JPH05161005A - Envelope signal correction method for picture reader - Google Patents

Abstract

PURPOSE:To correct a notch or a peak of an envelope signal by correcting the envelope signal obtained through reading a reference area by means of a weight mean of plural picture elements. CONSTITUTION:A scanner 11 reads a reference white background 10 under the control of a control circuit 22 and a picture signal enters an amplifier 13, an A/D converter 15 and a latch section 16 and a digital envelope signal is inputted to a shading correction memory 19. The circuit 22 corrects the envelope signal through a weight mean of plural picture elements so that the envelope signal including a notch signal or a peak signal is made close to the desired envelope signal, and the result is stored again in the memory 19 via a selector 21. The control circuit 22 executes the read of the original after the correction. In this case, a switch 18 is thrown to select the analog envelope signal after the correction from a D/A converter section 17 and inputs the result to a conversion section 15 as a reference signal Vref to correct the fluctuation in the signal from the original read by the scanner 11 and to output the result to a picture processing section 23.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an envelope signal (reading reference signal) correction method for an image reading apparatus.

[0002]

2. Description of the Related Art In an image reading device such as an image scanner device, an electronic blackboard device, a facsimile device, a digital copying device, etc., an image signal in the main scanning direction of an original is picked up by a photoelectric conversion section composed of, for example, a one-dimensional CCD sensor. ing. Even if the same color original is read for one line, the image signal for one line output from the CCD sensor varies. This is because the light amount of the light source that irradiates the original may vary in the main scanning direction, and the characteristics of the optical system that forms an image on the CCD sensor in the main scanning direction may vary.

Therefore, for example, an image signal in the main scanning direction when a reference white background is read is stored as a read reference signal (hereinafter referred to as an envelope signal), and an image signal when a document is read is stored in the envelope. Correction and processing are performed based on the signal. For example, the envelope signal is used as a reference level when analog / digital converting an image signal from a document.

If the envelope signal itself used as the reference is erroneous in this way, the processing accuracy of the image signal read from the original document is deteriorated. Therefore, conventionally, various image reading apparatuses are provided with an envelope signal correction circuit that corrects an envelope signal obtained from an image sensor to reduce noise components and the like contained in the envelope signal.

FIG. 2 is a block diagram showing a conventional envelope signal correction circuit of this type.

The correction circuit shown in FIG. 2 obtains a final corrected envelope signal by using a plurality of envelope signals when a reference white background is read, for example, and the one-dimensional CCD sensor 1 and the amplifier 2 are provided. , Analog / digital converter 3, latch circuit 4, magnitude comparator 5, envelope signal memory 6 and control circuit 7.

CCD sensor 1 read from a reference white background
The envelope signal of the first line, which is photoelectrically converted by the amplifier 2 and then converted into a digital signal through the analog / digital converter 3 in sequence, is controlled by the control circuit 7 and then the envelope circuit via the latch circuit 4 as it is. It is given to the signal memory 6 and stored.

The digital envelope signal for the next line output from the analog / digital converter 3 is processed as follows in pixel units under the control of the control circuit 7. That is, the pixel signal is given to the latch circuit 4 and latched, and at the same time, the magnitude comparator 5
Is compared with the pixel signal at the same position in the main scanning direction read from the envelope signal memory 6 and the pixel value of the envelope signal at this time is large, this is detected from the latch circuit 4 by the envelope signal. It is given to the signal memory 6 and replaced with the value at that position.

At the time when all the pixel-by-pixel processing for the digital envelope signal of the next line is completed,
Of the corresponding pixel values of the envelope signals for the two lines of the first line and the next line, the larger pixel value is the envelope signal memory 6
Are stored in each pixel area of.

With respect to the envelope signal from the subsequent line, the same processing as described above is executed, and as a result, each pixel area of the envelope signal memory 6 has the largest pixel value of all envelope signals. It is stored, and this stored content becomes the final corrected envelope signal.

In this conventional circuit, it is premised that the envelope signal from a plurality of lines includes an originally large value for the reference white background even if the reference white background has dust or dirt.

[0012]

However, in the envelope signal correction circuit described above, a plurality of lines of the reference white background are C
Since it is a method of reading with a CD sensor and determining the envelope signal with the maximum value as the white level, some dust and dirt can be corrected, but when dust and dirt spread over the entire line, this dust and dirt can be corrected. This causes a problem that the envelope signal level drops and the envelope signal is different from what it should be.

In order to avoid this, it is necessary to increase the number of lines for reading the reference white background, which results in a long processing time. Further, when a partial white area of the document is used as a reference white background, the increase in the number of lines is a serious problem.

Although the reference white background level should be the highest, in practice, a level higher than the reference white background level may occur due to external noise when the envelope signal is determined. In the conventional circuit, the high level due to this noise is included in the corrected envelope signal as it is and stored in the memory, which is problematic.

Further, in the conventional circuit, in order to read a plurality of lines of the reference white background and leave the highest level among the pixels at the same position in the main scanning direction as the white level in the memory,
The comparison is done by the hardware configuration,
Therefore, there is a problem that the hardware configuration becomes complicated and expensive.

The present invention has been made in consideration of the above points, and a good envelope signal can be obtained even if the reference area is dusty or dirty, and noise is mixed in the read envelope signal. It is an object of the present invention to provide an envelope signal correction method for an image reading apparatus, which is capable of obtaining the above-mentioned characteristics and which is capable of downsizing the entire structure.

[0017]

According to the present invention, an image signal obtained by reading a reference area of the same color in the main scanning direction is stored as an envelope signal, and then the envelope signal is trashed in the reference area. The present invention relates to an image reading apparatus that corrects a level drop due to dirt and dirt and a level fluctuation due to external noise, and corrects an image signal obtained by reading a document based on a corrected envelope signal. The method is characterized by a method for obtaining a final corrected envelope signal from the envelope signal obtained in (1).

That is, centering on the currently targeted pixel for which the value of the corrected envelope signal is to be obtained,
The values of a plurality of pixels of the envelope signal by the reading operation, which are separated by a predetermined number of pixels, are taken out, and these pixel values are weighted and averaged to obtain the pixel value of the corrected envelope signal which is the current target, It is characterized in that a corrected envelope signal is obtained from the envelope signal by the reading operation.

The values of a plurality of pixels of the envelope signal obtained by the reading operation, which are separated by a predetermined number of pixels, are taken out, and these pixel values are averaged to obtain the value of the pixel located in the middle of the plurality of pixels. By performing this processing for the entire range of the envelope signal to perform the first correction on the envelope signal, and repeatedly executing the same method as the first correction method, It is preferable to obtain the final corrected envelope signal after performing the weighted averaging process.

[0020]

The envelope signal obtained by reading the reference area in the main scanning direction contains a level drop due to dust and dirt in the reference area and a level fluctuation due to external noise.

As in the present invention, if the values of a plurality of pixels of the envelope signal related to the reading operation are extracted and these pixel values are weighted and averaged, it is possible to remove the level drop and the protrusion. A good corrected envelope signal can be obtained.

Since such a correction method is often executed by software, the values of a plurality of pixels of the envelope signal due to the reading operation, which are separated by a predetermined number of pixels, are taken out, and these pixel values are averaged to obtain the value. The value of a pixel located in the middle of a plurality of pixels is obtained, and such processing is executed for the entire range of the envelope signal to perform the first correction on the envelope signal, which is the same as the first correction method. It is preferable that the above method is repeatedly executed to execute the above-mentioned weighted averaging process so that the software can be easily changed and the memory capacity can be reduced.

[0023]

Embodiments of the present invention will be described below in detail with reference to the drawings.

First, a first embodiment of the present invention will be described.

FIG. 1 is a flow chart showing a method of correcting an envelope signal according to the first embodiment, and FIG. 3 is a block diagram showing a configuration around an input section of an image reading apparatus according to this embodiment.

In FIG. 3, a reference area 10 formed of, for example, a white background (hereinafter referred to as “reference white background”) 10 is a sub-scanning direction (Y) of a translucent platen surface 12 on which an original document (not shown) is placed.
Main scanning direction (X
Direction). The length of the reference white background 10 in the main scanning direction covers the maximum scannable range XR of the platen surface 12 in the main scanning direction, and is selected longer than that. The platen surface 12 is opened in the main scanning direction by a range XW slightly wider than the maximum readable range XR and in the sub-scanning direction by a range YW slightly wider than the maximum readable range YR.

The reading optical system (scanner) 11 is a CCD sensor 11b as a one-dimensional image sensor extending in the main scanning direction, and an image read from a document placed on the platen surface 12 or a reference white background 10 by this CCD sensor. The objective lens 11a for forming an image on 11b is included.
The scanner 11 is fixed to a predetermined position to perform the reading operation when reading the reference white background 10, and repeatedly performs the reading operation while sequentially traveling in the sub scanning direction by a motor or the like (not shown) when reading the document.

The control circuit 22 is composed of, for example, a microprocessor and controls the entire image reading apparatus. The control circuit 22 controls a pulse generator (not shown) to shift pulse SH to the scanner 11.
And a control clock CL. Based on these signals, the scanner 11 executes a reading operation in synchronization with the operation of each of the other parts, and outputs a read image signal photoelectrically converted by the CCD sensor 11b to the amplifier 13. The amplifier 13 is the scanner 1
The image signal output from 1 is amplified to a certain level and output to the normalization circuit 14.

The normalizing circuit 14 is a circuit for correcting reading optical system characteristics such as irradiation unevenness of a light source (not shown) and peripheral light amount deterioration characteristics due to the objective lens 11a. That is, the read image signal from the reference white background 10 is stored as an envelope signal indicating the characteristics of the reading optical system, and the read image signal from the document is corrected by using this envelope signal to vary the characteristics of the reading optical system. The effect of is removed. The envelope signal read from the reference white background is not used as an envelope signal used to remove the influence of variations in the characteristics of the reading optical system, but is used after being corrected. ..

The normalization circuit 14 includes an analog / digital converter (A / D converter) 15, a latch 16, and a digital / digital converter.
Analog converter (D / A converter) 17, analog switch 18, shading correction memory 19, counter 20
And a selector 21. Control circuit 22
Also constitutes a part of the normalization circuit 14.

The A / D converter 15 converts the analog image signal from the amplifier 13 into a digital image signal with the analog reference voltage Vref as a reference, and outputs the obtained digital image signal to the data line 15a. Standard white background 1
For analog image signals that read 0, a fixed value V
The analog reference voltage Vref of P is given to the A / D converter 15, and a digital image signal (envelope signal) proportional to the input analog image signal is output. On the other hand, with respect to the analog image signal obtained by reading the original, the analog reference voltage Vref corresponding to the envelope signal is given to the A / D conversion unit 15, which causes a variation in the characteristics of the reading optical system included in the analog image signal. The digital image signal without the influence is output.

The digital image signal output from the A / D conversion unit 15 at the time of reading the original is given to the image processing unit 23, which is the next stage circuit of the normalization circuit 14, and is subjected to various processes.
The image processing unit 23 receives a control signal from the control circuit 22 and a shift pulse SH and a control clock CL from a pulse generator (not shown), and executes a predetermined process in synchronization with other units. For example, processing such as binarization, halftone conversion, multi-value conversion, enlargement / reduction, etc. is performed on the input digital image signal.

The latch section 16 latches and outputs the digital image signal of the data line 15a to the data bus line 19a of the shading correction memory 19, which will be described later, under the control of the control circuit 22. The latch unit 16 functions during the reading operation from the reference white background 10, that is, the latch unit 16 functions when the digital image signal is an envelope signal.

The shading correction memory 19 stores the envelope signal. The shading correction memory 19 executes the writing operation of the envelope signal given via the data bus line 19 a under the control of the control circuit 22 during the reading operation from the reference white background 10. During the envelope signal correction operation described later (see FIG. 1), the data bus line 19a
The envelope signal to the data bus line 19a
Write the envelope signal above. During the reading operation from the original, the stored envelope signal is repeatedly output.

The counter 20 is an up-counter having a 1-increment structure and generates an address for the shading correction memory 19 based on the shift pulse SH and the control clock CL, and outputs the generated address to the address bus line 20a. To do. During the read operation from the reference white background 10, this output address becomes the write address for the shading correction memory 19, and during the read operation from the original, this output address becomes the read address for the shading correction memory 19.

In addition to the address bus line 20a from the counter 20, an address bus line 22a from the control circuit 22 is connected to the selector 21, and based on the control signal from the control circuit 22 via the control signal line 22b. , Any address is selected and output to the address bus line 19b for the shading correction memory 19. Standard white background 10
The address from the counter 20 is selected during the reading operation from the document and the reading operation from the document, and the address from the control circuit 22 is selected during the envelope signal correcting operation described later.

The D / A converter 17 converts the digital envelope signal output from the shading correction memory 19 to the data bus line 19a into an analog signal and outputs the analog signal to the analog switch 18. This conversion works effectively during the reading operation from the original.

The analog switch 18 selects the fixed voltage VP during the reading operation from the reference white background 10 based on the control signal from the control circuit 22, and the D / A converter 17 during the reading operation from the original. The analog envelope signal from is selected and given to the A / D converter 15 as the reference voltage Vref.

The control circuit 22 executes the above-mentioned various control operations and, in the case of this embodiment, also performs the envelope signal correction operation.

Next, the basic operation of the image reading apparatus having the configuration shown in FIG. 3 will be described with reference to the timing chart of each part of FIG.

The control circuit 22 first drives a motor (not shown) via a motor drive circuit (not shown) to move the scanner 11 to a position where the reference white background 10 is read. After that, the control circuit 22 controls the analog switch 18 so that the constant voltage VP is the analog reference voltage V of the A / D converter 15.
In addition, the selector 21 is controlled so that the address from the counter 20 is given to the shading correction memory 19. In this state, the scanner 1
In order to set the timing of starting reading of No. 1, a shift pulse SH (see FIG.
(A)) and a control clock CL in the main scanning direction (see FIG. 4).
(B)) is output. The shift pulse SH and the control clock CL are also given to the counter 20 and the like in order to synchronize the timing. Shift pulse S to the scanner 11
When H is input, the CCD sensor 11b of the scanner 11 serially outputs the image signals from all the C0 pixels to the amplifier 13 at time T0 from the time of the shift pulse SH, in synchronization with the control clock CL. The time T0 and the number of pixels C0 are determined by the characteristics of the CCD sensor 11b.

An analog image signal (envelope signal) obtained by reading the reference white background 10 is amplified by an amplifier 13 and then fixed at an analog reference voltage Vref which is a fixed voltage VP.
Is converted into a digital envelope signal by the A / D conversion unit 15. The digital envelope signal is sent to the latch unit 1
It is given to the shading correction memory 19 via 6 and stored.

Since the A / D converter 15 uses the analog reference voltage Vref of the fixed voltage VP as a reference, the waveform shape of the analog envelope signal is maintained even in the digital envelope signal. That is, if a drop signal or a protrusion signal is mixed in the analog envelope signal due to dust, dirt, noise, or the like, unnecessary signals are mixed in the digital envelope signal as shown in FIG. 4C. It

The irradiation unevenness of the light source and the objective lens 11b
Due to the deterioration characteristics of the peripheral light amount of the
However, in many cases, the envelope signal has a gentle mountain shape in which the level from the central portion in the main scanning direction is large and the level becomes smaller as the distance from the central portion increases.

When the envelope signal storing operation as described above is completed, the control circuit 22 executes the envelope signal correcting operation. That is, even if the stored envelope signal includes a drop signal or a protrusion signal, the shading correction memory is corrected by correcting it so as to approximate the original envelope signal shape as shown in FIG. 4D. Store again in 19. The correction operation is executed by software as shown in FIG. 1, and its details will be described later. The address for the memory 19 during this correction operation is the control circuit 22.
Occurs, is selected by the selector 21 and is given to the memory 19.

The control circuit 22 executes the reading operation of the document after the correction operation of the envelope signal is completed. At this time, the control circuit 22 outputs the shift pulse SH and the control clock CL before outputting the analog switch 18
To select the corrected analog envelope signal from the D / A converter 17, and set the selector 21 to select the address from the counter 20. The image signal from the original read by the scanner 11 in such a state is also amplified by the amplifier 13 and then given to the A / D converter 15. At this time, since the corrected envelope signal is given as the analog reference voltage Vref to the A / D conversion unit 15, the image signal read with this envelope signal as a reference is subjected to analog / digital conversion to obtain the light source of the light source. The fluctuation of the image signal due to the irradiation unevenness and the deterioration characteristic of the peripheral light amount of the objective lens 11b is corrected, that is, converted into the image signal with the same white level as a reference and output to the image processing unit 23.

Next, the correction operation of the envelope signal (correction processing)
Will be described in detail. It should be noted that this correction operation is executed by the control circuit 22 retrieving the envelope signal from the memory 19, correcting it, and then writing it, as described above.

First, the correction principle according to this embodiment will be described with reference to FIG. In FIG. 5, Dj (j = 0, 1,
..., N, ... i + N, ...) is the shading correction memory 1
It is the digital value (pixel value) of the envelope signal before correction written in the address 9 (corresponding to the pixel order). now,
If the pixel position to be corrected is i + N, the control circuit 22
In principle, calculates the corrected digital value D'i + N by the averaging process according to the following equation.

[0049]

[Equation 1]

In practice, the digital value Di from the memory 19 is
, Di + N and Di + 2N are read, and this equation (1) is executed,
The obtained corrected digital value D'i + N is stored again in the address of the memory 19 where the digital value Di + N was stored.

Such an averaging process is performed so that corrected digital values can be obtained for at least all the pixels within the maximum readable range XR in the main scanning direction.

According to such an averaging process, as is apparent from FIG. 5, the drop signal and the projection signal portion are averaged, and the drop amount and the projection amount are small in the corrected digital envelope signal. ..

FIG. 6 shows a corrected envelope signal obtained by correcting the envelope signal shown in FIG. 4 (C). Figure 6
(A) shows the result of executing the correction calculation once by the above equation (1). The correction reduces the drop amount and the protrusion amount, but the amount is still large. Therefore,
In the case of this embodiment, the correction calculation according to the equation (1) is repeated a plurality of times. FIG. 6B shows the correction calculation by the equation (1)
FIG. 6C shows the result of repeating the correction calculation by the equation (1) three times. Like this (1)
By repeating the correction calculation by the formula a plurality of times, the shape of the original envelope signal can be approximated.

It should be noted that N that defines the pixels to be subjected to the averaging process should be selected in consideration of the correction effect, and that correct correction digital values can be obtained for all the pixels within the maximum readable range XR. It is necessary to select it. The pixels D0, D1 ... DN-1 at the left end and the N pixels at the right end are not corrected, but in general, the pixel reading device has a CCD sensor 11b.
Not using all the total pixels C0 of the above, but the pixels at both ends are cut. Therefore, it is necessary to select the value of N according to the number of cut pixels, and if it is selected in this way, no problem will occur even if there is a pixel to which the equation (1) is not applied. Further, the value of N and the number of correction calculations should be empirically determined by the amount of drop and the amount of protrusion of the reference white plate 10 due to dust or dirt.

In the above description of the principle, an example in which three pixels are averaged has been shown for the sake of simplification of description, but a larger number of pixels may be used if the number of pixels is an odd number, and it is assumed that M pixels are used. The generalized expression of equation (1) is as shown in the following equation.

[0056]

[Equation 2]

Next, the correction process by the control circuit 22 when the correction calculation shown in the equation (2) is repeated L times will be described with reference to FIG.

In FIG. 1, h is a parameter indicating the current number of iterations of the operation of equation (2), i is a parameter indicating the pixel of the youngest address to be used in the average calculation, r is a summation parameter, and m is the average calculation. , D'is an average value parameter, and each parameter h, i, r, m, d'is stored in a register inside the control circuit 22. Further, M is the number of pixels used for the average calculation, L is the number of times the calculation of formula (2) is repeated, N is the distance between the closest pixels used for the average calculation (pixel number display), C0 is the final address, These fixed data are stored in advance in the data memory in the control circuit 22. The control circuit 22 first initializes the parameters h, i, r, and m in this order (steps 50 to 53). That is, the parameters h, i, and r are each set to 0, and the parameter m is set to 1.

Next, the control circuit 22 calculates i + (m-1) N from the current parameters i and m and the constant N, and the shading correction memory 19 specified by this value i + (m-1) N. The pixel data Di + (m-1) N is read out (step 54). After that, the read pixel data Di + (m-1) N is added to the total sum parameter r, and the added value is stored in the register as a new total sum parameter r (step 5).
5). Next, after incrementing the parameter m by 1,
After confirming that the parameter m is smaller than the fixed value M + 1, the process returns to the above step 54 (steps 56 and 57).

The processing loop consisting of steps 54 to 57 is repeated until the parameter m becomes equal to the fixed value M and a positive result is obtained in step 57. Step 5
When a positive result is obtained in 7, the content of the summation parameter r has a total of M pixels i, i + N, ..., I + (M−1), in which the pixel i is the first pixel and N pixels are different thereafter.
The sum total of N pixel data is ΣDi + (m-1) N.

When the calculation of the summation part to be used for the averaging is completed in this way, the value i + (M-1) N and the value C0 indicating the final pixel are compared in magnitude, and the current calculation by the equation (2) is performed. It is confirmed that one cycle has not been completed for each pixel (step 58).

Then, an average value is obtained by dividing the total sum parameter r by a fixed value M, and this average value r / M is stored in the register of the average value parameter d '(step 5
9). After that, the average value parameter d ′ is set to the value i +
(M-1) Write to the address of the shading correction memory 19 designated by N / 2, that is, the address of the pixel at the intermediate position of the plurality of pixels used for the averaging process (step 60). When the process of fixing the parameter i in steps 52 to 60 is completed in this way, the parameter i is incremented by 1 and the next pixel is set as the start pixel of the average calculation, and the process returns to step 52 described above (step 61).

By repeatedly executing the processing loop consisting of steps 52 to 61, the correction processing for the envelope signal in the memory 19 is sequentially performed for each pixel, and eventually the (h + 1) th correction calculation by the equation (2) is performed. Upon completion, a positive result is obtained at step 58.

At this time, after the parameter h is incremented by 1 and stored again in the register, it is confirmed that the calculation by the equation (2) has not been repeated L times, which is a predetermined number of times, and the process returns to the above-mentioned step 51 to proceed to the next step. The correction calculation of the number of times is executed (steps 62 and 63).

When the calculation by the equation (2) is repeated L times which is a predetermined number of times, a positive result is obtained in step 63, and the series of envelope signal correction processing ends.

When executing the processing shown in FIG. 1, it is necessary to prepare a shading correction memory 19 capable of storing envelope signals for two lines in the main scanning direction. Then, the read envelope signal from the reference white background 10 and the envelope signal for which the correction has been executed an even number of times and the envelope signal for which the correction has been executed an odd number are made different in storage area so that the corrected pixel values are the same. It is necessary to prevent it from being used for the calculation of the number of times.

According to the above-described embodiment, since the correction is performed by the averaging process of a plurality of pixel values and the averaging process is repeated, it is possible to reliably correct the dip and the protrusion of the envelope signal.

Further, the number of times of reading the reference white background 10 required to obtain the final corrected envelope signal may be one (1 line), and the area of the reference white background 10 may be smaller than in the conventional case.

According to the above-described embodiment, the reading operation of the reference white background 10 which takes the longest time may be performed for one line, and since the correction of the envelope signal is executed by software, the processing is performed as compared with the conventional method. The time can be shortened.
Further, the hardware configuration for correcting the envelope signal is reduced, and the entire device can be downsized.

Next, a second embodiment of the present invention will be described with reference to FIGS. 7 and 8.

In the second embodiment, the control circuit 22 stores the correction processing shown in the flowcharts of FIGS. 7 and 8. Also in this second embodiment, the correction calculation shown in the equation (2) is repeated L times.

In this embodiment, each parameter is almost the same as that in the first embodiment. Outline of each of these parameters, h is a parameter indicating the current number of iterations of the calculation of equation (2), i is a parameter indicating the pixel of the youngest address to be used in the average calculation, and m is the number of the pixel used in the average calculation. Is a parameter of the total sum Σd (k), k is a count parameter for calculating the total sum Σd (k) of these parameters, and d ′ is an average value parameter. Further, in this embodiment, d is used as a parameter.
(1), d (2), ... d (M−1), d (M) are used. These d (1), d (2), ... d (M-1), d
(M) is the pixel data Di, Di + N, ... Di + (M-
2) M parameters for storing N and Di + (M-1) N, respectively. Then, each of the above parameters is stored in a register inside the control circuit 22.

Further, M is the number of pixels used for averaging, and L is
Equation (2) is the number of iterations, N is the distance of the closest number of pixels used for averaging (pixel number display), C0 is the final address, and these fixed data are also controlled as in the first embodiment. It is stored in advance in the data memory in the circuit 22.

In FIGS. 7 and 8, the control circuit 22 first initializes the parameters h, i and m in this order (steps 70 to 72). That is, the parameters h and i
To 0 and the parameter m to 1.

Next, the control circuit 22 calculates i + (m-1) N from the current parameters i and m and the constant N, and the shading correction memory 19 designated by this value i + (m-1) N. The pixel data Di + (m-1) N is read (step 73). Then, the read pixel data Di + (m-1) N is set to the parameter d (m) (step 74). Next, the parameter m is incremented by 1 (step 75), and the parameter m is a fixed value M.
After confirming that it is smaller (step 76), the process returns to step 73 described above. The processing loop including steps 73 to 76 is repeated until the parameter m becomes equal to the fixed value M and a positive result is obtained in step 76. When a positive result is obtained in step 76, (M-
1) number of parameters d (1), d (2), ... d (M-
1) is set, and the parameter m becomes the fixed value M.

Next, the control circuit 22 controls the pixel data D of the shading correction memory 19 designated by i + (m-1) N.
i + (m-1) N is read (step 77). Here, the parameter x is replaced with m (step 78). Since the current parameter m is m = M, the read pixel data is Di + (M-1) N, and this pixel data is set in the parameter d (M) (step 7).
9). Up to this point, the parameters d (1), d (2), ...
All d (M) are set.

Next, the parameters k and r are initialized to 0 (steps 80 and 81). Further, the parameter k is incremented by 1 (step 82), the parameter d (k) storing each pixel data is added to the total sum parameter r, and the added value is added to the new total sum parameter r.
And store in the register (step 83). Next, it is confirmed that the parameter k is smaller than the fixed value M (step 84), and the process returns to step 82 described above. The processing loop including steps 82 to 84 is repeated until the parameter k becomes equal to the fixed value M and a positive result is obtained in step 84. At the time point when a positive result is obtained in step 84, the content of the summation parameter r is such that the pixel i is the first pixel, and the number of pixels is different by N pixels after that.
+ N, ..., i + (M−1) N sum of pixel data ΣDi
It is + (m-1) N.

When the calculation of the summation part to be used for the average calculation is completed in this way, the value i + (m-1) N and the value C0 indicating the final pixel are compared in magnitude, and the calculation of each pixel separated by N pixels is performed. It is confirmed that is not completed to the end (step 85). Next, the sum total parameter r is divided by a fixed value M to obtain an average value, and this average value r / M is stored in the register of the average value parameter d ′ (step 8).
6). Then, the average value parameter d ′ is set to the value i + (m
-1) Shading correction memory 1 specified by N / 2
The data is written in the address of 9, that is, the address of the pixel at the intermediate position of the plurality of pixels used for the averaging process (step 87). In this way, when the process of fixing the parameters i and m in steps 77 to 87 ends, the parameter k is initialized to 0 (step 88). Next, the parameter k is incremented by 1 (step 89), and the parameter d (k +
The pixel data stored in the register of 1) is moved to the register of the previous parameter d (k) (step 9).
0). After confirming that k is not M-1 (step 91), the process returns to step 89. And step 8
In the processing loop consisting of 9 to 91, the parameter k becomes 1 less than the fixed value M, that is, equal to M−1.
Repeat until a positive result is obtained with. When a positive result is obtained in step 91, the parameter d
(1), d (2), ... D (M-1) have pixel data Di
+ N, Di + 2N, ..., Di + (M−1) N are stored. When the process of fixing the parameter m in steps 77 to 91 is completed in this way, the parameter m is incremented by 1, that is, the parameter m is set to M + 1 (step 92), and the parameter x is set to x−1, that is, M (step 93) and returns to step 77 above. Then, in step 77, the pixel data Di + of the address i + (M) N
(M) N is read and the parameter x is replaced with m (step 78). That is, d (x) is set to d (M), and Di + (M) N is stored in d (M) in step 79. As a result, all the parameters d (1), d (2), ... D (M) are set.

Then, by repeatedly executing the processing loop consisting of steps 77 to 93, the correction processing for the envelope signal in the memory 19 is sequentially performed for each pixel separated by N pixels, and when the processing is completed to the end. At step 85, a positive result is obtained. At this time, the parameter i is incremented by 1 (step 94). Further, N times is repeated and it is confirmed that i = N is not satisfied (step 95), and the next adjacent pixel is set as the start pixel of the average calculation, and the process returns to step 72 described above. Then, the processing of steps 72 to 95 is repeated, and step 95
When an affirmative result is obtained in (2), the (N + 1) th correction calculation of all pixels for the envelope signal in the memory 19 by the equation (2) ends. At this time, the parameter h is incremented by 1 and stored again in the register (step 96), and it is confirmed that the calculation by the equation (2) is not repeated L times, which is a predetermined number of times (step 97), and the above step 71 is performed. After returning, the correction calculation of the next time is executed. If the calculation by the equation (2) is repeated L times which is a predetermined number of times, step 9
An affirmative result is obtained at 7, and the series of envelope signal correction processing ends.

As a result, the same operation and effect as those of the first embodiment can be obtained. Further, in this embodiment, step 89
Since the pixel data read once once is used as it is and the reading process of the same pixel data is omitted, the processing speed can be increased.

In the first embodiment described above, the averaging process is repeated to correct the envelope signal. However, since repeating the averaging process is equivalent to the weighted averaging process, The envelope signal may be corrected by performing the weighted averaging process once. However, the method of the above embodiment has the advantages that the software can be changed easily and the required capacity of the program memory can be reduced, as compared with one-time weighted averaging processing.

In the first embodiment described above, the number of pixels subjected to the averaging process is odd, but the number of pixels subjected to the averaging process may be even. In this case, it is necessary to select, as the plurality of pixels to be subjected to the averaging process, pixels at positions symmetrical with respect to the pixel position after correction without including those at the pixel position after correction.

Furthermore, in the first and second embodiments,
An empty area of the document may be used as the reference white background 10. In this case, unlike the conventional case, since the reading line is one line, the document surface can be effectively used. The reference area is not limited to the white background as in the above embodiment.

[0084]

As described above, according to the present invention, the envelope signal obtained by reading the reference area is corrected by the weighted average of the plurality of pixels. However, even if noise is mixed in the read envelope signal, a good corrected envelope signal can be obtained, and the overall structure can be downsized.

[Brief description of drawings]

FIG. 1 is a flowchart showing an envelope signal correction method for an image reading apparatus according to a first embodiment of the present invention.

FIG. 2 is a block diagram showing a conventional envelope signal correction circuit.

FIG. 3 is a block diagram showing an image reading apparatus according to the first embodiment of the present invention.

FIG. 4 is a signal waveform diagram for explaining the operation of the image reading apparatus in FIG.

FIG. 5 is an explanatory diagram showing a correction principle according to the first embodiment.

FIG. 6 is an explanatory diagram showing a relationship between the number of corrections and a corrected envelope signal according to the first embodiment.

FIG. 7 is a flowchart (No. 1) showing an envelope signal correction method for the image reading apparatus according to the second embodiment of the present invention.

FIG. 8 is a flowchart (No. 2) showing the envelope signal correction method for the image reading apparatus according to the second embodiment of the present invention.

[Explanation of symbols]

 10 ... Standard white background 11 Scanner 14 Normalization circuit 19 Shading correction memory 22 Control circuit.

Claims (2)

[Claims] 1. An image signal obtained by reading a reference area of the same color in the main scanning direction is stored as an envelope signal, and then the envelope signal is reduced in level due to dust or dirt in the reference area. An image reading apparatus that corrects so as to eliminate level fluctuations due to external noise and corrects an image signal obtained by reading a document based on the corrected envelope signal is trying to obtain the value of the corrected envelope signal. The values of a plurality of pixels of the envelope signal, which are separated from each other by a predetermined number of pixels around the pixel of current interest, are taken out, and these pixel values are weighted and averaged to correct the current objective of correction. An envelope signal correction method for an image reading apparatus, wherein a pixel value of a post-envelope signal is obtained, and a post-correction envelope signal is obtained from the envelope signal. 2. A plurality of pixels of the envelope signal separated by a predetermined number of pixels are taken out, these pixel values are averaged to obtain a value of a pixel located in the middle of the plurality of pixels, and such processing is performed. The weighted averaging process is performed by performing the first correction on the envelope signal by executing it on the entire range of the envelope signal and repeatedly executing the same method as the first correction method. 2. The envelope signal correction method for an image reading apparatus according to claim 1, wherein the final corrected envelope signal is obtained.