JPH01140857A - Color image reader - Google Patents

Abstract

PURPOSE:To prevent an outputted image from being adversely influenced by the dislocation of picture elements between line image sensors by arranging the picture elements of the line image sensors corresponding to the same position so that the picture element, to which a green component image is projected, comes to be a center. CONSTITUTION:An optical information exchange unit in an image reader is provided with a lens barrel 101a with a built-in image forming lens, an interference filter 102, a color separating prism 103 and CCDs 104-106. A color separated images by the prism 103 are projected on the CCDs 104-106. At that time, a green separated image is projected on the CCD 105 at the center.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a color image reading device, and more particularly to a color image reading device using a line image sensor in a color image forming device that forms color images by digital method. It is something.

[Background of the invention]

In recent years, the spread of color image equipment such as color CRT display devices for computers and videotex terminal devices, and the
2. Description of the Related Art As documents and other printed matter in offices and the like are increasingly produced in color, there is an increasing demand for color copies containing a larger amount of information.

Under the above circumstances, various methods and apparatuses for forming multicolor images have been proposed.

2. Description of the Related Art Conventional multicolor image forming apparatuses include those in which toner images of each color formed on an image forming body are sequentially transferred onto recording paper, and the toner images of each color are superimposed on the recording paper. However, such a multicolor image forming apparatus has problems such as the need for a transfer drum, which increases the size of the apparatus, and the transfer misalignment of each color toner image, making it impossible to obtain a clear multicolor image.

Therefore, for example, a plurality of latent image forming means and a plurality of developing means are arranged around a rotating drum-shaped photoreceptor, and by repeating latent image formation and development, visible images of different colors are formed on the drum-shaped photoreceptor. A method of forming layers in layers and transferring them all at once to recording paper (Japanese Unexamined Patent Publication No. 52-106743, No. 56-14)
No. 4452, No. 58-79261, No. 61-
170754) has been proposed.

The resist of the multicolor toner image formed on the image forming body by the above method is greatly improved compared to the above-mentioned color copying method using a transfer drum.

Further, one latent image forming means and a plurality of developing means are arranged around the rotating drum-shaped photoreceptor, and a latent image is formed and developed for one color each time the photoreceptor rotates. Multiple rotations form a multicolor visible image on the photoreceptor, which is then
A method of simultaneously transferring onto recording paper (Japanese Unexamined Patent Publications No. 60-76766, No. 60-95456, No. 61-17075)
4) has been proposed.

In this method, since there is only a single latent image forming means, the apparatus can be made more compact than in the former method, and in particular, since the latent image forming means is commonly used, it is advantageous in guaranteeing latent image registration. .

On the other hand, as a color imaging device that reads multicolor image information,
A method of switching filters and light sources is known.

Although this method is effective in preventing moiré because it basically does not cause color shift and can utilize a monochrome solid-state imaging device with high resolution, it has the disadvantage that color correction and undercolor removal cannot be performed.

There is also known an imaging means in which three color components of a document are input simultaneously. With this method, various color corrections can be performed without using a large capacity image memory.

In a color image reading device that separates the original image into colors and projects the color separated images onto each of three aligned line image sensors, it is difficult to achieve perfect alignment accuracy of the line image sensors. However, if the alignment accuracy is insufficient, the same pixels of each sensor will not correspond perfectly, causing pixel misalignment. Pixel deviation is for each B and G
, increases the error when performing masking based on R information.

There are many cases of misalignment of specific colors, but they can be roughly divided into: ■ When B pixels and R pixels are shifted to the left and right with respect to a G filter, ■ When B pixels and R pixels are shifted to the left and right with respect to a G74 filter. is shifted to the right, and (2) the B pixel and R pixel are shifted to the left with respect to the G filter. This deviation is not only in the main scanning direction but also in the sub-scanning direction.

If the above pixel deviation is uniformly corrected, that is, from B (blue), G (green), and R (red) signals to Y (yellow).

This causes color errors when converting into M (magenta) and C (cyan) signals. This error becomes particularly large when the image is a thin line and the thickness of the thin line is about the size of a pixel. Further, if UCR (undercolor removal) or the like is performed, the color error becomes even larger. In the case of a black line, this type of color error occurs in the form of a color surrounding the black line.

Pixel misalignment has long been recognized as a serious problem, but it is difficult to deal with it, and even if it were solved, it would only be applied to limited colors, as seen in Japanese Patent Application No. 61-15466. It was getting worse.

[Purpose of the invention]

Conventionally, line image sensors have been fixed using mechanical fixing or adhesives after aligning each line image sensor with a fixed lens system and color separation prism system using a jig. It was done to fix it using. Mechanical fixing methods tend to cause pixel misalignment due to vibrations and temperature cycles, so fixing using adhesives, which are stable in use, is considered the most effective fixing method. However, even with such an image sensor fixing method, pixel misalignment of about 0.0 to 0.5 pixels in both the main scanning and sub-scanning directions cannot be avoided.

The processing form of input image data in a color image copying apparatus is usually from B, G, R information to Y, M, C, BK.
(Black) Accompanied by a masking operation that converts it into information.

This masking operation is processed on the assumption that each color information is at the same position, and if there is a pixel shift, it will be processed incorrectly. This situation becomes more serious when the original has fine lines.

As a result of studying the influence of pixel misalignment on output images, the inventors investigated in which cases the degree of influence of pixel misalignment is large in B, G, and R color separation information, and based on the results. We also considered the installation method of the line image sensor so that we could ultimately obtain pixel misalignment that would have little effect.

The purpose of the present invention is to create a configuration in which the output image is not adversely affected by pixel misalignment between line image sensors, and to create a form that does not have the effect of pixel misalignment that is easy to correct by setting the pixel misalignment in a specific shape. The purpose is to

[Structure of the invention]

The above purpose is to provide a color image reading device in which a color-separated image is projected onto each of three aligned line image sensors, in which pixels of the line image sensor corresponding to the same position project a green component image. This is achieved by a color image reading device characterized in that pixels are arranged at the center.

[Effect]

In the present invention, the pixels of the B, G, and R image sensors corresponding to the same position are arranged so that the pixel corresponding to the green component image is centered.

Pixel misalignment means that the reading position is misaligned.

In the recording device, the image is recorded with no pixel shift. For example, in a laser optical system, image data corresponding to input pixels is sequentially output in synchronization with position detection of a polygon surface by a laser beam index sensor. The writing position of this output data corresponds to the number of pixels from the end of the line image sensor, and it should be noted that it does not correspond to the spatial position of the image sensor.

Furthermore, although the pixel deviation is a fraction of a pixel, the pixel deviation is not uniform due to distortion of the image sensor and imperfect parallelism between the image sensors. There is also deformation when it is fixed. This makes it more difficult to take measures against pixel misalignment.

Based on this, we investigated which type of pixel shift would be the least noticeable, and found that it would be less noticeable if the pixel shift was shifted so that the green component image was centered.

Figure 9 shows the number of pixels from the center (image height) and the pixels that deviate from the green component image when the line image sensor is tilted and the blue and red component pixel images are on the left and right with the green component image at the center. This shows the relationship with numbers.

The reason why it is least noticeable when the green component image is shifted to the center is thought to be that the human eye is sensitive to misalignment of the M color among Y, M, and toners. .Furthermore, Fig. 10 shows a state in which three to four colors of toner are overlapped, and the black information created in the masking operation (BK) toner (TBK) overlaps at the same position as the magenta color (TM). The pixel shift is hidden and becomes less noticeable.

〔Example〕

DESCRIPTION OF THE PREFERRED EMBODIMENTS The basic structure of an image forming apparatus equipped with a color image reading device of the present invention will be explained below with reference to FIGS. 1 and 2.

FIG. 1 is a diagram showing a general configuration of a color image forming apparatus and a control system.

FIG. 2 is a block diagram schematically showing an image forming process using the apparatus.

FIG. 1 shows a color image forming apparatus having the following configuration as a representative example. First, for color originals, an optical image is separated into 3' colors of blue, green, and red by a color separation means using a color separation prism 103 inserted in the optical path, and is imaged on three CCD image sensors, etc. After being converted, it is converted into a digital signal by an A/D converter.

Masking and UCR are performed on these color separation signals, and 4
Obtain color image data Y, M, C, and BK.

One of these image data is selected and written on the image forming body through a writing device using a semiconductor laser to form an electrostatic image, which is developed by a developing device corresponding to the selected color signal. A color toner image is formed. By performing the above process a plurality of times while changing the selected color signal, a multicolor toner image in which toner images of each color are superimposed is formed on the image forming body. This multicolor toner image is transferred and fixed onto a transfer material to obtain a color image.

In FIG. 1, the CPU 1 is controlled by the operation section circuit (304).
By pressing the copy button on the operation panel connected to the main body control CPU 2, the original reading section A is controlled by the optical drive CPU 2 connected to the main body control CPU 2 via serial communication.
is driven.

First, a document 82 on a document table 81 is optically scanned by an optical system. In this optical system, a carrier 7 is provided with a halogen lamp 85, a slit 86, and a reflecting mirror 87.
4. (2) It has a movable mirror unit 88 provided with mirrors 89 and 89', and the carriage 84 and the movable mirror unit 88 are stretched by a pulley 91 attached to a stepping motor 90 via pulleys 93 and 92. The wires are run on the slide rail 83 at speeds of V and 1/2V, respectively.

In this way, optical information obtained by illuminating the original 82 with the halogen lamp 85 is led to the optical information conversion unit 100 via the reflecting mirror 87 and the mirrors 89 and 89'.

Note that the halogen lamp 85 has undergone infrared cut treatment. A fluorescent lamp can also be used instead of the halogen lamp 85.

In addition, standard white plates 9 are installed on the back side of both ends of the platen glass 81.
7 and 98 are provided. This is for optically scanning the standard white plate to normalize image signals, which will be described later.

Next, the optical information conversion unit 100 of the image reading section A
is lens 101. An infrared (IR) cut filter 102 made of a multilayer film, and a color separation prism 10 made of a multilayer film.
3. C CD 104, which receives a blue component image, which is a line image sensor, and C CD 10, which receives a green component image.
5. It is composed of a CCD 106 that receives a red component image and a substrate 107 that amplifies and A/D converts the output from the CCD, and each CCD and the substrate 107 are connected by flexible wiring. In the optical information conversion unit lOO having such a configuration, the optical information derived from the optical system is focused by the lens 101, and is separated into blue, green, and red by a color separation means provided with a blue, green, and red color separation prism 103. The optical information is color separated into the CCD 103.104.
An image is formed on a light receiving surface 105, and an image signal is output. Note that a correction glass may be inserted to correct the lens 101 and each CCD focal position for each color.

The image signals output from the CCDs 104, 105, and 106 are subjected to calculations such as A/D conversion and shading correction, which will be described later, and then converted into yellow, magenta, cyan, and black image data by masking and UCR. .

The aforementioned Y, M, C, and BK image data are sequentially selected by a selector and then processed by a signal processing device such as binarization, multi-value conversion, dithering, etc., and outputted to the writing section B.

In the writing section B, the laser beam modulated by the selected color image data is first applied to the motor 1 at power supply 08.
The image forming body 120 is scanned by a polygon mirror 112 which is being rotated by a polygon mirror 10 .

A laser beam index sensor (198) for detecting the start of beam scanning, which will be described later, when the scanning is started.
The image data of the first selected color detected by
For example, modulation of the beam using cyan data) is started. The modulated beam scans over imager 120.

The image forming body 120 is uniformly charged in advance by a charger 121.

Main scanning is performed by the laser beam, and the image forming body 12
Sub-scanning is performed by the rotation of 0, and an electrostatic image corresponding to the first color image data is formed on the image forming body 120.

This electrostatic image is developed by applying a bias voltage from the high voltage power supply 2 (241a) to the developing device 123 containing cyan toner, and a cyan toner image is formed on the image forming body 120. Note that toner replenishment of the developing device 123 is performed at any time by C.
Toner replenishment SD 2 (313) controlled by PUI
will be replenished via. This cyan toner image is formed on the image forming body 1 with the cleaning blade 127 released.
The image forming member 120 moves as the image forming member 20 rotates, and is uniformly charged again by the charger 121 while holding the cyan toner image. Thereafter, an electrostatic image is formed based on the image data of the second color (for example, magenta) in the same manner as in the case of the first color, and the developing device 124 containing magenta toner is formed.
Similarly, by applying a bias voltage from the high voltage power supply 2 (241b), development is performed to form a magenta toner image on the cyan toner image. Note that the electrostatic image must be aligned with the cyan toner image with sufficient accuracy. Similarly, an electrostatic image is formed based on the third color image data (yellow), and is developed in the same manner as the previous time by applying a bias voltage from the high voltage power supply 2 (241c) to the developing device 125 containing yellow toner. A yellow toner image is formed on the cyan toner image and the magenta toner image.

Next, after being charged again, an electrostatic image is formed based on the image data (black) of the cylinder 4, and by applying a bias voltage from the high voltage power supply 2 (241d) to the developing device 122 containing black toner, the black toner image is placed first. is formed on the toner image.

A multicolor (four-color) toner image is obtained in the manner described above.

Although four-color image formation is explained here, Y
, M, and C color image formation is the same except that UCR is not performed.

The developing units 123, 124, 125, and 122 fly toner toward the image forming body 120 in a non-contact manner under the application of AC and DC bias voltages from the high voltage power supplies 2 (241a, b, c, d). It is preferable that the image be developed and that it be developed in reverse. Note that the toner replenishment to the developing devices 124, 125, and 122 is done by following the signal from the CPUI as in the case of the developing device 123.
4 (315) and SD I (312).

The thus obtained multicolor toner image is applied to the transfer material P, which is fed from the paper feeder 141 via the feed roller 142 and the timing roller 143 in synchronization with the rotation of the image forming body 120, to the high voltage power source 3 (324). Transferred by the transfer pole 130 to which high pressure is applied from the separation pole 131
It is separated from the image forming body 120 by.

The transfer agent P separated by the separation electrode 131 is transferred to the microcomputer C.
The paper is transported to a fixing device 132 whose temperature is controlled to a desired temperature by the PUI, where it is fixed, and then ejected to obtain a color image.

The image forming drum 120 after the transfer is cleaned by a cleaning device 126 and prepared for the next image formation.

In the cleaning device 126, a blade 127
In order to facilitate recovery of the toner cleaned by the blade 127, a metal roll 128 to which a DC voltage is applied from a high voltage power source 4 (325) is arranged without contacting the image forming body. Further, the blade 127 is released from the pressure bond when cleaning is completed, but a cleaning auxiliary roller 129 made of sponge is further provided to remove toner left behind on the image forming body 120 when the blade 127 is released. By rotating and pressing in the opposite direction, the image forming body 120 is sufficiently cleaned.

The above is the basic configuration of the color image forming apparatus of the present invention and when forming a color image using the apparatus, and the details thereof will be further explained below.

FIG. 3(a) shows a side view of the optical information exchange unit 100 incorporated into the image reading section A. The unit 100 is obtained by color separation using an imaging lens 101, an interference filter 102 that cuts ultraviolet and infrared light and transmits 400"700na+, and a prism 103 that separates the visible light that has passed through the filter 102. It has each element of COD 104, 105, and 106 which is a line image sensor that receives a separated image and photoelectrically converts it.Each CCD 104, 105, and 106 has panchromatic spectral sensitivity. There is.

Each of the CCDs 104, 105, and 106 has approximately 5,000 fine (
It consists of a densely arranged array of light-receiving elements (approximately 7 seconds wide).

, = (7) if each c CD 104.105.106
If the positional accuracy between pixels is not satisfied, color correction for each pixel becomes difficult, and the color reproduction of the resulting color image deteriorates.

In the present invention, each CCD is aligned using a jig and then fixed with an adhesive, so that pixel misalignment can be kept small.

Furthermore, a CCD having a 5,000-pixel element is used, and pixel misalignment is prevented by high alignment accuracy (Fig. 3(a)). Of course, in the writing system as well, it is necessary to maintain positional accuracy during multiple scans.

The imaging lens 101. Filter 102. Each element, such as the prism 103 and the CCDs 104, 105, 106, etc., must be constructed with optical precision and designed so as not to cause distortion due to changes in the external environment, mechanical vibrations, etc. Furthermore, the optical system must be designed to have no chromatic aberration.

Figure 7 is a cross-sectional view of the color separation prism and a diagram showing the characteristics of the dichroic filter.
A dichroic filter is formed on the and 103b surfaces. This is an interference filter obtained by alternately vacuum-depositing low refractive index dielectric films and high refractive index dielectric films on a transparent substrate such as glass. Has transmission properties. For example, 1fccD 104 can be used to read 103a in FIG. 7(b) and 103b in FIG. 7(C), 1fccD 104 can read culprit color information, 1fccD 105 can read red information, and 1fccD 106 can read green information.

The light that has passed through the lens 101 enters the prism surface 103c almost perpendicularly. Thereafter, the light is separated by a dichroic filter film (characteristics shown in FIG. 7(b)) on the 103a surface, and the blue component light (400 to 500 nm) is reflected toward the 103c surface. The blue component light is totally reflected at the boundary between the glass and the air layer on the surface 103c, is emitted from the surface 103f almost perpendicularly to the same surface, and is imaged on the CCD 104.

On the other hand, (500~
700n-) travels straight through the prism and reaches the 103(b) plane.

On the 103(b) side, a dichroic ivy filter membrane (7th
Because of the characteristic shown in Figure (c), the red component of 600 to 700 nw is reflected toward the 103d surface. Red component light is 103
Totally reflected at the boundary between the glass surface of the d-plane and the air layer, 103g
The light is emitted from the surface almost perpendicularly to the same surface and is imaged on the CCD 105.

The green component light of 500 to 600 nm passing through the 103b surface reaches the 103e surface, is totally reflected at the boundary between the air layer and the glass surface on the 103e surface, is emitted from the 103h surface almost perpendicularly to the same surface, and is imaged on the CCD 106. do.

The color separation prism 103 having the above structure can be easily constructed by polishing and bonding a large number of prisms having the illustrated surface angles. The prism 103 has a 103e surface, 10
The external dimensions of the 3c plane, 103f, 103g, and 103h plane are configured to form a large equilateral triangle.

With this configuration, the optical path lengths within the glass are all equal, and the lengths are equal to the height of the large equilateral triangle of the prism 103.

It will be understood that by using the prism as described above, the three-color separated optical images are projected onto substantially the same plane.

Now, when optical images separated into each color can be obtained on the same plane as described above, the following advantages arise.

Three CCDs 104, 105, 106 each have 5
Pixel positions cannot be aligned without adjusting the axes. If each color-separated optical image is obtained on the same plane, the CCD can also be arranged on the same plane.

As a result, as shown in Fig. 5, which will be described later, only one 5-axis adjustment mechanism was prepared, and after adjusting the CCD 104 first, the CCD was fixed to the prism 103 with adhesive, etc., and then the y-axis was adjusted. Using a mechanism, an adjustment mechanism is placed at the COD position of 105, and the CCD of 105 is adjusted so as to be based on 104, and similarly fixed to the prism 103 with adhesive or the like. Further, using the y-axis adjustment mechanism, the adjustment mechanism is shifted to the position of CCD 106, and CCD 106 is adjusted to be located between CCDs 105 and 104, and fixed to prism 103 by adhesive or the like.

In this way, adjustment can be easily performed by providing only one adjustment mechanism.

For reference, Fig. 8 shows the spectral characteristics of the document scanning light source, color separation prism, and line image sensor.
Figure (a) shows an example of the spectral characteristics of the document scanning light source 85. The horizontal axis represents wavelength (nIll), and the vertical axis represents relative intensity. Further, FIG. 8(b) shows an example of the spectral characteristics of the filter 102 and the color separation prism 103, which are an example of color separation means. The horizontal axis is wavelength (nm), and the vertical axis is transmittance (%)
It represents. Furthermore, FIG. 8(C) shows an example of the spectral characteristics of Al CCD 104.105.106 using a line image sensor. The horizontal axis represents wavelength (nm), and the vertical axis represents relative sensitivity (%).

Optical information conversion unit 10 in the image reading section A
The configuration of lens barrel 101a, color separation prism 103 (hereinafter referred to as prism), and CCD 10
4, 106 is fixed and held in order to meet the above requirements.
This is done as shown in Figures (a) and (b).

That is, FIG. 3 shows the optical information conversion unit 100 of the present invention.
This is an embodiment of the present invention, in which the lens barrel 101a and the prism 103 are mounted and supported by the following common means.

As shown in FIG. 3(a), the lens barrel 101a is housed in a 7-shaped receiver opening perpendicularly upward to the holding member 109a and is fixed by a fastener 109c before being attached to the device. It is adapted to be attached to a predetermined position on the board 109.

Further, in this embodiment, as shown in FIG. 3(0), the rear end surface of the lens barrel 101a is formed into a mounting surface 101b that can be joined to the front surface of the prism 103, and an adhesive is attached to the mounting surface 101b. The prism 103 is
It is designed to be fixed.

Since the mounting surface 101b is formed by a simple machining process, the distance to the imaging lens 101 housed in the lens barrel 101a and the perpendicularity to the optical axis are extremely accurate, and the mounting surface 101b can be mounted thereon. A predetermined optical image can be correctly formed on the light receiving surfaces of CODs 104, 105, and 106 connected to the substrate 107 through the prism 103 by flexible cables.

That is, as shown in FIG. 3(c), the perpendicularity LR of the plane between the lens barrel mounting surface 101b and the mounting surface 103c of the prism 103
, LR' deviation fi (mounting surface 103 with respect to lens optical axis)
The inclination amount of the perpendicularity LR, LR' of c) is determined by the resolution (M T F ) -111x IQQ%, which is determined from the signal output of the white line part and the black line part against the white background shown in Figure 5.y+x Normal For values of 30% or more, the amount of inclination is 10 in angle.
The surface accuracy is maintained during this period because the angle decreases by around 30% (9%) in 30 minutes, and by more than 50% (15%) in 30 minutes, which hinders the extraction of black and white discrimination signals. By fixing the lens barrel 101a and the prism 103 together in advance, it is possible to prevent a drop in the defective rate in the post-production process, and not only has cost benefits, but also improves the material of the holding member 109a, the mounting member 124, etc. The effect of this on pixel misalignment after the CCD is installed is also reduced, resulting in even greater effectiveness.

Examples of fixing methods include the above-mentioned bonding method, mechanical integration with the lens barrel, and attaching the lens barrel 1 to the holding member 109a.
01a, it is also possible as an effective method to set the prism 103 on both and hold it down.

In order to bond the prism 103 to the lens barrel 101a, an adhesive that has not only adhesive strength but also can withstand various environmental tests is selected and used. .105
and adhesion of 106 and the COD 104°105.
The mounting member and lens barrel 1 after adjusting the position of 106
Since the same adhesive is used for bonding CCD 01a to holding member 109a, the characteristics of the adhesive will be summarized in the section explaining the fixation of CCDs 104, 105 and 106. That's it.

Next, to explain the installation and fixing of c c D 104, 105, 106, as shown in FIGS. 3(a) and (b), a part of the holding member 109a is attached to the prism
A symmetrical support portion 109d is formed which extends to the side surface of the prism 103 and holds the left and right side surfaces of the prism 103 without contacting it.

while said CCD 104.105 and CCD 106
Attachment members 124a,
125a or 126a is fixed with adhesive, and is attached to the support portion 109d via the attachment members 124a, 125a and 126a,
The mounting order is as follows: First, the screw member S screwed into the support part 109d is temporarily fixed through each elongated hole T, and then the mounting member 124 is attached along the elongated hole T.
a.

125a and 126a are moved in a direction perpendicular to the light-receiving surface to adjust their positions and set correctly on their respective light-receiving surfaces of the prism 103, and then the screw members S are tightened and bonded to the holding member 109a. It is set to be fixed.

Furthermore, by partially deforming the shape of the mounting member 124a, 125a, 126a, or 109d, it is possible to support the solid-state image sensor only on one side, as shown in FIG. As shown in Fig. 4 (b) of cylinder 4, a plurality of bolts B and nuts N are attached and tightened so that the mounting members can be pressed together, so that the solid-state image sensor can be sandwiched and supported. .

In addition, as a mounting member having the above-mentioned adjustment mechanism, Fig. 3 and $
Although FIG. 4 shows a single component, it is of course possible to use a plurality of components.

As for the material of the mounting member, a material with a small coefficient of linear expansion is desired to prevent pixel misalignment due to temperature fluctuations.

Pixel misalignment due to temperature fluctuations can be reduced by making the conditions for fixing each CCD to the mounting member exactly the same. Normally, the linear expansion coefficient of a prism is as small as 7.4X101 (optical glass BK-7), so mounting materials such as glass, ceramic materials (7.0~g, 4X10-'), or low thermal expansion alloys (such as pine bar) are recommended. Alloy 1~3X 10-
'), Niresist cast iron (4~10-'), etc. are suitable.
Aluminum material (25 x 10-') is not very suitable.

The present inventors conducted tests using various materials as mounting members, but no image shift due to thermal expansion was detected when glass materials, other ceramic materials, and low thermal expansion alloys were used.

In the above embodiment, the holding member 10 of the lens barrel 101a
9d and mounting members 124, 125, 126, mounting member 12
4,125°126 and CCD 104.105.106
Adhesive was used to fix the CCDs 104, 105, 106 in relation to the divided optical images, and as shown in FIG. In particular, in Figure 3, even if iron (12 x 10'') with a large coefficient of linear expansion is used as the mounting member, the length dimension is short in practical terms, so the elongation due to heat is not affected much, and the depth direction is linear. This is the direction in which the sensors are arranged, and since the prism material and the line sensor package material are ceramic materials, their linear expansion coefficients are approximately the same, and no pixel misalignment occurred in this configuration.

The present inventors conducted comparative studies using many available adhesives. As a result, it was concluded that among two-component type adhesives and photocurable adhesives, especially ultraviolet curable adhesives are most preferable as adhesives to be used in the present invention.

Adhesives include epoxy and acrylic adhesives, and are further divided into one-component types and two-component types. L-liquid type products usually have a hardening agent mixed in during manufacture, and when used, they gradually harden and dry when exposed to air, etc., and harden and solidify.They take a long time to harden, and the degree of hardening and shrinkage during hardening are limited. Due to the irregularity of the surface, etc., it is necessary to use special equipment for adhesive fixation. Therefore, it is not suitable from the viewpoint of the purpose of use and productivity of the present invention. In this regard, two-component and instant-acting products cure in a few minutes at most by kneading the curing agent and the main ingredient during adhesion, shortening the curing time and stabilizing the degree of curing. , which is effectively suited for this purpose. Although there are 1-liquid type cyanoacrylate-based products that are quick-acting, they may cause adhesive peeling upon impact or deformation of the adhered object due to shrinkage of the adhesive when it dries, so this is not necessarily the preferred form of application. I can't say that.

The present inventors used Hardlock E as a two-component type adhesive.
When bonding was carried out at room temperature using 510K (trade name), good results were obtained in the environmental tests described later. During the above bonding process, we attempted to shorten the bonding time by significantly changing the temperature conditions, but the results showed that, although the results were slight, pixel misalignment during bonding was observed.
It became clear that this was not desirable.

On the other hand, photocurable adhesives can speed up the curing time of the adhesive simply by changing the intensity of light, improving workability, reducing costs, and stabilizing the product. Among photo-curable adhesives, ultraviolet-curable adhesives in particular have almost no change in heat even when exposed to ultraviolet rays, and can provide stable effects. The present inventors used ThreeBond T B as a photocurable adhesive.
3060B (product name), electrification 1045K (product name)
, Norland 65 (trade name), etc., and bonding was performed in a short time by irradiating ultraviolet light with a high-pressure mercury lamp.
Good results were also obtained in environmental tests, etc., which will be described later. Furthermore, using UV-curable urethane-based 3062B (trade name), L T 350 (trade name), etc., it was possible to obtain an adhesive that was even more effective in moisture resistance and had guaranteed strength.

In the adhesive bonding method described above, opposing surfaces of the adhesive members of each part are pressed together, and a small amount of adhesive is extruded from the sides of the pressed surfaces using an appropriate extrusion means. Due to its fluidity, the adhesive flows into the slight gap created between the pressure contact surfaces and firmly fixes the adhesive members to each other. The bonding method may be such that the adhesive is extruded so as to bond the front surface to the opposing surface of the bonding member, and then flows into the gap. Alternatively, the adhesive may be introduced with an appropriate gap.

Further, if the device is capable of arranging the adhesive members with appropriate positional accuracy, the adhesive may be applied in advance to the adhesive surface of each adhesive member in spots or areas, and the adhesive surfaces of the adhesive members may be immediately pressed and bonded.

C CD according to the invention 104. For bonding the CG D 105 and the CCD 106 to the prism 103, a mounting jig TC is prepared in advance.

The mounting jig is CCD 104. CCD 105 and CCD
106, as shown in FIG.
Its optical axis and X perpendicular to the optical axis as shown.

Rotation direction θA regarding the two directions of y and the x and y axes.

It is possible to adjust θB, and it is adjusted so that there is no pixel shift by fine adjustment of the mounting jig TC.

A black and white striped chart image placed at the document position is captured on a CCD 1.
04. Images are formed on the CCD 105 and CCD 106, and their output signals are superimposed and recorded on a synchroscope. By setting the black and white stripe interval based on the designed reduction magnification of the imaging lens lO1 and the size of the CCD pixel, and creating a chart in which one stripe corresponds to one pixel, the recorded signal superimposed on the synchroscope The amount of pixel shift can be easily read from. For example, the synchroscope CRT screen shown in FIG. 5 has CCD 104 and CCD 10.
This example illustrates a situation where there is a pixel shift between CCD 106 and CCD 106. Also, by adjusting the mounting jig TC while checking with a synchroscope, the relative position between the CCDs without pixel shift can be determined, and at this position, the CCD 104. CCD105
In the case of the embodiment, the mounting member 12 and the CCD 106 are respectively attached to the support portion 109d of the holding member 109a.
4a, 125a, and 126a at the imaging position. In FIG. 5, the pixel shift is enlarged for understanding. After this fixation, CCD 104 and CCD 1
Install and adjust the CCD 106 so that it is in the middle position of the 05 signal and fix it.

〔Effect of the invention〕

As explained above, the color image reading device of the present invention has the following features:
Since the pixels of the line image sensor on which the green component image is projected are adjusted to be centered, the toner image formed by the output from each pixel is centered around the magenta toner image, and the yellow , in a color image consisting of magenta, cyan, and black toner images,
The black image is constructed so as to overlap the magenta image, so even if there is an image shift, the colors that protrude from the black image will be a portion of yellow on one side of the black image and a portion of cyan on the other side of the black image. It becomes something that sticks out. Therefore, the pixel misalignment becomes extremely inconspicuous, the output image is not affected by the pixel misalignment between the line image sensors, and the pixel misalignment takes a specific shape and is easy to correct. A color image reading device can be provided.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing a general configuration of a color image forming apparatus and a control system equipped with a color image reading device of the present invention, and FIG. 2 is a block diagram showing an outline of an image forming process using the apparatus shown in FIG. 1 above. 3 and 4 are diagrams showing the fixed state of an embodiment of the optical information conversion unit of the present invention, and FIG. 5 is an explanatory diagram of the position adjustment method of the line image sensor (COD) of the present invention. Fig. 6 is a perspective view showing the adjustment direction of the line image sensor (CCD), Fig. 7 is a cross-sectional view of the color separation prism and a diagram showing the characteristics of the dichroic filter, Fig. 8 is the light source for document scanning, and the color separation prism. FIG. 9 is a diagram showing the spectral characteristics of the line image sensor, and FIG. 9 shows the other blue color relative to the green component image when the line image sensor is installed at an angle. FIG. 1O, which is a diagram showing the amount of pixel shift of a red component image, is an enlarged diagram showing a state in which toner images of three or four colors are overlapped. 100... Optical information conversion unit 101... Lens 101a... Lens barrel 101b... Mounting surface 102... Infrared cut filter 103... Color separation prisms 103a, 103b... Dichroic ivy filter 1
04-106...Line image sensor (CCD)
107...Substrate 109...Device board 109a...Holding member 109d...Supporting parts 124a, 125a, 126a...Supporting member TB
K...Black toner TC...Cyan toner TM...Magenta toner TY...Yellow toner P...Transfer material

Claims (2)

[Claims] (1) In a color image reading device in which a color-separated image is projected onto each of three aligned line image sensors, the pixels of the line image sensor corresponding to the same position are the pixels on which the green component image is projected. A color image reading device characterized in that it is arranged at the center. (2) The color image reading device according to claim 1, wherein in setting the line image sensor, setting of the image sensor on which the green component image is projected is performed last.