JPH0998288A - Digital image reader - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide the digital image reader by which a sharp image is acquired by converging shading data more quickly so as to attain accurate shading correction even when the length of a reference white board is shorter. SOLUTION: A signal formed on a sensor of an image forming optical system is amplified by an analog processing section to convert the signal into digital image data, shading data based on image data 511 are stored in a storage means 505, updated by the weight means in an update means for each picture element, shading data of the storage means 505 updated for a prescribed number by a correction means are used to apply shading correction to the image data. The storage means 505 stores shading data based on image data of a reference white board read in preceding to the reading of an original image as an initial value, and the update means updates the data through weight means for a prescribed number of times.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a digital image reading apparatus which is applied to a copying apparatus, a facsimile apparatus, a scanner apparatus and the like, and which digitizes a signal obtained by reading an original image by an image forming optical system to perform signal processing. In particular, the present invention relates to a digital image reading apparatus that enables faster convergence of shading data to a target value in shading correction and realizes more accurate shading correction.

[0002]

2. Description of the Related Art An example of a conventional digital image reading apparatus is disclosed in Japanese Patent Application Laid-Open No. 4-236569.

This conventional digital image reading apparatus removes noise from the image of the reference white plate read before reading the original image by a low-pass filter before storing the image, and stores the noise when reading the original. The shading correction is performed based on the data and the read data.
As a result, the S / N ratio is improved, the peak value in one line does not exceed the actual value, and the density unevenness of the output image is
It is possible to prevent the generation of linear image noise.

[0004]

However, in the above-described conventional digital image reading apparatus, when the low-pass filter is formed in the sub-scanning direction by using the FIR type (non-recursive type) filter, in the sub-scanning direction. There was a problem that the hardware scale increased because many lines of delay memory were required.

Further, in order to reduce the hardware scale,
If the low-pass filter is configured to calculate the weighted average using an IIR type (recursive type) filter, it may take time to converge depending on the initial value. Especially, the length of the reference white plate is short and When the number of lines cannot be read, there is a problem in that the reading of the reference white board is completed before the true value is reached.

The present invention has been made in view of the above conventional problems, and in the shading correction processing,
Aiming to provide a digital image reading device that enables faster convergence of shading data to a target value, performs more accurate shading correction even if the length of the reference white plate is short, and can obtain a clear image There is.

[0007]

In order to solve the above-mentioned problems, the digital image reading apparatus of the first feature of the present invention comprises:
An imaging optical system that irradiates a document with light to form an image of the document on a sensor, an analog processing unit that amplifies a signal formed on the sensor and converts it into digital image data, and a storage unit that holds shading data. Update means for updating the shading data in the storage means with a weighted average for each pixel, and correction means for performing shading correction on the image data using the shading data in the storage means updated a predetermined number of times by the updating means. In the digital image reading apparatus, the updating unit weights a predetermined number of times after the storage unit holds shading data based on the image data of the reference white plate read as an initial value prior to reading the document image. The average is updated.

According to a second aspect of the present invention, there is provided a digital image reading apparatus according to claim 1, wherein the digital image reading apparatus includes a pseudo shading data generating means for generating predetermined pseudo shading data. The updating means updates the weighted average a predetermined number of times after the pseudo shading data is held in the storage means as an initial value.

Further, in the digital image reading device of the third feature, in the digital image reading device according to claim 1 or 2, the coefficient in the weighted average of the updating means can be arbitrarily set.

Further, the digital image reading device of the fourth feature is the digital image reading device according to claim 1, wherein the analog processing section outputs the image data in line units of the original image, The storage means holds the shading data on a line-by-line basis, and the updating means carries out a weighted average update on a line-by-line basis after the initialization of the storage means.

[0011]

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, an outline of a digital image reading apparatus of the present invention and an embodiment of the digital image reading apparatus of the present invention will be described. ] Will be described in detail with reference to the drawings.

[Outline of Digital Image Reading Apparatus According to the Present Invention] The digital image reading apparatus according to the present invention optically reads original image information using a line image sensor such as a digital copier, a facsimile, and a scanner, and converts it into an electric signal. It is applied to devices, etc., and is characterized in that even if the length of the reference white plate is short, accurate shading correction can be performed and a beautiful image can be obtained.

In the digital image reading apparatus having the first and fourth characteristics of the present invention, as shown in FIGS. 3 and 5, the analog processing section 302 amplifies the signal formed on the sensor of the image forming optical system. Then, after converting into digital image data, the following image processing is performed. Although the shading data based on the image data 511 is held in the storage unit 505, it is updated by the weighting average for each pixel by the updating unit. After that, the correction unit 305 performs shading correction of the image data using the shading data of the storage unit 505 that has been updated a predetermined number of times, and corrects the sensitivity variation for each pixel of the sensor and the uneven illumination. Here, in the updating means, after the shading data based on the image data of the reference white plate read prior to the reading of the original image is held in the storage means 505 as the initial value, the updating by the weighted average is performed a predetermined number of times. I am trying. The update means corresponds to the configuration excluding the storage means 505 shown in FIG. In particular, in the digital image reading apparatus of the fourth characteristic, each processing described above is performed line by line in the original image.

As described above, since the shading data based on the image data of the reference white plate (first line) read prior to the reading of the original image is used as the initial value of the storage means 505, the conventional method is used. Initial value "0"
It is possible to obtain the shading data that converges to the target value faster than the method in which the weighted average is updated, and as a result, it is possible to perform accurate shading correction even if the length of the reference whiteboard is short, and clear It is possible to obtain a clear image.

Further, in the digital image reading apparatus having the second and fourth characteristics, as shown in FIGS. 3 and 6, the updating means is generated in the storage means 505 by the pseudo shading data generating means 604 as an initial value. After the pseudo shading data is held, the weighted average is updated a predetermined number of times. The updating means is shown in FIG.
Pseudo shading data generation means 604 shown therein
And the configuration excluding the storage unit 505 is applicable. In particular, the 4th
In the digital image reading apparatus having the above feature, each of the above processes is performed in line units of the original image.

As described above, since the pseudo shading data larger than "0" is used as the initial value of the storage means 505, updating is performed from the initial value "0" by the weighted average as in the prior art. Compared with, it is possible to obtain shading data that converges to the target value faster, and even if the image data of the first line of the reference white board has a large value due to noise, its influence can be removed. As a result, even if the length of the reference white plate is short, accurate shading correction can be performed and a clear image can be obtained.

Further, in the digital image reading apparatus having the third and fourth characteristics, the coefficient in the weighted average of the updating means can be arbitrarily set (for each line). Therefore, for example, a line that can read a reference white board By changing the weighted average coefficient depending on the number, the number of lines until the shading data converge can be controlled. As a result, the scanning speed of the scanner changes due to scaling in the sub-scanning direction, etc., and accurate shading data can be obtained even when the data of the reference white board of a predetermined number of lines cannot be read, resulting in noise resistance and shading. It is possible to achieve both faster convergence of the data to the target value.

[Embodiment 1] FIG. 1 is a front view of a digital copying apparatus to which a digital image reading apparatus according to an embodiment of the present invention is applied. 2 is a circuit configuration diagram of an analog processing unit in the digital image reading apparatus of the embodiment, FIG. 3 is a configuration diagram of a control unit in the digital image reading apparatus of the embodiment, and FIG. 4 shows an operation of the control unit. 5 is a timing chart to be described, and FIG. 5 is a circuit configuration diagram of a shading correction unit of the digital image reading apparatus according to the first embodiment.

First, the configuration of a digital copying apparatus to which the digital image reading apparatus of this embodiment is applied will be described with reference to FIG. In FIG. 1, a contact glass 101 for placing a read document is a light source 102a and a light source 10a.
The light reflected by the image surface of the read document is illuminated by the mirror 2b and the mirrors 103, 104, 105, 106 and 1
An image is formed on the light receiving surface of the line image sensor 109 via 07 and the lens 108.

In addition, the light source 102 and the mirror 103 are such that the lower surface of the contact glass 101 is the contact glass 101.
The mirrors 104 and 105 are mounted on the first traveling body 110 that moves in the sub-scanning direction in parallel with the second traveling body 111 that moves in the sub-scanning direction at a speed of 1/2 in conjunction with the traveling body 110. The image plane of the read document is scanned in the sub-scanning direction by moving these optical systems.

The image reading resolution of the digital image reading apparatus of this embodiment is 16 [pixels / mm].
The A3 size document can be read. Therefore, A3 size (297 [mm] x 420 [m
m]) for reading the short side 297 [mm] in the main scanning direction.
A CCD line image sensor of [pixel] is used.

Next, the circuit configuration and operation of the analog processing section in the digital image reading apparatus of this embodiment will be described with reference to FIG. First, the image data formed on the CCD line image sensor 109 is stored in the CCD 201.
Is separated into an ODD output OS1 and an EVEN output OS2 and output in a staggered manner. In this embodiment, the CCD 201 having the outputs DS1 and DS2 for compensating the noise of the reset clock is used.

Next, in the first differential amplifier 202, the CCD 2
The ODD output OS1 of 01 and the noise compensation output DS1 are differentially amplified by the operational amplifier 211 to cancel the noise component of the reset clock to cancel the signal component O.
Only DDOUT is taken out. Similarly, for the second differential amplifier 203, the EV of the CCD 201
Input EN output OS2 and noise compensation output DS2,
A signal component EVENOUT in which the noise component of the reset clock is canceled is taken out.

Since the ODD output OS1 and the EVEN output OS2 from the CCD 201 are output in a staggered manner, the output signal ODDOUT of the first differential amplifier 202 and the output signal EVENOUT of the second differential amplifier 203 are A switch circuit 2 whose ON / OFF switching is controlled by a first gate signal 241 and a second gate signal 242.
04 and 205 will be combined.

This combined signal is amplified and impedance-converted by the operational amplifier 213 and input to the terminal Vin of the A / D converter 209. In the A / D converter 209, a digital image signal of 8 [bit] and 256 [gradation] is obtained by A / D converting the input signal to the terminal Vin. The operational amplifier 214
And 215 are reference voltages Vref + and V, respectively.
It is for generating ref-.

Next, the configuration of the control unit in the digital image reading apparatus of this embodiment will be described with reference to FIG. 3, and the operation of the control unit will be described with reference to FIG.

First, in FIG. 3, the control unit of the digital image reading apparatus is shown in FIG. 2 and a CPU 303, a stepping motor 301 for driving and controlling the first traveling body 110 and the second traveling body 111 under the control of the CPU 303. And an image processing unit (hereinafter abbreviated as IPU) 304 that takes in a digital image signal from the analog processing unit 302 and performs image processing under the control of the CPU 303. There is. The IPU 304 includes a shading correction unit 305.

Next, referring to FIG. 4, a schematic operation of the control unit will be described along with an actual scanning operation. First, when the carriage starts to move from the home position by driving the stepping motor 301, the CPU 303 outputs the signal SLEAD indicating the effective reading range of the reference white board at the active (“L”) level. At this time, the pulse of the synchronization signal RDSYNC in the main scanning direction is also periodically output.

Inside the IPU 304, the gate signal WTGTN is synchronized with the main scanning synchronization signal RDSYNC after the reading effective range signal SLEAD becomes active level.
Are generated as active (“L”) levels. When the gate signal WTGTN becomes active ("L") level, the carriage reads the data of the reference white board while moving. At this time, the gain of the operational amplifier 213 is determined from the first two lines of the reference whiteboard data.

When the carriage further advances and enters the effective range for reading the original, the scan signal SSCAN is output as an active ("L") level. Similarly to the generation of the gate signal WTGTN by the read effective range signal SLEAD, the gate signal FGATEN is generated as the active (“L”) level in synchronization with the main scanning synchronization signal RDSYNC because the scan signal SSCAN becomes the active level. To be done.

Next, the description of the shading correction unit in the digital image reading apparatus of the present embodiment will be made. Before that, the operation of the conventional shading correction unit will be described based on the block diagram of the conventional shading correction unit shown in FIG. Explain and clarify the problem.

In FIG. 8, the conventional shading correction unit comprises a calculation unit 501, a selector 802, a line buffer 505, a comparator 506, and gate circuits 507 and 808. In the figure, 511 is input image data, 812 is shading data, cl
rgate is a clear gate signal and WTGTN is a gate signal.

In the conventional shading correction section, as shown in FIG.
As shown in (d), the clear gate signal cl is generated in the first one line when the gate signal WTGTN becomes "L" level.
rgate is generated, the shading data is cleared to zero, and stored in the line buffer 505 having a capacity of 5000 × 8 [bit].

The next one-line image data 511 and the shading data stored in the line buffer 505 are
A weighted average is calculated by the calculation unit 501 for each pixel and compared with the original shading data. Then, the larger one is selected by the selector 802, and the line buffer 50
5 is rewritten to the same address and updated. At this time, the weighted average is calculated based on the following equation. Ds (n) = (3 × Ds (n-1) + Dr (n)) / 4 (1) where Ds (n-1) is the shading data of one line before, and Dr (n) is the reading of the reference white board. Data, n is a pixel pointer (1 to 5000).

Furthermore, while the gate signal WTGTN is at "L" level, the reference white board is read over several tens of lines and the data in the line buffer 505 is updated.
Avoid the influence of dust and scratches, and eliminate the uneven sensitivity of lighting, CCD, etc.
It can be expressed as unevenness of the white level data in the line buffer 505.

After the gate signal WTGTN becomes "H" level, the selector 802 always keeps the line buffer 50.
Since the output data from 5 is selected, the contents of the line buffer 505 will not be updated.
However, in this conventional method, the shading data is accurately obtained when the number of effective reading lines of the reference white plate cannot be secured due to the layout of the machine such that the reference white plate cannot be widened or when the size is reduced in the sub-scanning direction. There was a problem that it could not be done.

Therefore, the shading correction unit 3 of this embodiment
In No. 04, the shading data has no noise and the shading data converges to a steady value (target value) earlier. FIG. 5 shows a circuit configuration diagram of the shading correction section of the digital image reading apparatus according to the present embodiment.

In FIG. 5, the shading correction unit of the present embodiment comprises a calculation unit 501, selectors 502 and 503,
A line buffer 505, a comparator 506, and a gate circuit 507 are provided. In the figure, 51
1 is input image data, 512 is shading data,
clgate is a clear gate signal, and WTGTN is a gate signal.

At the first one line when the gate signal WTGTN becomes "L" level, the clear gate signal clrgate is generated as active ("L") level as shown in FIG. 4 (e). At this time, as the shading data written in the line buffer 505, not the zero-cleared data but the input image data 511 obtained by reading the reference white board is selected, and the shading data in the line buffer 505 having a capacity of 5000 × 8 [bit] is selected. Stored in.

The next one-line image data 511 and the shading data stored in the line buffer 505 are
A weighted average is calculated by the calculation unit 501 for each pixel and compared with the original shading data. Then, the larger one is selected by the selector 503, and the line buffer 50
5 is rewritten to the same address and updated.

Further, while the gate signal WTGTN is at the "L" level, the reference white plate is read over ten lines and the shading data of the line buffer 505 is updated to avoid the influence of dust and scratches and to prevent illumination, CCD, etc. The sensitivity unevenness can be expressed as the unevenness of the white level data in the line buffer 505.

After the gate signal WTGTN becomes "H" level, the calculation unit 501 always keeps the line buffer 50.
5 is selected, the shading data 512 held in the line buffer 505 is not updated.

Shading correction is performed using the shading data 512 generated by the shading correction unit 305. The shading correction method is performed by calculating the following equation which is already known. DOUT (n) = DIN (n) / Ds (n) × 255 (2) where DOUT (n) is post-shading correction data, DIN (n) is pre-shading correction data, and Ds
(N) is shading data, n is a pixel pointer (1
~ 5000).

As described above, in the shading correction unit 305 of this embodiment, when updating the shading data,
Since the initial value when calculating the weighted average is not “0” but the read raw data of the reference white board, as shown in FIG. 4 (g), the shading data 512 is converged by reading the reference white board data of several lines. Even when the number of read lines on the reference white board is not large, accurate shading data 512 without noise can be obtained.

[Second Embodiment] Next, FIG. 6 is a circuit diagram of a shading correction section in a digital image reading apparatus according to a second embodiment of the present invention.

In FIG. 6, the shading correction unit of this embodiment comprises a calculation unit 501, selectors 602 and 603.
Line buffer 505, comparator 506, gate circuit 50
7 and a pseudo shading data generator 604. In the figure, 511 is input image data, 612 is shading data, clrgat
e is a clear gate signal, WTGTN is a gate signal, lg
aten is a gate signal.

In the pseudo shading data generator 604, pseudo shading data is generated in synchronization with the active period of the gate signal lgaten. This pseudo shading data uses pseudo shading data which is slightly smaller than the shading data when the reference white board is actually read. The pseudo shading data is generated by a known method such as reading out data stored in advance in a ROM or the like.

In the first one line when the gate signal WTGTN becomes the "L" level, the clear gate signal clrgate is generated as the active ("L") level as shown in FIG. 4 (e). At this time, the data written in the line buffer 505 is stored in the line buffer 505 having a capacity of 5000 × 8 [bit] with pseudo shading data selected.

Image data 51 of one line of the next reference white board
The shading data stored in 1 and the line buffer 505 is weighted and averaged by the calculation unit 501 for each pixel and compared with the original shading data. Then, the larger one is selected by the selector 603, rewritten to the same address of the line buffer 505, and the update processing is performed.

Further, while the gate signal WTGTN is at "L" level, the reference white board is read and the line buffer 50 is read.
By updating the shading data of No. 5, it is possible to avoid the influence of dust and scratches and express the unevenness of the sensitivity of the illumination, CCD, etc. as the unevenness of the white level data in the line buffer 505.

After the gate signal WTGTN becomes "H" level, the calculation unit 501 always keeps the line buffer 505.
The shading data in the line buffer 505 is not updated because it is configured to select the output data from. Next, shading correction is performed using the shading data 612 thus generated. The shading correction method is the same as that in the first embodiment.

As described above, since the shading correction unit of this embodiment uses a large initial value in advance, the shading data shown in FIG.
Shading data converges faster than the conventional example. Therefore, accurate shading data 612 without noise can be obtained even when the number of read lines on the reference white plate is not large.

Further, in the case of the shading correction unit of the first embodiment, when the value of the first line of the reference whiteboard read data is large due to noise, there is a risk that the data will be held. However, in this embodiment, since the pseudo shading data is used as the initial value, the influence of noise can be removed.

[Third Embodiment] Next, in the shading correction section in the digital image reading apparatus according to the third embodiment,
The configuration equivalent to that of FIG. 5 (Embodiment 1) or FIG. 6 (Embodiment 2) is provided, and the weighted average coefficient of the calculation unit 501 can be changed for each line under the control of the CPU 303.

As a method of performing the weighted average, the following three equations are used. Ds (n) = (Ds (n-1) + 3 * Dr (n)) / 4 (3) Ds (n) = (2 * Ds (n-1) + 2 * Dr (n)) / 4 (4) Ds (n) = (3 × Ds (n−1) + Dr (n)) / 4 (5) where Ds (n−1) is the shading data of one line before, and Dr (n) is the reference whiteboard. Read data of
n is a pixel pointer (1 to 5000).

The weighted average by the equation (3) has a characteristic that the convergence is fast but the effect of removing noise is small. Further, the weighted average according to the equation (5) has a characteristic that the effect of removing noise is great, but the convergence is slow. Furthermore, the weighted average according to (4) has a characteristic in between.

FIG. 7 is a timing chart for explaining the operation of the control section (shading correction section) of this embodiment. The operation of the shading correction unit of this embodiment will be described with reference to FIG.

At the time of reading the reference white board, the weighted average according to the equation (3) is performed for the first several lines. With this coefficient, the shading data increases monotonically and tries to converge to the target value quickly, so the length of the reference white board can be shortened.

In the next 5-line section, the weighted average according to the equation (4) is performed. With this coefficient, the shading data converges more slowly than the weighted average by equation (3),
Noise resistance is improved.

Furthermore, for the last 10 lines, the weighted average according to equation (5) is performed. By this time, the shading data seems to have almost converged to the target value, so the noise removal is mainly performed.

As described above, in the shading correction unit of the present embodiment, the coefficients are switched to perform the weighted averaging, whereby the shading data can be converged quickly to the target value, and accurate shading data without noise can be obtained. It becomes possible. Although the number of lines in each section is set as described above in this embodiment, it is desirable to adjust the number of lines in each section according to the configuration of the reading system such as the length of the reference white plate and the scanning speed.

[0062]

As described above, according to the digital image reading apparatus of the first and fourth features of the present invention, the updating means is read in the storage means as an initial value prior to reading the original image. After the shading data based on the image data of the reference white board is held, the weighted average is updated a predetermined number of times, so that the shading data that converges to the target value can be obtained earlier, and as a result, It is possible to provide a digital image reading device capable of performing accurate shading correction even when the length is short and obtaining a clear image.

Further, according to the digital image reading apparatus of the second and fourth characteristics, the updating means holds the pseudo shading data generated by the pseudo shading data generating means as the initial value in the storage means. Since the weighted average is updated a predetermined number of times, the shading data that converges to the target value can be obtained earlier, and the image data on the first line of the reference whiteboard has a large value due to noise. In addition, it is possible to eliminate the effect, and as a result, to provide a digital image reading device capable of performing a precise shading correction even when the length of the reference white plate is short and obtaining a clear image. You can

Further, according to the digital image reading apparatus having the third and fourth characteristics, the coefficient in the weighted average of the updating means can be arbitrarily set (for each line), so that the magnification in the sub-scanning direction is changed. Accurate shading data can be obtained even when the scanning speed of the scanner changes and the data of the reference white board of a predetermined number of lines cannot be read. As a result, noise resistance and faster convergence of the shading data to the target value can be obtained. It is possible to provide a compatible digital image reading apparatus.

[Brief description of drawings]

FIG. 1 is a front view of a digital copying apparatus to which a digital image reading apparatus according to an embodiment of the present invention is applied.

FIG. 2 is a circuit configuration diagram of an analog processing unit in the digital image reading apparatus of the embodiment.

FIG. 3 is a configuration diagram of a control unit in the digital image reading apparatus according to the embodiment.

FIG. 4 is a timing chart illustrating the operation of the control unit.

FIG. 5 is a circuit configuration diagram of a shading correction unit of the digital image reading apparatus according to the first embodiment.

FIG. 6 is a circuit configuration diagram of a shading correction unit in the digital image reading apparatus according to the second embodiment of the present invention.

FIG. 7 is a timing chart illustrating an operation of a control unit (shading correction unit) according to the third exemplary embodiment.

FIG. 8 is a configuration diagram of a conventional shading correction unit.

[Description of Reference Signs] 101 Contact Glass 102a, 102b Light Source 103-107 Mirror 108 Lens 109 Line Image Sensor 110, 111 Traveling Body 201 CCD 202, 203 Differential Amplifying Unit 204-208 Switch Circuit 209 A / D Converter 210 Constant Voltage Diode 211-215 operational amplifier 221-239 resistance OS1 ODD output OS2 EVEN output DS1, DS2 reset clock noise compensation output ODDOUT output signal of the first differential amplifying unit 202 EVENOUT output signal 241,242 gate signal 243 of the second differential amplifying unit 203 ~ 245 Gain signal Vref +, Vref- Reference voltage 301 Stepping motor 302 Analog processing unit 303 CPU 304 Image processing unit (IPU) 305 Fading correction unit SLEAD Effective reading range signal RDSYNC Main scanning synchronization signal WTGTN gate signal SSCAN scan signal FGATEN gate signal 501 calculation unit 502,503,602,603,802 selector 505 line buffer 506 comparator 507,808 gate circuit 511 input image data 604 Pseudo shading data generator 512, 612, 812 Shading data clrgate Clear gate signal lgaten Gate signal

Claims (4)

[Claims] 1. An image forming optical system for forming an image of an original on a sensor by irradiating an original with light, an analog processing unit for amplifying a signal formed on the sensor and converting it into digital image data, and shading data. Shading correction is performed on the image data by using storage means for holding the shading data, updating means for updating the shading data in the storage means with a weighted average for each pixel, and shading data in the storage means updated a predetermined number of times by the updating means. In the digital image reading apparatus having the correction means, the shading data based on the image data of the reference white plate read prior to the reading of the original image is stored in the storage means as the initial value. After that, the digital image reading device is characterized in that updating is performed by a weighted average a predetermined number of times. 2. The digital image reading apparatus has a pseudo shading data generating unit for generating predetermined pseudo shading data, and the updating unit stores the pseudo shading data as an initial value in the storage unit. The digital image reading apparatus according to claim 1, wherein after being held, the weighted average is updated a predetermined number of times. 3. The digital image reading apparatus according to claim 1, wherein the coefficient in the weighted average of the updating means can be set arbitrarily. 4. The analog processing unit outputs the image data in line units of the original image, the storage unit holds shading data in line units, and the updating unit sets line units after initial setting in the storage unit. 4. The digital image reading apparatus according to claim 1, wherein the updating is performed by weighted averaging for a predetermined number of lines.