

  In the multicolor image forming apparatus of FIG. 21 as well, the photoreceptors 1Y, 1M, 1C, and 1K are uniformly charged by the chargers 2Y, 2M, 2C, and 2K.

Thereafter, exposure is performed with a light beam (for example, laser light) whose intensity is modulated in accordance with image information by the optical scanning device 20 serving as a latent image forming unit, and an electrostatic latent image is formed.
The electrostatic latent images formed on the photoreceptors 1Y, 1M, 1C, and 1K are respectively developed by the developing devices 114Y, 114M, 114C, and 114K, and are visualized as toner images of yellow, magenta, cyan, and black. The
A sheet-like recording medium is fed from a paper feed unit (not shown) in synchronization with the development process, and is conveyed by the conveyance belt 106 and carried on the conveyance belt 106 via a conveyance unit (not shown).
The sheet-like recording medium carried on the conveyance belt 106 is conveyed toward the photoreceptors 1Y, 1M, 1C, and 1K, and the color toner images visualized as described above are transferred to the transfer units 6Y, 6M, and 6C. , 6K and sequentially transferred onto a sheet-like recording medium.
The “4-color superimposed toner image” transferred to the sheet-like recording medium is conveyed to the fixing unit 30 and fixed on the sheet-like recording medium by the fixing unit 30 to obtain a multicolor or full-color image.

  The sheet-like recording medium after fixing is discharged to a paper discharge unit (not shown) or a post-processing device.

  The photoreceptors 1Y, 1M, 1C, and 1K after the toner image transfer are cleaned by cleaning members (blades, brushes, and the like) of the cleaning units 5Y, 5M, 5C, and 5K to remove residual toner, paper dust, and the like.

In the multicolor image forming apparatus configured as shown in FIGS. 17 and 18, the movement of the belt surface of the intermediate transfer belt 105 is performed in the intermediate transfer system, and the belt surface of the transport belt 106 is moved in the direct transfer system. If not performed with high accuracy, color misregistration will occur.
In order to drive the belt with high accuracy, a method of making all the components with high accuracy can be considered, but there are many components and it is practically difficult to realize from the viewpoint of cost.

  Therefore, it is preferable to provide a “detection means for detecting speed fluctuations and displacements” of the intermediate transfer belt 105 and the conveyance belt 106 and feed back the detection result to the belt drive motor.

  For this purpose, a movement information detection device that detects the speed fluctuation and displacement amount of these belts as movement information becomes important.

  In order to detect the speed fluctuation of the belt, it was previously detected by forming a mark directly on the belt. However, it is difficult to form a mark by processing directly on the belt, and processing takes time. Mass production was poor, which was a major factor in increasing costs.

  In the present invention, the speed fluctuation of a moving member such as a belt is easily detected without directly processing the intermediate transfer belt or the conveyance belt.

  The “movement information detection device” described above can be used to detect speed fluctuations in the circumference of the belt of the intermediate transfer belt or the conveyance belt used in the multicolor image forming apparatus. Belt speed fluctuation can be detected.

  In the belt speed correcting means for correcting the speed of the intermediate transfer belt or the conveying belt, the speed fluctuation information can be fed back to, for example, a belt drive motor, so that the speed fluctuation can be controlled to be substantially zero. It is possible to realize a high-quality color image in which “expansion / shrinkage and color misalignment” are suppressed to be small.

  Further, with respect to the belt-like member and the roller-like member used as the fixing device, it is also possible to detect and correct the speed fluctuation using the movement information detecting device.

Further, the detection result of the speed fluctuation of the intermediate transfer belt or the conveyance belt is fed back to a writing start position correcting means (for example, a liquid crystal deflecting element provided in the optical scanning apparatus) for correcting the writing start position by the optical scanning apparatus 20. It is also possible.
The liquid crystal deflecting element can shift the position of light reaching the photoconductor in a direction parallel to the rotation direction of the photoconductor by a voltage applied to the liquid crystal.
When the belt speed fluctuation occurs, the superposition of each color image shifts, or each color image itself expands or contracts. By using a liquid crystal deflecting element, each color toner image is corrected so as to correct the belt speed fluctuation. Since the formation position and the expansion / contraction of the image can be corrected, it is possible to obtain a high-quality output image with no color shift or image expansion / contraction as a result.

Next, an optical scanning device using a liquid crystal deflection element will be described.
FIG. 19 is a schematic configuration diagram illustrating an example of an optical scanning device used in a multicolor image forming apparatus.

  The image forming apparatus in FIG. 19 has four drum-shaped photoreceptors 301, 302, 303, and 304 (corresponding to the photoreceptors 1Y, 1M, 1C, and 1K in FIGS. 17 and 18) and a transfer belt 400 (see FIG. 19). 17 is a multi-color image forming apparatus that forms a color image by sequentially transferring toner images of different colors.

The optical scanning device is integrally formed, and all light beams are scanned by a single optical deflector (polygon mirror in this example) 1050.
The polygon mirror 1050 has six deflection reflection surfaces and a two-stage structure.

  That is, the optical scanning device includes a light source unit 1001, a single polygon mirror 1050 that deflects and scans the light beam from the light source unit 1001, and a scanning beam scanned by the polygon mirror 1050 on the photosensitive members 301, 302, 303, and 304. Scanning lenses 1061, 1062, 1063, and 1064 that form an image on the surface to be scanned.

  Scanning is performed for two stations in a direction facing the polygon mirror 1050.

  For simplification of explanation, only one station is illustrated for the optical system after the light source unit and the scanning lens.

A light source (for example, a semiconductor laser (LD), an LD array, etc.), a coupling lens, and an aperture (not shown) are mounted on the light source unit 1001.
A light beam emitted from a light source (not shown) of the light source unit 1001 is converted into a substantially parallel light beam, a substantially divergent light beam, or a substantially convergent light beam by a coupling lens (not shown).

Thereafter, the light beam is cut out to a desired light beam width by an aperture, and is once condensed in the sub-scanning direction in the vicinity of the polygon mirror 1050 by a line image forming lens (for example, a cylindrical lens) 1041, and the scanning optical system comprising the scanning lens L1: 1061. A beam spot is formed on the image plane (scanned surface) 2001 by the system.
As described above, in the normal optical scanning device, in order to reduce the deterioration of the optical characteristics due to the surface tilt between the mirrors of the polygon mirror 1050, the “surface tilt correction optical system” that collects light once in the sub-scanning direction in the vicinity of the polygon mirror. It has been adopted.
The scanning lens is made of resin, and the diffraction grating may be formed on one or a plurality of optical surfaces.

  Usually, folding mirrors 1111 and 1121 are inserted between the light polarization means and the image plane, and the optical path is folded. Moreover, although the scanning optical system has been shown as an example constituted by a single scanning lens, two or more scanning lenses may be used.

Writing start position correction means (for example, a liquid crystal or a deflection element) for correcting the writing start position by the optical scanning device is preferably provided between the coupling lens of the light source unit 1001 and the cylindrical lens 1041.
The liquid crystal deflecting element is an element that deflects light by utilizing a change in refractive index with respect to light having a certain polarization direction when a voltage is applied. The liquid crystal deflecting element is not limited to liquid crystal, and other electro-optical materials such as LiNbO ₃ can be used. It can be realized even if used.

Hereinafter, a liquid crystal will be described as an example of the electro-optic material.
FIG. 20 shows a schematic diagram of the structure of the liquid crystal deflection element.

The structure of the liquid crystal deflection element is a structure in which a plurality of transparent electrodes are electrically connected via a resistance member. When a potential difference is applied to the terminals 1 and 2 in the figure, the potential at the transparent electrode changes substantially linearly from the terminal 1 toward the terminal 2.
When a voltage is applied to the liquid crystal, the refractive index of polarized light along the optical axis of the liquid crystal changes in proportion to the voltage. Therefore, the refractive index of the portion where the transparent electrode exists is “from terminal 1 to 2”. Change linearly.
This refractive index distribution is equivalent to that of a prism, and the light transmitted through the portion where the transparent electrode is present is deflected as shown in the figure.
The amount of light deflection can be changed by changing the potential difference between the terminals 1 and 2. By aligning the deflected direction with the rotation direction (sub-scanning direction) of the photosensitive member, the formation position of each color toner image and the expansion / contraction of the image can be corrected.

FIG. 21 shows an embodiment in which the movement information detection device is applied to the detection of the speed fluctuation of the intermediate transfer belt of the multicolor image forming apparatus.
FIG. 21 shows only the main part around the intermediate transfer belt 105 of the color image forming apparatus of FIG.

  21 is installed in the vicinity of the driving roller 105A of the intermediate transfer belt 105.

When the installation position is in the vicinity of the driving roller 105A as in the movement information detecting device A, the “belt surface is regulated by the driving roller 105” when the belt is driven, and “the direction in the direction orthogonal to the belt surface is "Battery" can be reduced.
The movement information detection device A is arranged at such a position, and the movement speed of the intermediate transfer belt 105 is detected as movement information as described above, and the drive of the transfer belt 105 is stabilized using the fluctuation data. be able to.

  As in the case of the movement information detection device B, if the belt is disposed so as to correspond to the “tensioned portion” of the intermediate transfer belt 105, it is easily affected by the above-described belt fluttering.

FIG. 22 shows an embodiment of a multicolor image forming apparatus that prints and records pixels by an ink discharge head. The figure shows only the main part of the “recording paper (sheet-like recording medium) conveyance system”.
As shown in FIG. 22A, the recording paper S loaded in the paper tray 251 is fed by the feed roller FR and sent to the printing unit by the print head 253 by the transport roller 252.
The recording sheet S is further sandwiched between the conveyance rollers 254 and intermittently conveyed in a “stable conveyance state” as shown in FIG.
The print head 253 prints pixels on the recording paper S while scanning in a direction orthogonal to the drawing. During this printing scan, “sheet feeding is stopped”, and when one scanning of printing is completed, an intermittent feeding operation of “sending a predetermined amount” is repeatedly performed by the transport rollers 252 and 254.

  The movement information detection device 255 is arranged at the position shown in the figure, and can detect a predetermined feed amount of intermittent feeding as movement information of the actual recording paper S itself, and can control the conveyance to realize high-precision paper feeding.

S sheet recording medium (moving member)
LTS light source
CL coupling lens
LN imaging lens
IS image sensor
PSN image pattern acquisition means
CON movement information calculation means

JP 2009-15240 A JP 2003-267591 A Japanese Patent Laid-Open No. 5-18713 JP-A-8-261717 JP-T-2007-520762

Digital image processing (CG-ARTS Association), p. 205)

Claims (9)

A movement information detection device for detecting the speed and displacement of a movement member as movement information,
A light source;
A coupling lens for guiding the beam emitted from the light source and irradiating the moving member;
An image sensor having a plurality of pixels and capable of acquiring a one-dimensional or two-dimensional image;
An imaging lens that forms an image of the scattered beam scattered by the moving member on the image sensor;
Image pattern acquisition means for acquiring the image pattern on the image sensor in time series;
Movement information calculation means for calculating a cross-correlation function between two image patterns P1 and P2 adjacent to each other in acquisition time and calculating a speed and displacement of the moving member;
The calculation of the cross-correlation function in the movement information calculation means includes a step of executing a discrete Fourier transform, a step of executing a background removal process in a frequency space, and a step of executing an inverse discrete Fourier transform. A moving information detecting apparatus characterized by the above. The movement information detection apparatus according to claim 1,
The calculation of the cross-correlation function in the movement information calculation means is
Performing a Fourier transform of the image pattern P1 to calculate a Fourier pattern F1;
Performing a discrete Fourier transform of the image pattern P2 to calculate a Fourier pattern F2, and
Performing a product of the Fourier pattern F1 and the complex conjugate of the Fourier pattern F2 to calculate a composite Fourier pattern F3;
Performing a background removal process in the frequency space on the synthesized Fourier pattern F3 to calculate a BG removal Fourier pattern F4;
And a step of calculating a correlation pattern PS by performing an inverse discrete Fourier transform on the BG-removed Fourier pattern F4. In the movement information detection device according to claim 2,
A movement information detecting apparatus that performs an operation of a product of a synthetic Fourier pattern F3 and a low frequency removal pattern LC as background removal processing. In the movement information detection device according to claim 2,
A movement information detection apparatus that performs a calculation of a difference between the synthesized Fourier pattern F3 and the background removal pattern FB as background removal processing. In the movement information detection device according to claim 3 or 4,
A plurality of patterns are prepared in advance at least one of the low-frequency removal pattern LC and the background removal pattern FB,
A movement information detecting apparatus that performs background removal processing by selecting one of the plurality of facilitated removal patterns according to at least one of image patterns P1 and P2. The movement information detection apparatus according to any one of claims 3 to 5,
The movement information detecting apparatus characterized in that the calculation of the speed and displacement of the moving member by the movement information calculating means includes a step of calculating a correlation peak position PP on the correlation pattern PS with a resolution equal to or less than a pixel unit of the image sensor. A plurality of image carriers, a charging device that uniformly charges the plurality of image carriers, and an optical scanning device that forms an electrostatic latent image by optically scanning the plurality of uniformly charged image carriers with a beam spot; The electrostatic scanning image formed on the plurality of image carriers by the optical scanning device is visualized with toner of a predetermined color, and can be moved to face the plurality of image carriers. An intermediate transfer belt, a first transfer device, a second transfer device, and a fixing device.
The color toner images visualized on the plurality of image carriers are transferred onto the intermediate transfer belt by the first transfer device, and the color toner images are superimposed on the intermediate transfer belt. The respective color toner images thus transferred are transferred onto a sheet-like recording medium by the second transfer means, and the respective color toner images transferred onto the sheet-like recording medium are fixed on the sheet-like recording medium by the fixing device. In a multicolor image forming apparatus for forming a color or color image,
7. A multi-color image forming apparatus comprising the movable information output device according to claim 1 and detecting a speed of the intermediate transfer belt. A plurality of image carriers, a charging device that uniformly charges the plurality of image carriers, and an optical scanning device that forms an electrostatic latent image by optically scanning the plurality of uniformly charged image carriers with a beam spot; The electrostatic scanning image formed on the plurality of image carriers by the optical scanning device is visualized with toner of a predetermined color, and can be moved to face the plurality of image carriers. A conveyance belt that conveys a sheet-like recording medium, a transfer device, and a fixing device,
Each color toner image visualized on the plurality of image carriers is directly transferred to the sheet-like recording medium conveyed by the conveying belt by the transfer device, and each color toner image is transferred onto the sheet-like recording medium. In a multicolor image forming apparatus that forms a multicolor or color image by superimposing and fixing each color toner image superimposed on the sheet-like recording medium on the sheet-like recording medium by the fixing unit.
A multi-color image forming apparatus, comprising the movement information detecting device according to claim 1, wherein the moving information detecting device detects a speed of the conveying belt. In a multicolor image forming apparatus that prints and records pixels on a sheet-like recording medium with an ink ejection head,
A multi-color image forming apparatus comprising the movement information detection apparatus according to claim 1, wherein the movement information detection apparatus detects a movement amount of the sheet-like recording medium.

Images