JP2677208B2 - Image sensor output level correction device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image sensor for use in a facsimile or the like, which optically reads a document such as characters and patterns and outputs it as an electric signal, and more particularly to a device for correcting its output level.

[0002]

2. Description of the Related Art An image sensor such as a facsimile machine scans a document with a CdS photoelectric conversion element to read black and white light and shade of the document as an electric signal. FIG.
Is a waveform that shows the output voltage level of the image sensor when reading a grayscale original (some grays with different densities are arranged so that the densities are visually equidistant) with this type of image sensor. It is a figure. As described above, in the actual output voltage level of the image sensor, the output level of the black original increases by about 20% as shown by the broken line d, compared with the ideal output characteristic c shown by the solid line. Black level floating e occurs.

It is known that this black level floating is caused by stray light generated mechanically and optically. Figure 9 is Cd
It is a schematic sectional drawing of the reading part of the contact image sensor using the S photoelectric conversion element. A support glass 2 is installed on a base 1, and a sensor glass 3 and a space glass 4 are arranged on the support glass 2. A light-shielding layer 6 for regulating incident light from an LED array 5 serving as an illumination light source is provided on the sensor glass 3, and the CdS photoelectric conversion elements 7 are arranged on the light-shielding layer 6 at a density of 8 pieces / mm. They are arranged in the main scanning direction. A window 6a for entering light is provided so as to form a pair with the CdS photoelectric conversion element 7. Furthermore, a thin glass plate 8 is coated with UV resin 9 as a wear-resistant layer.
It is adhered by. The original A is conveyed while being pressed by the platen roller 10 on the thin glass plate 8.

From the LED array 5 to the CdS photoelectric conversion element 7
The incoming light is mainly shown by the solid line L1, and if there is no incident light other than this, the CdS photoelectric conversion element 7 can obtain an appropriate output according to the shading of the document. However, in reality, the light that is once reflected by the base 1 and then enters (broken line L2), the light that enters between the sensor glass 3 and the spacer glass 4 (broken line L3), and the light-shielding layer 6 leaks. There is stray light that is mechanically generated such as incident light (broken line L4) and other incident light (not shown). Therefore, due to this stray light, an output is generated from the CdS photoelectric conversion element even in the case of a black original, and the above-mentioned black level floating is generated.

The intensity of this stray light is due to variations in mechanical precision,
For example, the non-linearity of the LED array, the bending of the base, the unevenness of the cut state of the glass, the unevenness of the thickness of the UV resin, etc. are highly dependent on the arrangement direction of the CdS photoelectric conversion elements and the LED array. Will also be uneven.

In order to solve the above problem, Japanese Patent Laid-Open No. 4-160
Japanese Patent No. 870 proposes an output level correction device for an image sensor shown in FIG. The output level correction device is configured to output the output of the contact image sensor 11 driven by the timing circuit 12 to the image signal processing circuit 30 on the facsimile side via the differential amplifier 13. The image signal processing circuit 30 includes a binarization circuit, a halftone processing circuit, and the like. In addition, the contact image sensor 11
Address counter 14 for outputting an address corresponding to each of the elements, and correction data is stored for each pixel of the photoelectric conversion elements arranged in the main scanning direction of the contact image sensor 11 based on the address signal output from the address counter 14. A memory (ROM) 19 is provided. Then, the correction data read from the memory 19 is input to the differential amplifier 13 via the D / A converter 20.

In this output level correction apparatus, in the manufacturing process, an all black original is scanned for each image sensor to check the black level floating for each photoelectric conversion element, and the correction amount is written in the memory 19. . Therefore, when the contact image sensor 11 reads a document and sequentially inputs the output from each photoelectric conversion element to one input terminal of the differential amplifier 13, the black level floating of the photoelectric conversion element at that time is caused by the address counter 14. Then, it is read from the memory 19, converted into a digital signal by the D / A converter 20, and sequentially input to the other input end of the differential amplifier 13. Therefore, the differential amplifier 13 amplifies and outputs the difference between the image signal output of the contact image sensor 11 and the signal corresponding to the black level floating which is the output of the D / A converter 20, and is input to the image signal processing circuit 30. The image signal is a proper signal in which the floating of the black level is removed.

[0008]

In such a conventional output level correction apparatus, the correction data of the black level floating is written in the memory (ROM) 19 in the manufacturing process of the module including the contact image sensor 11 and its periphery. For,
This is effective in correcting the black level floating due to stray light.
However, if the characteristics of the photoelectric conversion element 7 and the LED array 5 change with time, the black level floating as shown in FIG. 8 also changes with time. Will be difficult. In this case, just before scanning the original, it is sufficient to scan the all-black original once to check the floating of the black level for each photoelectric conversion element and write it in the memory (RAM). However, every time the original is read, the all-black original is read. There is a problem in that scanning is extremely tedious and impractical.

[0009]

SUMMARY OF THE INVENTION It is an object of the present invention to provide an output level correction device for an image sensor, which is capable of correcting a black level floating which changes with time without scanning an all black original every time an original is read. is there.

[0010]

A first output level correction apparatus according to the present invention provides an image sensor output level when an image sensor reads and scans an all-black original and an image sensor output level when the image sensor reads and scans at zero incident light. A first memory for pre-storing the difference value of the image sensor output level;
A second memory for storing the output level of the image sensor when the image sensor reads and scans with zero incident light; an adder for adding the output levels of the image sensors read from the first and second memories; and the image sensor The image signal output level and the addition result are added or subtracted to remove the black level floating, and an arithmetic means is provided.

In this case, the first memory and the second memory store the difference in output level and the output level corresponding to each pixel of the plurality of photoelectric conversion elements forming the image sensor, and the adder is The respective output levels of the first and second memories are added to correspond to the respective pixels, and the calculating means is configured to calculate the output level of each pixel of the image sensor using the output value of the adder. To be done.

The second output level correction apparatus of the present invention is a difference value between the image sensor output level when the image sensor scans and scans an all-white original and the image sensor output level when the scan is performed at zero incident light. Is stored as a bright correction value, and the difference value between the image sensor output level when the image sensor reads and scans an all-black original and the image sensor output level when the image sensor reads and scans with no incident light is divided by the bright correction value. Means for storing the calculated value as a dark correction value, a dark correction value detection / correction circuit for calculating the dark correction value, and performing dark correction by the dark correction value stored for the output level of the image sensor, A bright correction value detection / correction circuit for calculating the bright correction value and performing the bright correction based on the stored bright correction value for the output level of the image sensor is provided. And wherein the Rukoto.

[0013]

In the first output level correction apparatus of the present invention, the first
The value of the difference between the output level of the all-black original and the output level of zero incident light is stored in advance in the memory of, and the original is read in the state of zero incident light. By correcting the output level of the image sensor based on, the black level floating caused by stray light can be corrected without reading the all black original every time the original is read, and it is possible to cope with the change with time of the output level. In addition, simplification of the document reading operation can be achieved.

In the second output level correction apparatus of the present invention,
The difference in the output level of the incident light zero with respect to the read output level of the all-white original is calculated as the bright correction value, and the difference of the output level of the incident light zero with respect to the read output level of the all-black original is calculated and this Divide by the correction value to obtain the dark correction value, and correct the output level of the image sensor with these bright correction value and dark correction value to correct the black level floating that occurs mechanically and to cope with changes over time. In addition, the document reading operation can be simplified.

[0015]

Next, an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing a first embodiment of an output level correction device for an image sensor according to the present invention. In the figure, 11 is a contact image sensor in which a plurality of pixels such as CdS photoelectric conversion elements are arranged in series in the main scanning direction, and outputs an electric signal according to the black and white density of the original. A timing circuit 12 drives the contact image sensor 11. Here, the contact image sensor 11 is cleared by the start pulse S1, and the electric signal is output from each pixel of the contact image sensor 11 by the clock pulse S2. The original reading output of the contact image sensor 11 is input to one input terminal of the differential amplifier 13. Also, 14
Is an address counter, which is driven by each pulse of the timing circuit 12 and outputs an address signal corresponding to each pixel of the contact image sensor 11.

Reference numeral 15 is a first memory, 16 is a second memory, and each of the memories 15 and 16 stores correction data at an address corresponding to each pixel of the contact image sensor 11, as described later. Is possible. The correction data stored in and read from the first and second memories 15 and 16 are added in the adder 17, converted into an analog signal in the D / A converter 18, and then the difference is added. It is input to the other input terminal of the dynamic amplifier. The output of this differential amplifier is output to the image signal processing circuit 30 including a binarization circuit and the like.

In the first memory 15, the value of the image signal level f of zero incident light when the illumination for illuminating the original is turned off corresponds to each pixel of the photoelectric conversion element forming the contact image sensor 11. It is stored. For example, in the example of the contact image sensor shown in FIG. 7, reading is performed by the contact image sensor when the LED array as the light source is not lit, and the output level of each pixel of each photoelectric conversion element at that time is determined. Store. The value of the image signal level f is written by scanning the contact image sensor immediately before reading the original. Therefore, the first memory 15 is composed of a rewritable RAM.

In the second memory 16, the difference between the image signal level b when the all black original is read and the image signal level f in the non-illuminated state, that is, the value of bf is the value of the contact image sensor 11. It is stored corresponding to each pixel of the photoelectric conversion element. The value of b-f is obtained by scanning the all black original document with the illumination turned on and scanning with the illumination turned off in the manufacturing process, and is written in the second memory 16. Therefore, the second memory 16
Is configured with ROM or RAM backed up by a battery so that the value is not erased.

The output level correction in the image sensor having this structure will be described. FIG. 2 is a waveform diagram showing the output level corresponding to each pixel of the photoelectric conversion elements arranged in the main scanning direction of the contact image sensor 11, where a is the image signal level when an all-white original is read, and b is the total. The image signal level when the black original is read, and f indicates the image signal level when the illumination for illuminating the original is turned off. Here, the difference b-f between the image signal level b when the all-black original is read and the image signal level f when the illumination for illuminating the original is turned off is the black level floating due to the influence of stray light.

Therefore, when reading the original,
First, the contact image sensor 11 performs reading with the illumination turned off to obtain the image signal level f,
Write to the memory 15. Then, the original document is read, and the obtained image signal level is input to one input terminal of the differential amplifier 13. Further, in synchronization with this, the image signal level f written in the first memory 15 and the second memory 1
The value (bf) of the corresponding pixel written in 6 is input to the adder 17, and the addition is performed here. As a result, the calculation of f + (b-f) is executed, and as a result, the image signal level b at the time of reading the all-black original is obtained. This image signal level b is input to the other input terminal of the differential amplifier 13, where the image signal level c for reading the original document is set.
The corrected image signal level can be obtained by taking the difference between and. That is, it is possible to correct the black level floating due to stray light.

Therefore, in this output level correction device, if the contact image sensor is simply read in a state where the illumination light is turned off and the incident light is zero, the original is read every time the original image is read in order to correct the black level floating due to stray light. It is not necessary to mount an all-black original on the contact image sensor and read it, and the operability of original reading can be simplified. Further, by reading this zero incident light, it is possible to cancel the temporal change in the image sensor, and it is possible to cope with the accompanying temporal change in the output level. It should be noted that this reading in the state of zero incident light can be automatically performed as a part of the initialization operation of the contact image sensor, in which case the operator does not need to perform any operation. It is possible to improve operability.

The value b-f for the first memory 15
May be stored when the operator considers it necessary. In this case, the first memory 15 is composed of a rewritable ROM (EPROM). The value f is stored in the second memory 16 at the beginning when reading a plurality of originals, but it is preferable to store the value f for each original to be read in the case of highly accurate reading.

Here, in general, the image signal level b of the contact image sensor when an all-black original is read largely changes with temperature, but the level difference b-f hardly changes. Therefore, even if the value of the level difference b−f is written in the second memory in advance in the manufacturing process, it does not affect the actual application. On the other hand, since the value of the image signal level b is written immediately before reading the original, it is possible to perform correction by following the temperature change.

Further, if the image signal level a when an all-white original is read changes due to the sensitivity of the contact image sensor and changes with time of illumination, the gain of the amplifier circuit built in the contact image sensor is adjusted, or The value of the level a can be kept constant by controlling the illuminance of the illumination. By controlling in this way, the values of the levels b and f are also kept substantially constant. In particular, when a point light source such as an LED is used for illumination, the light distribution of illuminance may change, but since the level difference b-f hardly changes at this time as well, a suitable output level as described above is obtained. Can be corrected. This is because the image signal level f when the illumination for illuminating the original is turned off does not change even if the illuminance distribution of the illumination changes. The level difference b−f does not become zero because
Since it is the effect of stray light, which is also affected by the illuminance of the dots around the relevant dot, the change in the light distribution of illuminance also affects the image signal level b, but it is as affected as the image signal level a. Absent.

FIG. 3 is a block diagram of a second embodiment of the output level correction apparatus for an image sensor according to the present invention. The output level correction device for this image sensor includes a contact image sensor 50 including a direct current reproduction circuit, and a dark correction value detection / correction circuit 51 for correcting the black and white levels of the data read by the contact image sensor 50. And a bright correction value detection / correction circuit 52. Further, in these circuits, a brightness / darkness correction memory 53 for storing brightness / darkness correction values described later, an address counter 54 for designating an address of the brightness / darkness correction memory 53, and the contact image sensor 5 are provided.
0 and other timing circuits 55 for supplying timing signals to other control circuits are connected.

The dark correction value detection / correction circuit 51, for each sensor element, with respect to the output level when the sensor is not illuminated,
A correction value is obtained by performing a calculation using the stored value of the brightness / darkness correction memory 53, and dark correction is performed on the output level of the contact image sensor 50. Also, the bright correction value detection / correction circuit 5
Reference numeral 2 calculates a bright correction value for each sensor element by performing a calculation on the output level of all white from the contact image sensor 50 by using the stored value of the bright / dark correction memory 53. Based on this, the output level of the contact image sensor 50 is corrected.

In the present embodiment, the signal processing system after correction is composed of a follow-up comparison type analog-digital converter, and the white level peak voltage V of the backside white density during the reading of the original document.
a peak detector circuit 56 for detecting the _peak, and ABC gain control circuit 57 for controlling the peak value from the history of the white level peak voltage, the ABC gain peak DA converting the back white density in document reading detected by the control circuit 57 Hold 58 and finally the analog input signal after the light / dark correction
7-bit image signal A for analog-to-digital conversion
DC59 and 2-bit DA for generating a relative voltage from the crank level V _CLP of the output of the peak hold 58 and the dark level output as the light / dark reference voltage of the image signal ADC 59.
C60 and 61, and a memory 62 for giving a coefficient to be input to the first 2-bit DAC 60.

The operation of the output level correction device for the image sensor of the second embodiment will be described. A document illumination light source (not shown) is turned off, and the image signal level f of the contact image sensor 50 at this time is converted to a 6-bit follow-up comparison method analog-digital conversion incorporated in the dark correction value detection / correction circuit 51. A / D conversion is performed with a container. In addition, the brightness correction memory 53 stores the image signal level b of the contact image sensor 50 that has read the all-black original in advance and similarly AD-converted the original illumination light source. Then, the dark correction value detection / correction circuit 51 obtains the obtained image signal level f,
The difference b-f between the two is calculated from the value b read from the memory, and stored in the brightness correction memory 53.

Then, the light source for illuminating the original is turned on, an original feeding roller (not shown) is read as the white reference image signal level a, and the bright correction value detection / correction circuit 52 stores it in the brightness correction memory 53. The stored image signal level f is read, the difference a−f between them is calculated, and stored in the brightness correction memory 53 as a brightness correction value. Then, the dark correction value detection / correction circuit 51 stores the stored b-f and the a
Division with −f is performed, and (b−f) / (a−f) is used as the dark correction value.

The value of af obtained in the bright correction value detection / correction circuit 52 is input to the peak detection circuit 56, and the peak value is detected as the white level peak voltage _Vpeak . The white level peak voltage V _{peak is set} to 100%, while the clamp level V _CLP regenerated by the direct current regeneration circuit built in the contact image sensor 50 is set to 0%.
The white reference value is detected for each sensor element with the full scale in the range of 40% to 100%. At this time, the white level peak voltage V _peak is held by the peak hold 58. The detected white reference value is stored in the brightness correction memory 53.

When an actual document is read, first, an all black document is read to detect the output level b, and a document feed roller is read to detect the output level a. And the output level f of incident light zero stored in advance in the brightness correction memory 53, the brightness correction value and the dark correction value are calculated and stored in the brightness correction memory 53. Then, with respect to the output level of the contact image sensor 50 for the document to be read, the dark correction value detection / correction circuit 51 causes the dark correction value (b−
f) / (a-f) is read out and the dark correction value is analogically calculated for each sensor element. Similarly, the light correction value detection / correction circuit 52 reads the light correction values af from the light / dark correction memory and normalizes the white reference value for each sensor element.

Further, during the reading of the original, the peak detection circuit 56 detects the white peak for each line, and the ABC gain control circuit 57 controls the white level peak voltage _Vpeak from the history of the white peak detection result to obtain it. The peak hold 58 holds the obtained white level peak voltage _Vpeak, and the signal processing after the above-mentioned brightness correction is performed as follows.

The brightness-corrected analog image signal is analog-digital converted by a 7-bit image signal ADC 59. Eshingo brightness reference voltage of ADC59 is 2 bits D for relative voltage generated from the white level peak voltage V _peak and the clamp level V _CLP dark level output as shown in FIG. 4
Determined from AC60 and AC61. That is, the bright reference voltage V
_{DAC2 is, V DAC2 = V CLP + β} × (V peak -V CLP) dark reference voltage V _{DAC1 is, V DAC1 = V CLP + α} × (V peak -V CLP) where, α = 0,0.08,0 .17, 0.25 β = 1, 0.92, 0.83, 0.75. It should be noted that α and β are coefficients relating to the accuracy of correction, and the larger the value, the more accurately the output level can be corrected. These α and β are preset in the memory 62.

Generally, by setting the reference voltage V _DAC2 lower than the white level peak voltage V _peak and setting the dark reference voltage higher than the dark level output clamp level V _CLP , the contrast of the image signal AD-converted by the image signal ADC 59 is set. Emphasize. However, here, in particular, the 2-bit DAC 60 is used in order to subtract the floating of the image signal level from the dark level when the all-black original shown in FIG. 2 is read. That is,
In a general contact image sensor, the floating of the image signal level from the dark level when reading an all-black original is about 15 to 20% of the white level. Therefore, by setting α = 0.17, there is an effect that the image signal level when reading the all-black original is corrected instead of the image signal level when the illumination is turned off.

Here, it is also possible to increase the set value of α and pre-write it in the memory 62 for each sensor element and each contact image sensor in the manufacturing process. By doing so, substantially the same effect as that of the first embodiment can be obtained.
This is shown in FIGS. Here, a is an image signal level when an all-white original is read, b is an image signal level when an all-black original is read, and f is an image signal level when the illumination is off. Further, for simplification, the sensor element is divided into four blocks B1, B2, B3 and B4, and the waveforms of a, b and f are the same in the same block.

In FIG. 5, (A-0), (B-0),
(C-0) and (D-0) are the image sensor outputs before the correction, (A-0) is the first image sensor output, and (B
(0) is an image sensor output when the sensor sensitivity is deteriorated to 80%, (C-0) is an image sensor output when the light distribution by the LED array changes with time, (D-0)
Is the image sensor output when the dark level rises due to the rise in environmental temperature. Further, the waveform of (A-0) is (B
−0), (C-0), (D-0), the output when the gain adjustment of the amplifier circuit of the contact image sensor 11 of the first embodiment is adjusted (A-1) in FIG. (B-
1), (C-1), and (D-1). The respective gains are 1.25, 1.11 and 0.91. (A-1) has a gain of 0 and has the same waveform as (A-0).

On the other hand, in FIG. 7, (A-2), (B-
2), (C-2) and (D-2) are the outputs of the image sensor 50 of the second embodiment, respectively (A-0) and (B) in FIG.
It is the output of the bright correction value detection / correction circuit 52 after the light / dark correction when −0), (C-0), and (D-0).

In terms of the corrected output levels shown in FIGS. 5 to 7, the first embodiment has (A-1) and (B- in FIG. 6), respectively.
In the waveforms 1), (C-1), and (D-1), it is possible to set, for each sensor element, a level higher than the waveform f as the black level. The second embodiment is shown in FIG.
2), (B-2), (C-2), and (D-2), the black level can be set for each sensor element by dark correction. Further, the white level can be set to a constant level by the bright correction.

It should be noted that the difference between the first embodiment and the second embodiment lies in the extent to which correction is possible with respect to the temperature change and the secular change of the output of the contact image sensor, but the second embodiment is different. However, it is unavoidable that the configuration becomes complicated as compared with the first embodiment, and therefore either configuration of both embodiments may be adopted depending on the quality, price, etc. required of the image sensor.

[0040]

As described above, the output level correction apparatus for a first image sensor according to the present invention uses the first memory when the image sensor reads and scans an all-black original and when it scans with zero incident light. The value of the difference between the respective output levels is stored in advance, the image sensor output level when the image sensor reads and scans at zero incident light is stored in the second memory, and these stored values are read out and stored. By performing addition and calculating the added value with respect to the image signal output level of the image sensor, stray light mechanically generated without scanning an all-black original for every scanning of the original causes stray light. Black level floating can be corrected. Also, since this correction is effective against the temporal change in the floating of the black level, the document reading operation can be simplified, while an appropriate image signal can be obtained, resulting in more faithful image reproduction and halftone expression. Has the effect that can be realized.

Further, in the present invention, the difference in output level and the output level corresponding to each pixel of the plurality of photoelectric conversion elements forming the image sensor are stored in the first memory and the second memory, respectively. Since the output level of each pixel of the image sensor is corrected based on the output level, it is also effective for floating of the black level in each pixel and temporal change.

The output level correction circuit of the second image sensor according to the present invention determines the difference value between the output levels of the image sensor when the image sensor reads and scans an all-white original and when it scans with zero incident light. Means to store as a correction value and a value obtained by dividing the difference value between the image sensor output level when the image sensor scans and scans an all-black original and the image sensor output level when scanning and scanning at zero incident light by the bright correction value. Is stored as a dark correction value, and a bright correction and a dark correction are performed on the output level of the image sensor based on the bright correction value and the dark correction value. Level drift can be corrected with high accuracy, and changes over time can be dealt with. In addition, the brightness correction can be used to correct the output level of a white document at the same time. Simplification of the document reading operation to increase the reliability of the collected can be achieved.

[Brief description of the drawings]

FIG. 1 is a block configuration diagram of a first embodiment of an output level correction device of the present invention.

FIG. 2 is a diagram showing a relationship between an output level of a contact image sensor and its correction value.

FIG. 3 is a block diagram of a second embodiment of the present invention.

FIG. 4 is an explanatory diagram of a bright / dark reference voltage in the image signal ADC.

FIG. 5 is a diagram of an output level before correction in the device of the present invention.

FIG. 6 is a diagram of output levels after correction in the first embodiment of the present invention.

FIG. 7 is a diagram of corrected output levels in the second embodiment of the present invention.

FIG. 8 is a diagram for explaining black level floating in a contact image sensor.

FIG. 9 is an optical mechanism diagram for explaining the influence of stray light on the image sensor.

FIG. 10 is a block diagram of an example of a conventional output level correction device.

[Explanation of symbols]

 11 Contact Image Sensor 13 Differential Amplifier 15 First Memory 16 Second Memory 17 Adder

Claims (7)

(57) [Claims] 1. An output level correction device for correcting fluctuations in output level caused by stray light in the image sensor, the image level sensor comprising an image sensor for outputting an electric image signal according to the black and white density of a document. A first memory for pre-storing a value of a difference between an image sensor output level when the image sensor reads and scans an all-black original and an image sensor output level when the image sensor reads and scans at zero incident light; A second memory for storing an image sensor output level when the image sensor reads and scans with zero incident light; an adder for reading the values stored in the first and second memories and adding them; A calculator that removes the black level floating by adding or subtracting the image signal output of the image sensor and the addition result. An output level correction device for an image sensor, comprising: a step. 2. The first memory and the second memory store the difference in output level and the output level corresponding to each pixel of a plurality of photoelectric conversion elements forming an image sensor, and the adder stores each difference. The image according to claim 1, wherein the output levels of the first and second memories are added in correspondence with the respective pixels, and the calculation means calculates the output levels of the respective pixels of the image sensor by using the outputs of the adders. Sensor output level correction device. 3. The first memory stores the difference in output level of the image sensor only when the image sensor is manufactured or when the user determines that it is necessary, and the second memory stores the output level of the image sensor each time a document is read. The output level correction device for an image sensor according to claim 1, wherein 4. The image sensor according to claim 1, wherein the first memory is a non-volatile memory or a volatile memory backed up by a battery, and the second memory is a volatile memory. Output level correction device. 5. An output level correction apparatus for correcting an output level fluctuation caused by stray light in the image sensor, the image level sensor comprising an image sensor for outputting an electric image signal according to a black-and-white density of a document. A means for storing a difference value between the image sensor output level when the image sensor reads and scans an all-white document and the image sensor output level when the image is read and scanned at zero incident light as a bright correction value; Means for storing, as a dark correction value, a value obtained by dividing a difference value between the image sensor output level when the original is read and scanned and the image sensor output level when the original is read and scanned at zero incident light, as a dark correction value; A dark correction that calculates a correction value and performs dark correction based on the dark correction value stored for the output level of the image sensor. A positive value detection / correction circuit, and a bright correction value detection / correction circuit that calculates the bright correction value and performs bright correction with the stored bright correction value for the output level of the image sensor. Image sensor output level correction device. 6. The storage means stores the respective bright correction values and dark correction values corresponding to the respective pixels of the plurality of photoelectric conversion elements forming the image sensor, and the bright correction value detection / correction circuit and the dark correction value detection are performed. 6. The output level correction device for an image sensor according to claim 5, wherein the / correction circuit performs bright and dark corrections on the outputs of the respective pixels of the image sensor. 7. The image sensor output level when reading and scanning an all-black original is stored in the storage means only when the image sensor is manufactured or when the user determines that it is necessary. 7. The output level correction device for an image sensor according to claim 5, wherein the output of the image sensor of the reading scan is stored in the storage means each time the document is read.