JPH1127525A - Image reader - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image reader by which correcting precision of shading is improved. SOLUTION: An image signal S4 of a form 1 is outputted from an image sensor 4, amplified by an amplifier 5, turned into an image signal S5, a black level of the image signal S5 is clamped to zero V by a black level clamping circuit 6, and then an image signal S6 is outputted. The image signal S6 is converted into an image signal S8 by defining a peak value 7 of the image signal S6 as a reference voltage by an A/D converter 8 and transmitted to correcting operation table memory 10. A corrected value S11 corresponding to an A/D converted pixel is transmitted from a corrected value memory 11 to the correcting operation table memory 10 based on an address S12. Since the results of correcting operation corresponding to the corrected value S11 and the image signal S8 are written in the correcting operation table memory 10, a corrected image signal S10 is outputted, even if the image signal S8 whose level is lowered is inputted.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image reading apparatus in which the accuracy of shading correction when optically reading a form image is improved.

[0002]

2. Description of the Related Art In a conventional image reading apparatus, a blank sheet is read in advance in order to correct shading caused by a decrease in the amount of light around a lens of an image reading section for optically reading an image of a form or uneven illumination of a light source. For example, a white level signal output from an image sensor such as a CCD (Charge Coupled Device) image sensor is converted into an analog / digital signal by a fixed reference voltage set to a maximum signal level (hereinafter, referred to as A / D conversion). Then, the digital white level signal is used as a correction value in a RAM (Randam Acc.
ess Memory). At the time of normal reading, shading is corrected by digitally calculating a digital image signal of a form with this correction value (that is, the measured white level). For example, A /
Assuming that the number of quantization bits for the D conversion is 8, the measured white level is W, the level of the image signal at the time of reading is S, and the level of the image signal after correction is H, H = S × 2 ⁸ / W expressed.

[0003]

However, the image signal output from the image pickup device is A / A at a fixed reference voltage.
In a conventional image reading apparatus that performs D conversion, the reference voltage is fixed even when the output level of the image sensor decreases due to a decrease in the amount of light, such as when the light source has deteriorated or the form has a low reflectance. , The conversion accuracy of the A / D conversion is reduced. For this reason, there has been a problem that the accuracy of correction for shading is reduced.

[0004]

According to a first aspect of the present invention, there is provided an image reading apparatus for optically reading an image of a form to generate an image reading signal. Such means are provided. An image reading unit that scans the form to generate an analog first image signal; an amplification circuit that amplifies the first image signal at a predetermined amplification to generate a second image signal; A black level clamp circuit that generates a third image signal by shifting the level so that the black level of the waveform of the second image signal is clamped to a fixed voltage;
A peak detection circuit for detecting and holding a peak value of the third image signal, and converting the third image signal to A using the voltage of the peak value with respect to the fixed voltage as a reference voltage.
An A / D converter for generating a digital fourth image signal by performing a D / D conversion, and converting the fourth image signal generated by using a blank white sheet for measurement in place of the form for each pixel of the image reading unit. And a correction value memory for storing the correction value corresponding to each pixel when reading the form, and the fourth image corresponding to the form generated by the A / D converter. Correction operation means for performing a correction operation on the signal level using the correction value output from the correction value memory, and outputting the result of the correction operation as the image reading signal.

According to the first aspect of the present invention, since the image reading apparatus is configured as described above, first, a fourth image signal generated by using a blank measuring sheet as a form is used for each pixel of the image reading unit. The correction value is stored in the correction value memory as a correction value. Thereafter, the form is scanned by the image reading unit to generate a first image signal, and the first image signal is amplified by the amplifier circuit to become a second image signal. The black level of the second image signal is clamped to a fixed voltage by the black level clamp circuit to become the third image signal, and the peak value is detected and held by the peak detection circuit. The third image signal is A / D-converted by an A / D converter using the voltage of the peak value with respect to the fixed voltage as a reference voltage to become a fourth image signal. The fourth image signal is corrected by a correction value corresponding to each pixel by a correction calculation unit, and is output as an image reading signal. According to a second aspect of the present invention, an image reading apparatus for optically reading an image of a form and generating an image reading signal includes the following means.

According to the first aspect of the present invention, there is provided an image reading section, an amplifier circuit, a black level clamp circuit, a peak detection circuit, an A / D converter, a correction value memory and a correction operation means, and a white level measurement area on the upper side as the form. A white level detection and holding circuit for detecting and holding a white level signal of the form from the image reading signal generated using the generated form, and reading the image using the white level signal held by the white level detection and holding circuit Normalization means for normalizing the signal to a preset reference level. According to the second aspect, the white level is detected and held by the white level detection and holding circuit of the image reading signal output from the correction calculation means using the form having the white level measurement area provided on the upper side.
This image reading signal is based on the white level signal held in the white level detection holding circuit by the normalizing means,
It is normalized to a preset reference level. Therefore, the above problem can be solved.

[0007]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS First Embodiment FIG. 1 is a block diagram of an image reading apparatus showing a first embodiment of the present invention. This image reading apparatus has a light source 2 for irradiating a form 1, and an image of the form 1 is read by an image reading unit (for example, an image sensor) 4 via an optical system 3 such as a lens. The image sensor 4 is configured by, for example, a CCD or the like, and has a function of scanning the form 1 and generating an analog first image signal S4. An output terminal of the image sensor 4 is connected to an amplifier circuit (indicated by “amplifier” in FIG. 1) 5 that amplifies the image signal S4 to a predetermined level and generates a second image signal S5. I have. The output terminal of the amplifier circuit 5 is connected to the black level clamp circuit 6. The black level clamp circuit 6 is a circuit composed of, for example, an analog switch, a capacitor, and the like, and is a circuit that performs a level shift so that the black level of the image signal S5 is clamped to a fixed voltage of 0 V, for example. Therefore, when the image signal S5 passes through the black level clamp circuit 6, the image signal S5
Is output as an image signal S6 having a minimum value of 0V. The output terminal of the black level clamp circuit 6 is connected to the input terminal of the peak detection circuit 7 and the analog signal input terminal Vin of the A / D converter (indicated by “ADC” in FIG. 1) 8. .

[0008] The peak detection circuit 7 is composed of an analog switch, a hold capacitor and the like, and is a circuit for detecting and holding the maximum level (ie, peak value) S7 of the image signal S6. The output terminal of the peak detection circuit 7 is connected to the + side reference voltage input terminal + Ref of the A / D converter 8, and the − side reference voltage input terminal -Ref of the A / D converter 8 is connected to the ground (0V). It is connected. The A / D converter 8 is
This circuit generates an image signal S8 by A / D-converting the image signal S6 with a quantization bit number of, for example, 8 bits using the peak value S7 as a reference voltage on the positive side and 0V as a reference voltage on the negative side. An output terminal of the A / D converter 8 is connected to an image signal input terminal Iin of a central processing unit (hereinafter, referred to as a CPU) 9 and a correction operation unit (for example, a correction operation table memory) 10. The CPU 9 has a function of calculating a correction value S9 for each pixel of the image reading unit 4 from an image signal S8 generated using blank white paper as the form 1. The output terminal of the CPU 9 is connected to the correction data writing input terminal Rin of the correction value memory 11, and the address output terminal of the address control circuit 12 is connected to the address input terminal Ain of the correction value memory 11. The correction value memory 11 stores an address S12 corresponding to each pixel of the correction value memory 11 output from the address control circuit 12.
Has the function of storing the correction value S9 based on the An output terminal of the correction value memory 11 is connected to a correction value input terminal Tin of the correction operation table memory 10, and an output terminal of the correction operation table memory 10 is connected to an image memory (not shown). The correction calculation table memory 10 uses the correction value S11 output from the correction value memory 11 for the level of the image signal S8 generated by the A / D converter 8 when the image reading unit 4 reads the form 1. A correction operation is performed, and the result of the correction operation is used as an image reading signal S
10 is provided. The CCD control circuit 13 has a function of controlling the scanning of the form 1 by the image sensor 4.

FIG. 2 is a diagram showing an example of a shading state, in which the ordinate represents image plane illuminance (CCD output), and the abscissa represents CCD pixels. FIG. 3 is a diagram showing a waveform of an image signal S6 which is an output signal of the black level clamp circuit 6, wherein the vertical axis represents the image signal S6 and the horizontal axis represents C.
The pixels of the CD are taken. In FIG. 3, the waveform A is
FIG. 5 shows a waveform of the image signal S6 before the light source 2 is deteriorated;
The maximum value is the peak value S6a. The waveform B indicates the waveform of the image signal S6 in which the light source 2 has deteriorated and the level has decreased, and the maximum value is the peak value S6b. FIG.
7A and 7B are diagrams showing examples of correction values, and FIG. 7A shows an image signal S8 when a measurement blank sheet is read.
The vertical axis indicates the level of the image signal S8, and the horizontal axis indicates CCD pixels. FIG. 4B shows the respective values of FIG. FIGS. 5 (a) and 5 (b)
FIG. 2 is a diagram showing an example of a correction operation table memory 10 in FIG. 1. FIG. 1A shows the configuration of the correction operation table memory 10, and FIG.
The equation for calculating the image reading signal S10 output from the device is shown.

FIGS. 6 (a) and 6 (b) are diagrams showing a state of correction for shading. FIG. 6 (a) shows an image signal S8, and the vertical axis represents the level of the image signal S8 and the horizontal axis. The axis is the CCD pixel. This figure 6
In (a), the waveform A shows the waveform of the image signal S8 before the light source 2 deteriorates, and the maximum value is the peak value 255. The waveform B shows the waveform of the image signal S8 in which the light source 2 has deteriorated and the level has decreased, and the maximum value is the peak value 20
It is 0. FIG. 6B shows an image reading signal S10 obtained by correcting the image signal S8. The vertical axis represents the level of the image reading signal S10, and the horizontal axis represents CCD pixels. Next, the operation of FIG. 1 will be described with reference to FIGS. The form 1 is illuminated by the light source 2, and the image of the form 1 is read by the image sensor 4 via the optical system 3. However, the light distribution characteristics of the light source 2 are not uniform, and the amount of light around the optical system 3 such as a lens. Due to the decrease or the like, the brightness of the image of the form 1 formed on the image sensor 4 does not become flat as shown in FIG. 2 even if the reflectance of light on the surface of the form 1 is uniform. This is shading.

The image reading apparatus according to the present embodiment has a function of electrically correcting the shading. The image reading apparatus has the following two modes: (1) a correction value measuring mode and (2) a normal reading mode. Perform mode operation. (1) In the correction value measurement mode, the data of the image signal S8 when the measurement white paper is read is taken into the CPU 9, the correction value S9 for shading is calculated, and the correction value memory 1 is stored.
This is a mode in which the data is stored in 1. (2) The normal reading mode is a mode in which the read image of the form 1 is corrected using the correction value S11 output from the correction value memory 11. (1) Correction value measurement mode First, set a clean white sheet for measurement at the reading position,
The image is scanned by the image sensor 4 to read the image of the blank sheet for measurement. The image signal S4 output from the image sensor 4 becomes an image signal S5 input to the amplifier 5 and amplified to a predetermined voltage level.
5 is sent to the black level clamp circuit 6. The black level clamp circuit 6 generates the image signal S6 by clamping the black level of the image signal S5 to, for example, 0V. This image signal S6 is input to the input terminal Vin of the A / D converter 8.
And to the peak detection circuit 7. Peak detection circuit 7
Is the peak value S6 of the image signal S6 as shown in FIG.
a and S6b are detected and held as the peak value S7. This peak value S7 is equal to the reference input terminal of the A / D converter 8 + R
ef. Also, a black level clamp voltage (for example, 0 V) is applied to the − side reference input terminal −Ref of the A / D converter 8.
V) is applied. The image signal S6 clamped to the black level is converted by the A / D converter 8 into an 8-bit digital image signal S8 using the voltage of the peak value S7 with respect to 0 V as a reference voltage, and transmitted to the CPU 9.
The CPU 9 calculates a correction value S9 from the image signal S8 of the white level for each pixel as shown in FIG. 4, and stores the correction value S9 at an address corresponding to each pixel in the correction value memory 11. (2) Normal reading mode The image signal S4 of the form 1 read by the image sensor 4 becomes an image signal S5 inputted to the amplifier circuit 5 and amplified to a predetermined voltage level, and the image signal S5 becomes black. The signal is sent to the level clamp circuit 6 and the black level is clamped to 0V. The image signal S6 output from the black level clamp circuit 6 is, as in (1) the correction value measurement mode, an 8-bit signal with the A / D converter 8 using the voltage corresponding to 0V of the peak value S7 of the image signal S6 as a reference voltage. And sent to the image signal input terminal Iin of the correction operation table memory 10.

On the other hand, based on the address S12, the A /
The correction value S11 corresponding to the D-converted pixel is read out from the correction value memory 11, and the correction operation table memory 1
It is sent to the zero correction value input terminal Tin. As shown in FIG. 5, the correction operation table memory 10 stores 8 bits (0 to 0).
255) and the 8-bit image signal S8
Correction results for all combinations of the levels are previously written as a table in a memory cell whose address is the correction value S11 and the level of the image signal S8.
Even when the image signal S8 whose level has decreased as shown by the waveform B in FIG. 6A is output, the shading correction as shown in FIG. An image signal S10 is output.
As described above, in the first embodiment, the analog image signal S6 is converted to the peak value S7 of the image signal S6.
A / V is defined as a voltage with respect to the black level (0 V)
Since the D conversion is performed, the white level at that time is always full even when the light amount is reduced and the white level is reduced due to the light source 2 being deteriorated or the surface of the form 1 having a low reflectance. Since the A / D conversion is performed as a scale, the accuracy of the A / D conversion does not decrease, and the accuracy of correction for shading is improved.

Second Embodiment FIG. 7 is a block diagram of an image reading apparatus according to a second embodiment of the present invention, and is common to the elements in FIG. 1 showing the first embodiment and common elements. Are given. The first embodiment relates to an apparatus for reading a form with a fixed width, but the second embodiment relates to an apparatus for reading a form with a narrow width. In this image reading apparatus, a white level detection and holding circuit 13 and a normalizing means (for example, a normalizing table memory) 14 are provided in the image reading apparatus of FIG. The output terminal of the address control circuit 12 is connected to the address input terminal Ain of the correction value memory 11 and to the address input terminal Ain of the white level detection and holding circuit 13. The output terminal of the correction operation table memory 10 is connected to the image signal input terminal Iin of the white level detection and holding circuit 13 and the image signal input terminal Iin of the normalization table memory 14. An output terminal of the white level detection and holding circuit 13 is connected to an address input terminal Ain of the normalization table memory 14. The white level detection and holding circuit 13 is a circuit configured by, for example, a memory or the like, and detects and holds the white level signal S13 in the white area of the form 1. The white level signal S13 held by the white level detection and holding circuit 13 may be held for each pixel, or may be represented by the white level of a specific pixel. The white level signal S13 may be any value that can evaluate the white level of the form 1, such as the average value of a plurality of lines or the maximum value in a plurality of lines. The normalization table memory 14 stores the white level signal S13 held in the white level detection and holding circuit 13.
Has a function of generating a fifth image signal in which the image reading signal S10 is normalized to an absolute white level set in advance. The output side of the normalization table memory 14 is connected to an image memory (not shown). Other configurations are the same as those in FIG.

FIG. 8 is a diagram showing an example of a correction value. In FIG. 8, an image signal S8 when a blank measuring sheet is read is shown.
Is represented by a waveform A, the vertical axis represents the level of the image signal S8,
And the horizontal axis shows the pixels of the CCD. FIG. 9 is a diagram showing an image reading signal S10 obtained by correcting the image signal S8 for shading, in which the vertical axis represents the level of the image reading signal S10 and the horizontal axis represents CCD pixels. FIG. 10 is a diagram illustrating an image signal S8 when a blank white sheet for measurement and a form having a smaller width than the white blank sheet for measurement are read. In FIG. 10, a waveform A indicated by a broken line indicates an image signal S8 when a blank sheet for measurement is read, and a waveform B indicated by a solid line indicates an image signal S8 obtained when the document 1 is read. FIG. 11 shows an image reading signal S obtained by correcting the image signal S8 of the waveform B in FIG. 10 for shading.
10, FIG. 12 shows a form provided with a white level measurement area W, FIG. 13 shows a white level signal S13 corrected for shading, and FIG. 14 shows a normalized image reading signal S10. FIG. 4 is a diagram showing an image signal S14. Next, referring to FIGS. 8 to 14,
The operation of FIG. 7 will be described.

As in the first embodiment, (1) the correction value measurement mode is executed, the CPU 9 calculates the correction value S9 from the white level of each pixel, and the address corresponding to each pixel in the correction value memory 11. The correction value as shown in FIG. Further, as in the first embodiment, (2) the normal reading mode is executed, and an image signal S10 corrected for shading as shown in FIG. 9 is output from the output terminal of the correction operation table memory 10. Is done. For example, as shown in FIG. 8, the white level is 200 (that is, the correction value is 200).
Is corrected by the CPU 9 by 255/200 = 1.275 times.
Therefore, for the form 1 whose white level is represented by the waveform A, the level of the image signal S8 of the pixel whose white level is 200 is 200 × 1.275 = 255, and the form 1 as a whole is shown in FIG. Thus, the image reading signal S10 whose level is corrected to 255 is obtained.

When the form 1 having a narrow width as shown in FIG. 10 is read, if the correction value S9 of the image signal S8 of the pixel having the peak value of the form 1 is 200, the image signal of the form 1 is read. Since the A / D conversion of S6 is performed using the level of the white peak as the reference voltage, the level of the image signal S8 after the A / D conversion of the image signal S6 of the peak pixel becomes 255 instead of 200. Therefore, when the image signal S8 is corrected for shading, as shown in FIG.
10 is corrected to a level of 255 × 1.275 = 325. As described above, when the width of the form 1 to be read changes, the position of the pixel having the peak value of the form 1 changes, so that the corrected level varies depending on the width of the form 1.
Absolute standard of white level cannot be taken. However, if the peak pixel at the time of the correction for the shading is within the form, the corrected level does not change even if the width of the form 1 changes. In order to eliminate the fluctuation of the white level, as shown in FIG. 12, a white level measurement area W is provided on the upper side of the form 1, the white level value is measured in this area W, and the white level is measured using this value. May be normalized to, for example, 255.

An image reading signal S10 output from the correction operation table memory 10 when the form 1 shown in FIG. 12 is read is sent to the white level detection holding circuit 13 and the normalization table memory 14. Normalization table memory 1
4, data D calculated using the following equation (1) is written in advance. D = 255 × IM / WH (1) where, WH: white level of the form IM; level of the image signal S10 The normalization table memory 14 outputs the corrected image signal S10 output from the correction operation table memory 10. And the form 1 output from the white level detection and holding circuit 13
Data D shown in equation (1) using the white level of
And outputs the image signal S10 as a normalized image signal S14.

If the white level detection and holding circuit 13 holds the white level of each pixel, the white level detection and holding circuit 13 has a dynamic shading waveform for each form 1 to be read. Is held. Therefore, even if the shading waveform in the correction value measurement mode is different from the shading waveform due to uneven light amount deterioration of the light source 2, the shading is further corrected by the white level S13 held in the white level detection holding circuit 13. Correction for shading is performed stably even with time.
As described above, in the second embodiment, the white level of the image reading signal S10 is normalized by using the white level signal S13 in the normalization table memory 14, so that the advantages of the first embodiment are obtained. In addition, even if the width of the form 1 changes or the light source 2 changes with time, the shading can be corrected without changing the white level. Note that the present invention is not limited to the above embodiment, and various modifications are possible.
For example, there are the following modifications.

(A) The correction value in FIG. 4 may use the white level of each pixel in one line. However, in order to reduce an error due to noise, an average of image data of a plurality of lines for each pixel is used. May be used. (B) The correction calculation table memory 10 may be replaced with a calculation circuit that calculates the image reading signal S10. (C) The normalization table memory 14 stores the image signal S
14 may be replaced with an arithmetic circuit for calculating the value of.

[0020]

As described in detail above, according to the first aspect, the analog third image signal is converted to the A / A voltage using the voltage with respect to the black level of the peak value of the third image signal as the reference voltage. Since the D-conversion is performed, even when the white level decreases, the A / D conversion is always performed with the white level at that time being a full scale, so that the accuracy of the A / D conversion does not decrease and the accuracy of the correction for shading is reduced. Can be improved. According to the second invention, the white level of the image reading signal is normalized by the normalizing means using the white level signal. Therefore, in addition to the effect of the first invention, the width of the form changes. Even if the light source changes with time, shading can be corrected without changing the white level.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of an image reading device according to a first embodiment of the present invention.

FIG. 2 is a diagram illustrating a state of shading.

FIG. 3 is a diagram showing a waveform of an image signal S6.

FIG. 4 is a diagram illustrating an example of a correction value.

FIG. 5 is a diagram showing a correction operation table memory 10 in FIG. 1;

FIG. 6 is a diagram illustrating a state of correction for shading.

FIG. 7 is a configuration diagram of an image reading device according to a second embodiment of the present invention.

FIG. 8 is a diagram illustrating an example of a correction value.

FIG. 9 is a diagram showing an image reading signal S10.

FIG. 10 is a diagram showing an image signal S8.

FIG. 11 is a diagram showing an image reading signal S10.

FIG. 12 is a diagram showing a form provided with a white level measurement area W;

FIG. 13 is a diagram showing a white level signal S13.

FIG. 14 is a diagram showing an image signal S14.

[Explanation of symbols]

4 Image reading unit 5 Amplifying circuit 6 Black level clamp circuit 7 Peak detection circuit 8 A / D converter (analog /
Digital converter) 10 Correction calculation table memory (correction calculation means) 11 Correction value memory 13 White level detection holding circuit 14 Normalization table memory (Normalization means)

Claims (2)

[Claims] 1. An image reading apparatus for optically reading an image of a form to generate an image read signal, comprising: an image reading section for scanning the form to generate an analog first image signal; The image signal is amplified with a predetermined amplification degree and the second
An amplifier circuit that generates an image signal of the following; a black level clamp circuit that generates a third image signal by performing a level shift so that a black level of a waveform of the second image signal is clamped to a fixed voltage; A peak detection circuit for detecting and holding a peak value of a third image signal; and a fourth digital signal obtained by converting the third image signal from analog to digital using the voltage of the peak value with respect to the fixed voltage as a reference voltage. An analog / digital converter that generates an image signal of the above, and the fourth image signal generated using a blank sheet for measurement instead of the form is stored in advance as a correction value for each pixel of the image reading unit, A correction value memory for outputting the correction value corresponding to each pixel at the time of reading a form; and a correction value memory generated by the analog / digital converter. Correction operation means for performing a correction operation on the level of the fourth image signal corresponding to the form using the correction value output from the correction value memory, and outputting the correction operation result as the image reading signal; An image reading device, comprising: 2. An image reading apparatus for optically reading an image of a form to generate an image reading signal, wherein the image reading section, an amplifier circuit, a black level clamp circuit, a peak detection circuit, and an analog / digital conversion are provided. Detector, a correction value memory, and a correction operation means, and a white level detection and holding circuit for detecting and holding a white level signal of the form from the image reading signal generated by using the form having a white level measurement area on the upper side as the form An image reading apparatus, comprising: a circuit; and normalizing means for normalizing the image reading signal to a preset reference level using the white level signal held in the white level detection holding circuit.