JP2006262182A - Image reading apparatus, image forming apparatus and image reading method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To make a linearity correction of each terminal about a portion larger than image data of a maximum reflectance and a portion smaller than the image data of a minimum reflectance. <P>SOLUTION: As a chart (white side) with a highest reflectance does not go to a full-scale size and a chart with a lowest reflectance does not go to 0, correction data cannot be created. Then, an F, L linearity detection and correction table creator 15 is provided, and the correction data is created therein to make a linearity correction, and an output value on the L side is output matching data on the F side based on an L-side gamma correction table 14. When this takes place, for instance, as for the white side, a differential is decreased in inverse proportion to the output value from data of the chart with the highest reflectance and the maximum 255 is set as 255, and as for the black side, a differential is decreased in proportion to the output value from data of the chart with the lowest reflectance and the minimum 0 is set as 0. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an image reading apparatus using a CCD image sensor having a plurality of output terminals (First, Last or Odd, a combination of Even, First, Last) for each line, an image forming apparatus including the image reading apparatus, and an image reading method. .

As a technique of this means, for example, the invention disclosed in Patent Document 1 is known. The present invention relates to a color linear image sensor of right and left division reading method, a plurality of signal processing means for performing predetermined signal processing on image signals output from respective output terminals of the color linear image sensor, a white plate, a gray plate The level of the image signal of the reference plate and the white plate is made to substantially match the first predetermined level, the level of the image signal of the gray reference plate is made to substantially match the second predetermined level, and other densities are accommodated. A CPU for obtaining adjustment data for adjusting the level of the image signal to a level obtained by interpolating between the first and second predetermined levels, and output from the corresponding signal processing means using the adjustment data. And an LUT that adjusts the image signal to reduce the signal level difference and match the linearity characteristics of the output channels.
Japanese Patent Laid-Open No. 2002-21818

  In the invention described in Patent Document 1, an average value of the reading area of each reflectance is obtained for each of FE (odd number in the first half), FO (even number in the first half), LE (odd number in the second half), and LO (even number in the second half). Adjustment data for each terminal is obtained to match a single reference terminal. In this case, interpolation is performed between the respective reflectances, but adjustment data for a portion higher than the maximum reflectance and adjustment data for a portion lower than the minimum reflectance are not particularly taken into consideration. Although the EO (even number, odd number) difference and the FL (first half, second half) difference also occur in the signal processing circuit, most of them occur due to variations in the linearity of the CCD output itself. Here, the linearity means the characteristic of the incident light quantity of the CCD versus the analog output, but the analog output is basically in a straight line because it is proportional to the light quantity. However, it may not be on a straight line. FIG. 13 shows the relationship between the amount of incident light and the scanner output value. From this figure, it can be seen that there is a difference in linearity between F and L.

  Therefore, it is necessary to correct at least one of the output values of F and L. Therefore, conventionally, when creating correction (adjustment) data, the data obtained by reading each reference density plate (or reference document) is interpolated between the respective reflectances. Since there is no reference density plate for the adjustment data of the part higher than the reflectance and the part lower than the minimum reflectance, the adjustment data cannot be created by interpolation.

  Further, in Patent Document 1, there is no particular mention of the case where the data obtained by reading each reference density plate reaches or approaches the upper limit or lower limit of the A / D conversion output. In addition, the average value of the area becomes a representative value of the read data of the reference original. However, the fact that the A / D conversion output data becomes full scale or 0 is actually unknown as to the read value of the pixel. This is because the full scale or 0 becomes a reading value even when the full scale is exceeded or below 0.

  That is, if at least one of the image data used for calculating the average value is full scale or 0, the average value obtained by that is lower or higher than the actual value. In addition, the representative value when the reference density is read may be the maximum value in the histogram. In this case as well, when full scale or 0 corresponds to it, it becomes a representative value, so it is not an appropriate value.

  The present invention has been made in view of the actual situation of the prior art, and a first object of the present invention is to provide linearity of each terminal for a portion larger than the image data of the maximum reflectance and a portion smaller than the image data of the minimum reflectance. It is to enable correction.

  The second object is that when one of the representative values of the read data of each reference document when the reference document is read reaches or approaches the upper limit or lower limit of the A / D conversion output data, It is to enable appropriate correction between the output terminals.

  In order to achieve the object, the first means includes a photoelectric conversion element array that distributes and outputs pixels to a plurality of output terminals, and an A / D conversion means that converts an image signal from the photoelectric conversion element array into digital data. Means for holding image data after A / D conversion when reading a plurality of reflectance reference documents, and means for obtaining adjustment data between output terminals by interpolating the data held by the holding means And an image reading apparatus comprising means for acquiring at least one of adjustment data of a portion larger than the image data of the maximum reflectance and a portion smaller than the image data of the minimum reflectance among the plurality of reference documents separately from the interpolation. And

  According to a second means, in the first means, at least one of the adjustment data of a portion larger than the image data of the maximum reflectance and a portion smaller than the image data of the minimum reflectance among the plurality of reference documents is a part thereof. Is linear.

  According to a third means, in the first means, at least one of adjustment data of a portion larger than the image data with the maximum reflectance and a portion smaller than the image data with the minimum reflectance among the plurality of reference documents is set as an output value. The difference is proportionally reduced.

  According to a fourth means, in the first means, at least one of adjustment data of a portion larger than the image data with the maximum reflectance and a portion smaller than the image data with the minimum reflectance among the plurality of reference documents is set as an output value. The difference is reduced in inverse proportion.

  According to a fifth means, in the first means, at least one of adjustment data of a portion larger than the image data with the maximum reflectance and a portion smaller than the image data with the minimum reflectance among the plurality of reference documents is proportional to the output value. The difference is once increased and then decreased.

  According to a sixth means, in the first means, at least one of adjustment data of a portion larger than the image data with the maximum reflectance and a portion smaller than the image data with the minimum reflectance among the plurality of reference documents is inversely proportional to the output value. The difference is once increased and then decreased.

  The seventh means includes a photoelectric conversion element array that distributes and outputs pixels to a plurality of output terminals, an A / D conversion means that converts an image signal from the photoelectric conversion element array into digital data, and a plurality of reflectances. Means for holding representative values of image data after A / D conversion when a reference document is read, means for obtaining adjustment data between output terminals by interpolating data held by the holding means, and the reference An image is provided with means for separately acquiring a representative value of the reference original when any of the representative values of the read data of each reference original when the original is read reaches the upper limit or lower limit of the A / D conversion output data. Features a reader.

  The eighth means includes a photoelectric conversion element array that distributes and outputs pixels to a plurality of output terminals, an A / D conversion means that converts an image signal from the photoelectric conversion element array into digital data, and a plurality of reflectances. Means for holding representative values of image data after A / D conversion when a reference document is read, means for obtaining adjustment data between output terminals by interpolating data held by the holding means, and the reference And a means for separately obtaining a representative value of the reference document when any of the representative values of the read data of each reference document when the document is read is close to the upper limit or lower limit of the A / D conversion output data. Features a reader.

  A ninth means is the seventh or eighth means, wherein the obtaining means obtains a true representative value of the reference document from a configuration of each pixel data based on the representative value. To do.

  A tenth means is the ninth means, wherein the obtaining means refers to a configuration of image data of a reference document having a reflectance other than the configuration of the pixel data in addition to the configuration of the pixel data. And

  The eleventh means is characterized in that the image forming apparatus includes the image reading apparatus according to any one of the first to tenth means.

  The twelfth means converts the image signal from the photoelectric conversion element array that distributes and outputs pixels to a plurality of output terminals into digital data, and performs A / D conversion after reading a plurality of reflectance reference documents. When image data is stored and adjustment data between output terminals is obtained by interpolating the stored data, a portion larger than the maximum reflectance image data in a plurality of reference documents, and smaller than the minimum reflectance image data The image reading method is characterized in that at least one of the partial adjustment data is acquired separately from the interpolation.

  The thirteenth means converts the image signal from the photoelectric conversion element array that distributes and outputs pixels to a plurality of output terminals into digital data, and performs A / D conversion after reading a plurality of reflectance reference documents. When the representative value of the image data is held, and when the adjustment data between the output terminals is obtained by interpolating the held representative value, one of the representative values of the read data of each reference document when the reference document is read is The image reading method is characterized in that when the upper limit or lower limit of the A / D conversion output data is reached, or when the upper limit or the vicinity is reached, the representative value of the reference original is separately acquired.

  According to the present invention, at least one of the adjustment data of the plurality of reference documents larger than the maximum reflectance image data and smaller than the minimum reflectance image data is obtained separately from the interpolation. The linearity correction of each terminal can be performed for a portion larger than the image data of the reflectance and a portion smaller than the image data of the minimum reflectance.

  In addition, when any of the representative values of the read data of each reference original when the reference original is read reaches or approaches the upper limit or lower limit of the A / D conversion output data, the representative value of the reference original is Since it is obtained separately, when one of the representative values of the read data of each reference document when the reference document is read reaches or approaches the upper limit or lower limit of the A / D conversion output data, a plurality of output terminals Appropriate corrections can be made in between.

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

<First Embodiment>
FIG. 1 is a diagram showing a schematic configuration of a general document reading apparatus. In the figure, an original reading apparatus as an image reading apparatus includes a first carriage 5 on which a light source 3 and a first mirror 4 are mounted, a second carriage 8 on which second and third mirrors 6 and 7 are mounted, and a first carriage. Or an optical system 2 comprising a lens 9 for condensing the reflected light guided by the third mirrors 4, 6, and 7, and a CCD 10 for reading and photoelectrically converting an image formed on the imaging surface by the lens 9. The image information on the document placed on the contact glass 1 is read. The first carriage 5 and the second carriage 8 are moved in the sub-scanning direction (rightward in the drawing) by a motor (not shown) at a speed of 2 to 1 to read the entire surface of the document.

  FIG. 2 is a diagram showing an internal configuration of the F / L type CCD 10. In the CCD 10 according to this embodiment, in order to increase the CCD transfer frequency, a terminal (First F) sequentially transfers and outputs the first half of the pixels from the first pixel, and a terminal that sequentially transfers and outputs the second half of the pixels from the last pixel. Each of (L of Last) has odd (O) and even (E) terminals. That is, the CCD 10 has a total of four outputs (RS1 to RS4) as shown in FIG.

  FIG. 3 is a block diagram of the document reading apparatus when an F / L type CCD is used in the circuit. The image signal read by the CCD 10 is divided into F E and O, L E and O, and outputted (FE, FO, LE, LO). In the analog signal processing circuit, E and O of F and E and O of L are respectively received by one LSI 11F and 11L, and an appropriate analog signal is generated by combining amplification factor adjustment and the output of EO. In the next A / D converters 12F and 12L, the image signals are converted from analog to digital, and shading correction circuits (comprising a memory and arithmetic elements) 13F and 13L for correcting illuminance distribution and pixel sensitivity difference, and further L digital. The data is output from the document reading device through the gamma correction table 14 for matching the F digital data only with the data. Here, the gamma correction table 14 reads several types of reflectance gray charts in advance so that, for example, L data after EO synthesis can be matched with F data. In this embodiment, the gray chart corresponds to a reference document having a plurality of reflectances.

  Specifically, it consists of a lookup table such as a ROM, and the L data is matched to the F data by writing the F data to the output in response to the input of the L data. In this case, correction data for all inputs is obtained by linear interpolation between the reflectances corresponding to the gray charts. Therefore, the L-side correction table and a CPU (not shown) for calculating the table value function as means for obtaining adjustment data. Although the case where L data is matched with F data is shown here, F data may be matched with L data, and FL data may be matched for each EO. In addition, there is an ideal linearity, and F and L data may be matched to this.

  By the way, it is a gray chart of several kinds of reflectivity, but even the chart with the highest reflectivity (white side) does not go to full scale (255 for 8-bit), but the lowest chart (black side) But it will not be 0. For example, if the white side is 230/255 and the black side is 20/255, correction data cannot be created because there is no data above 230 and below 20. In order to solve this, correction data is created by a method different from interpolation, and linearity correction is executed. For this purpose, an F, L linearity detection and correction table creation unit 15 is provided, where correction data is created, linearity correction is performed, and the L-side output value is matched to the F-side data based on the L-side gamma correction table. Output. Note that the F and L linearity detection and correction table creation unit 15 corresponds to a means acquired separately from the interpolation.

  The characteristic diagram of FIG. 4 is one example, and is an example in which the scanner output is decreased in inverse proportion to the reflectance. That is, the white side decreases the difference in inverse proportion to the output value from the chart data having the highest reflectance, the maximum 255 is set to 255, and the black side is proportional to the output value from the chart data having the lowest reflectance. Thus, the difference is reduced, and the minimum 0 is set to 0. Here, reducing the difference in inverse proportion to the reflectance means, for example, that when the correction data from 230 to 255 is created when the FL difference of 230 is −3 digit (that is, F is 230 and L is 227), the correction data is shown. As the data increases from the input value and output value of 5, the difference decreases and 255 becomes 255. The same applies to the black side. Such correction data is created separately from the interpolation data by the F and L linearity detection and correction table creation unit 15.

  The characteristic diagram of FIG. 6 is another example. On the white side, the difference is reduced in inverse proportion to the output value until halfway, but from there, it is assumed to be one-to-one linear (ie, the input and output are the same value), and the black side is The difference is reduced in proportion to the output value until halfway, but from there, it is set to one-to-one linear (that is, the same value for input and output). An example of white side correction data is shown in FIG. In many cases, the FL difference between the output value close to the full scale on the upper side and the output value close to 0 on the lower side is small in the first place, and there is no need for correction. For this reason, it is basically good to use a one-to-one linear, but if it is made linear between 230 and 20 suddenly, it becomes discontinuous at that portion, so as shown in FIG. After that, it is linear after that. The same applies to the black side.

  The characteristic diagram of FIG. 8 is related to FIG. 4, but in another example, for example, the upper side when the highest chart has only a reflectance of 40%, and the lowest chart has a reflectance of 60%. It corresponds to the lower side when there is only up to. An example of the correction data at this time is shown in FIG. In this case, the difference is once increased in proportion to the output value and then decreased, or is increased in inverse proportion to the output value and then decreased. This is because the FL difference is maximized because the reflectance is around 50%.

  The document reading apparatus according to the present embodiment is provided integrally or separately with an image forming apparatus including an image forming unit that forms an image on a sheet, for example, and functions as a copying machine. Since the image forming apparatus itself may include any known image forming means such as an electrophotographic system or an ink jet system, description thereof is omitted here.

  As described above, according to the present embodiment, either or both of the adjustment data of the portion larger than the image data of the maximum reflectance and the adjustment data of the portion smaller than the image data of the minimum reflectance are acquired separately from the interpolation. The linearity of each terminal can be corrected for the part.

  In addition, one or both of the adjustment data of the portion larger than the image data with the maximum reflectance and the portion smaller than the image data with the minimum reflectance among the plurality of reference originals are linear (one-to-one straight line). In this case, the linearity of each terminal can be further corrected for this portion.

  In addition, either or both of the adjustment data of a part larger than the image data of the maximum reflectance and the part of the reference document smaller than the image data of the minimum reflectance among the plurality of reference documents may be reduced in proportion to the output value, Alternatively, since the difference is decreased in inverse proportion to the output value, the linearity of each terminal can be further corrected for this portion.

  Further, among the plurality of reference originals, one or both of the adjustment data of the portion larger than the image data of the maximum reflectance and the adjustment data of the portion smaller than the image data of the minimum reflectance once increased in proportion to the output value. Since the difference is once increased in inverse proportion to the output value and then decreased, the upper side when the maximum reflectance of the reference density member is less than half, and the minimum reflectance of the reference density member are The linearity of each terminal with respect to the lower side when exceeding half can be corrected.

<Second Embodiment>
Next, a second embodiment will be described. FIG. 10 is a block diagram of an original reading apparatus using an F / L CCD in the circuit according to the second embodiment.
In the figure, the image signal read by the CCD 10 is divided into F E and O and L E and O in the same manner as in the first embodiment described above. In the analog signal processing circuit, E and O of F and E and O of L are respectively received by one LSI 11F and 11L, and an appropriate analog signal is generated by combining amplification factor adjustment and the output of EO. In the next A / D converters 12F and 12L, image signals are converted from analog to digital, and shading correction circuits 13F and 13L are provided for correcting illuminance distribution and pixel sensitivity difference. Then, F and L linearity detection for calculating and holding each representative value when the gray chart of several kinds of reflectance is read, and a gamma correction table for adjusting the F digital data only to the L digital data are created. A circuit having a function, an F / L linearity detection and correction table creation unit 15 is provided, and only L data is output from the document reading device via the gamma correction table 14. Here, in the gamma correction table 14, for example, the L representative value data is matched with the F representative value data.

  Specifically, the gamma correction table 14 includes a lookup table such as a ROM, and the L data is matched with the F data by writing the F data to the output in response to the input of the L data. In this case, correction data for all inputs is obtained by linear interpolation between the reflectances corresponding to each gray chart. Here, the case where L data is matched with F data is shown, but F data may be matched with L, and FL data may be matched for each EO. In addition, there is an ideal linearity, and F and L data may be matched to this. These are the same as those in the first embodiment.

  By the way, it is a representative value of several types of reflectance gray charts, and an average of reading values of several to several tens of pixels in both vertical and horizontal directions is used as a representative reading value at the reflectance. If even one of these several to several tens of pixels has full scale or zero data, it is no longer a representative value of the reflectance. Taking the case of 8 bits as an example, the full scale is 255, but a read value of a certain pixel of 255 indicates that it may be just 255, but may actually be higher. This is because, from the characteristics of the A / D converters 12F and 12L, the digital value is 255 even for an analog input exceeding 255. The same applies when the value is smaller than zero. In addition, instead of the average value, the maximum value in the histogram may be used as the representative value. Also in this case, if the read value of more than half becomes 255, the representative value becomes 255.

  Therefore, in this embodiment, it is detected whether or not there is 255 in the read data of each gray chart, and the true representative value of the gray chart is obtained from the data configuration. In addition, the data structure of the previous reflectance may be used.

Therefore, in this embodiment, an area information acquisition circuit 16 is provided as shown in FIG. The area information acquisition is to obtain the number of data corresponding to the reading value in each gray chart, the reading value that maximizes the number (maximum value in the histogram), the average value, the standard deviation, and the like. For example, in a gray chart of 100 pixels with a reflectance of 50%,
5 readings of 125 digit,
10 readings of 126 digit,
There are 21 readings of 127 digits,
28 readings of 128 digits,
129 digit readings are 21
10 readings of 130 digit,
131 digit readings are 5
Thus, the total number is 100, the read value with the maximum number is 128 digits, the average value is also 128 digits, and the standard deviation is 1.45 digits.

Now consider the following example: When the reflectivity is 90%,
One 251 digit,
3 252 digits,
7 253digits,
10 254 digits,
There are 79 255 digits, and the solid line in FIG. 11 corresponds to the graph. In this case, the maximum reading value is 255 digits, the average value is 254.63 digits, and the standard deviation is 0.82 digits. However, this has not been appropriate as described above.

In the case of appropriate data, first, the standard deviation is estimated from the area information of the gray chart having a low reflectance. As shown in FIG. 12, assuming that the previous reflectance is 70%, the standard deviation is 1.76 digits, and the average value or representative value is 200 digits, the standard deviation is proportional to the reflectance route. .00 digit. Next, when estimating from the data distribution of the portion that does not reach 255 of 90%, it is about 253.5 digits that reaches 16% (value of −σ) from the lowest data, and it is assumed that the distribution is a normal distribution. Then
253.5 + 2.00 = 255.5 (digit)
Is a true average value (or may be referred to as a representative value) (a broken line in FIG. 11 in the graph).

  In other words, the reflectivity is 90%, the number becomes maximum at 255 or 256 digits, the average value becomes 255.5 digits, and the standard deviation becomes 2.00 digits. Basically, a true representative value can be obtained in the same manner even if the ratio of clipping to 255 is larger.

  The difference between the case where the representative value of the reflectance 90% is 254.63 and 255.5 is different when obtaining interpolation data from 70% to 90% as shown in FIG. When the value exceeds 255, the value can only be set to 255, but from the 70% of 200 digits, the true representative value is higher than the actual value. If the true representative value is larger than 255, the difference is further increased, which is more effective. The same is true for the low reflectance side.

  In the present embodiment, the area information acquisition circuit 16 corresponds to means for separately acquiring a representative value of the reference document.

  Other parts that are not particularly described are configured in the same manner as the first embodiment described above and function in the same manner, and thus redundant description is omitted.

  As described above, according to this embodiment, when one of the representative values of the read data of each reference document when the reference document is read reaches or approaches the upper limit or lower limit of the A / D conversion output data Since the true representative value can be acquired by having a method of separately acquiring the representative value of the reference document, an appropriate correction can be made between a plurality of output terminals.

  Further, each reference document is obtained from data of a plurality of pixels, and each pixel data based on the representative value of the reference document that reaches or approaches the upper limit or lower limit of the A / D conversion output data. Since the true representative value can be obtained by analogizing the representative value of the reference document from the configuration (specifically, the numerical distribution of the read data), appropriate correction can be performed between the plurality of output terminals.

  Further, in addition to the configuration of each pixel data based on the representative value of the reference document that has reached or approached the upper limit, or the lower limit, the configuration of the image data of the reference document whose reflectance is one lower or higher ( Specifically, the true representative value can be acquired by referring to the standard deviation), so that appropriate correction can be made between a plurality of output terminals.

1 is a diagram illustrating a schematic configuration of a general document reading apparatus. It is a figure which shows the internal structure of F / L type CCD. 1 is a block diagram of a document reading apparatus according to a first embodiment using an F / L type CCD for a circuit. FIG. It is a figure which shows the characteristic in case an output reduces in inverse proportion to a reflectance with respect to incident light quantity. FIG. 5 is a diagram illustrating an example of output correction data with respect to an input in the case of the characteristics of FIG. 4. It is a figure which shows the characteristic in case an output is partially linear with respect to incident light quantity and others decrease in inverse proportion to a reflectance. FIG. 5 is a diagram illustrating an example of output correction data with respect to an input in the case of the characteristics of FIG. 4. It is a figure which shows the characteristic in case an output increases once in inverse proportion to a reflectance with respect to incident light quantity, and decreases. FIG. 5 is a diagram illustrating an example of output correction data with respect to an input in the case of the characteristics of FIG. 4. FIG. 6 is a block diagram of a document reading apparatus according to a second embodiment using an F / L type CCD for a circuit. It is a figure which shows the read value of a pixel with a histogram. FIG. 10 is a characteristic diagram illustrating a difference between an incident light amount corrected using an actual value and a true representative value and a scanner output in the second embodiment. It is a figure which shows the relationship between the incident light quantity of F / L type CCD, and a scanner output.

Explanation of symbols

2 Optical system 3 Light source 4, 6, 7 Mirror 9 Lens 10 CCD
11F, 11L analog LSI
12F, 12L A / D converter 13F, 13L Shading correction circuit 14 L side correction table 15 F, L linearity detection and correction table creation unit 16 Area information acquisition unit

Claims (13)

A photoelectric conversion element array that distributes and outputs pixels to a plurality of output terminals; and
A / D conversion means for converting an image signal from the photoelectric conversion element array into digital data;
Means for holding image data after A / D conversion when a reference document having a plurality of reflectances is read;
Means for interpolating data held by the holding means to obtain adjustment data between output terminals;
Means for obtaining at least one of adjustment data of a portion larger than the image data of the maximum reflectance and a portion smaller than the image data of the minimum reflectance among the plurality of reference documents separately from the interpolation;
An image reading apparatus.   2. The adjustment data of at least one of a portion larger than the image data with the maximum reflectance and a portion smaller than the image data with the minimum reflectance among the plurality of reference documents is linear. Image reading device.   The difference between at least one of adjustment data of a portion larger than the image data with the maximum reflectance and a portion smaller than the image data with the minimum reflectance among the plurality of reference documents is reduced in proportion to the output value. Item 2. The image reading apparatus according to Item 1.   The adjustment data of at least one of a portion larger than the image data with the maximum reflectance and a portion smaller than the image data with the minimum reflectance among the plurality of reference documents is reduced in inverse proportion to the output value. Item 2. The image reading apparatus according to Item 1.   At least one of the adjustment data of the reference data larger than the maximum reflectance image data and smaller than the minimum reflectance image data in a plurality of reference documents should be increased and then decreased in proportion to the output value. The image reading apparatus according to claim 1.   At least one of the adjustment data for the parts larger than the image data with the maximum reflectance and the parts smaller than the image data with the minimum reflectance in the plurality of reference documents should be decreased after increasing the difference in inverse proportion to the output value. The image reading apparatus according to claim 1. A photoelectric conversion element array that distributes and outputs pixels to a plurality of output terminals; and
A / D conversion means for converting an image signal from the photoelectric conversion element array into digital data;
Means for holding a representative value of the image data after A / D conversion when a reference document having a plurality of reflectances is read;
Means for interpolating a representative value of data held by the holding means to obtain adjustment data between output terminals;
Means for separately obtaining a representative value of the reference document when any one of the read values of each reference document when the reference document is read reaches an upper limit or a lower limit of A / D conversion output data;
An image reading apparatus. A photoelectric conversion element array that distributes and outputs pixels to a plurality of output terminals; and
A / D conversion means for converting an image signal from the photoelectric conversion element array into digital data;
Means for holding a representative value of the image data after A / D conversion when a reference document having a plurality of reflectances is read;
Means for interpolating data held by the holding means to obtain adjustment data between output terminals;
Means for separately obtaining a representative value of the reference original when any of the representative values of the read data of each reference original when the reference original is read approaches the upper limit or lower limit of the A / D conversion output data;
An image reading apparatus.   9. The image reading apparatus according to claim 7, wherein the acquiring unit acquires a true representative value of the reference document from a configuration of each pixel data based on the representative value.   The image reading apparatus according to claim 9, wherein the acquiring unit refers to a configuration of image data of a reference document having a reflectance other than the configuration of the pixel data in addition to the configuration of the pixel data.   An image forming apparatus comprising the image reading apparatus according to claim 1. The image signal from the photoelectric conversion element array that outputs the pixels by distributing the pixels to a plurality of output terminals is converted into digital data,
Holds the image data after A / D conversion when scanning a reference document with a plurality of reflectances,
When obtaining adjustment data between output terminals by interpolating the held data, at least adjustment data of a portion larger than the maximum reflectance image data and smaller than the minimum reflectance image data among a plurality of reference documents. An image reading method characterized in that one is acquired separately from interpolation. The image signal from the photoelectric conversion element array that outputs the pixels by distributing the pixels to a plurality of output terminals is converted into digital data,
Holds the representative value of the image data after A / D conversion when reading a reference document with a plurality of reflectances,
When obtaining the adjustment data between the output terminals by interpolating the held representative value, one of the representative values of the read data of each reference document when the reference document is read is the upper limit of the A / D conversion output data Alternatively, when the lower limit is reached, or when the upper limit or the vicinity is reached, a representative value of the reference original is separately acquired.

Images