JP2006079318A - Image reader and signal processing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image reader and a signal processing method capable of interpolating missing pixels with high accuracy and thus minimizing the memory capacity required for the interpolation, even if sensitivity characteristics vary among individual imaging elements that form an adhesion type image sensor. <P>SOLUTION: The image reader includes a characteristic correcting part 3 for correcting the variation in characteristic among imaging elements 36(1), 36(2) of a sensor chip, and an interpolating part 4 for interpolating the lack of pixel data using nearby pixels, the lack of the pixel data occurring at the connections between a plurality of sensor chips. The characteristic correcting part 3 performs corrections for pixels within a particular area, for example, an area containing part or the whole of the pixels used for the interpolation by the interpolating part 4. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an image reading apparatus using a contact image sensor and a signal processing method in the image reading apparatus.

  In a conventional image reading apparatus of this type, for example, a plurality of sensor chips are connected to each other so that image pickup devices are arranged in a line. In the connection between the sensor chips, the interval between the image pickup elements is larger than the place where the image pickup element is located. Therefore, interpolation is performed assuming that the image pickup element (pixel) is missing at that place. Interpolation methods include the method of using the average value of neighboring pixels of missing pixels as interpolation data, the method of calculating the regression line of pixels adjacent to the missing pixel using the least square method, and calculating the interpolation data from the regression line, missing A method of interpolating missing pixels using a method of calculating a quartic curve from four pixels adjacent to a pixel and calculating interpolation data from the quartic equation is used (see, for example, Patent Document 1).

JP2003-101724 (pages 5-6, FIGS. 3-5)

  Prior to the interpolation as described above, black level variation correction (hereinafter black correction) and image sensor sensitivity correction are performed. In order to perform such interpolation, data for correction is created and stored in a memory, and correction data is used when reading an image. A large-capacity memory is used for storing correction data. There was a problem that was necessary.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide an image reading apparatus and a signal processing method capable of reducing the memory capacity required for correction.

This invention
An imaging unit in which a plurality of sensor chips having a plurality of imaging elements are connected to each other and arranged in a row,
An A / D converter that converts an image signal output from the imaging unit into digital pixel data;
A characteristic correction unit for correcting variations in characteristics of the image sensor of the sensor chip;
An interpolation unit that interpolates the lack of pixel data that occurs at the connection parts of the plurality of sensor chips using pixels in the vicinity thereof,
The characteristic correction unit provides an image reading device that performs the correction on pixels in a specific region including part or all of pixels used for interpolation by the interpolation unit.

  In the image reading apparatus according to the present invention, a high-quality read image can be obtained with a small memory capacity with respect to the image data read by the contact image sensor.

Embodiment 1 FIG.
FIG. 1 is a diagram showing a configuration of an image reading apparatus according to Embodiment 1 of the present invention.
As illustrated, the image reading apparatus according to the present embodiment includes an imaging unit 1, an A / D conversion unit 2, a characteristic correction unit 3, an interpolation unit 4, and a reading control unit 5.

As will be described later, the imaging unit 1 has a plurality of imaging elements arranged one-dimensionally, and is controlled by the reading control unit 5 to read an image one-dimensionally and output an image signal.
The A / D converter 2 converts the image signal output from the imaging unit 1 into a series of digital pixel data. Each pixel data represents a value of a pixel corresponding to the image sensor.

  The characteristic correction unit 3 includes switching units 11 a and 11 b, a correction calculation unit 12, a correction data generation unit 13, and a region designation unit 14.

  The correction calculation unit 12 includes a black correction unit 21 and a sensitivity correction unit 22. The black correction unit 21 corrects the black level variation of the image sensor 36 and outputs the pixel data BK subjected to the black correction process. The sensitivity correction unit 22 corrects variations in sensitivity of the image sensor 36 and outputs pixel data WH subjected to sensitivity correction processing.

  The correction data generation unit 13 calculates black correction data BC for correcting black level variations and sensitivity correction data generation for calculating sensitivity correction data WC for correcting sensitivity variations. And a data memory 25 for storing the black correction data BC and the sensitivity correction data WC generated by the black correction data generation unit 23 and the sensitivity correction data generation unit 24, and the data is transmitted between them. It has a data bus 26 and a correction control unit 27 for controlling operations including data transmission between them.

The image reading apparatus shown in FIG. 1 can operate in any one of an image reading mode, a black correction data generation mode, and a sensitivity correction data generation mode, and the switching unit 11a includes the image reading mode and the sensitivity correction data. In the generation mode, the pixel data RD output from the A / D conversion unit 2 is supplied to the black correction unit 21 of the correction calculation unit 12, and in the black correction data generation mode, the pixel data RD is output from the A / D conversion unit 2. Pixel data RD is supplied to the correction data generation unit 13. The switching unit 11b supplies the pixel data BK output from the black correction unit 21 to the sensitivity correction unit 22 in the image reading mode, and the pixel data BK output from the black correction unit 21 in the sensitivity correction data generation mode. Is supplied to the correction data generation unit 13.
In the following description, the black correction data generation mode and the sensitivity correction data generation mode may be collectively referred to as a correction data generation mode.

  In the black correction data generation mode, the black correction data generation unit 23 of the correction data generation unit 13 corrects the black level variation of the image sensor based on the output BK from the A / D conversion unit 2. Data BC is generated and stored in the data memory 25. In the sensitivity correction data generation mode, the sensitivity correction data generation unit 24 of the correction data generation unit 13 corrects variations in sensitivity of the image sensor based on the output BK from the black correction unit 21 of the correction calculation unit 12. Sensitivity correction data WC is generated and stored in the data memory 25. The correction data generation unit 13 supplies the stored correction data to the correction calculation unit 12 in the image reading mode. Also in the sensitivity correction data generation mode, the stored black correction data is supplied to the black correction unit 21 of the correction calculation unit 12.

In the image reading mode and the sensitivity correction data generation mode, the black correction unit 21 of the correction calculation unit 12 uses the black correction data supplied from the black correction data generation unit 23 of the correction data generation unit 13 to perform an A / D conversion unit. 2 is corrected, and corrected pixel data BK is output.
In the image reading mode, the sensitivity correction unit 21 of the correction calculation unit 12 uses the sensitivity correction data supplied from the sensitivity correction data generation unit 24 of the correction data generation unit 13 to output the pixel data BK output from the black correction unit 21. And the corrected pixel data WH is output.

  The correction control unit 27 receives the read timing signal from the read control unit 5 for controlling the above-described operation, and based on this, the data currently output from the A / D conversion unit 2 and the black correction unit 21 is captured by the image pickup unit. 1 is recognized, the correction data generation unit 13 generates correction data based on the recognition result, and the correction data to the black correction unit 21 and the sensitivity correction unit 22 of the characteristic correction unit 12. Control the timing of supply.

  The data memory 25 is used not only for storing the above correction data, but also for storing data necessary for generating the black correction data BC and the sensitivity correction data WC and intermediate results of the calculation for generation.

  As will be described in detail later, the characteristic correction unit 3 of the present embodiment performs black correction for all pixels, while sensitivity correction is not performed for all pixels, but within a specific region. Pixels, for example, pixels in a region in the vicinity of the missing pixel, which are used for interpolation of the missing pixel by the interpolation unit 4, are subjected to sensitivity correction data, while the pixel used for interpolation by the interpolation unit 4 Correction is not performed for pixels other than.

  Therefore, the sensitivity correction data generation unit 24 generates correction data only for pixels in a specific region, the data memory 25 stores the generated sensitivity correction data, and the sensitivity correction unit 22 is stored. Based on the sensitivity correction data, correction is performed only for pixels in a specific region.

The area designating unit 14 designates the specific area to the correction data generating unit 13. That is, based on a signal synchronized with reading output from the reading control unit 5, a signal (inside / outside detection signal) RU indicating whether the pixel data output from the imaging unit 1 is for a pixel in the region is generated. . For example, the signal RU has a first value, for example, “1” when it is within the region, and a second value, for example, “0” when it indicates that it is outside the region.
When the specific area is composed of pixels in the vicinity of the missing pixel, the area is defined by the maximum distance represented by the number of pixels from the missing pixel. For example, when the specific area is configured by the i th pixel counted from the pixel adjacent to the missing pixel, the specific area is defined by the number “i”. This number “i” is stored in, for example, the area specifying unit 14.

  The correction data generation unit 13 determines whether or not to generate correction data for each pixel and supplies the correction data to the correction calculation unit 12 based on the region inside / outside detection signal RU from the region specifying unit 14. More specifically, the sensitivity correction data generation unit 24 outputs the pixel signal BK that is output from the black correction unit 21 and supplied via the switching unit 11b when the area inside / outside detection signal RU is “1”. Sensitivity correction data is generated, and sensitivity correction data is generated for the pixel signal BK that is output from the black correction unit 21 and supplied via the switching unit 11b when the intra-region detection signal RU is “0”. Do not do. In the image reading mode, the correction calculation unit 12 for the sensitivity correction data is output for the pixel signal BK output from the black correction unit 21 and supplied via the switching unit 11b when the intra-region detection signal RU is “1”. Is supplied from the black correction unit 21 when the area inside / outside detection signal RU is “0”, and for the pixel signal BK supplied through the switching unit 11b, a correction calculation unit for sensitivity correction data. 12 is not supplied.

  The interpolation unit 4 interpolates missing pixels using the pixel data WH output from the correction calculation unit 12, and outputs pixel data IP including the interpolation data.

  FIG. 2 is a diagram illustrating the imaging unit 1. The illustrated imaging unit 1 includes a contact image sensor 31. The contact image sensor 31 includes a plurality of, for example, a sensor chips 33 (1) to 33 (a), a lens 34, and a light source 35 that are connected to each other on a substrate 32 and arranged in a row. .

FIG. 3 is a diagram showing the sensor chip 33 (1), the sensor chip 33 (2), and a part of the sensor chip 33 (3) of the contact image sensor 31 in more detail.
Each of the sensor chips 33 (1), 33 (2), and 33 (3) includes a plurality of, for example, b imaging elements 36 (1) to 36 (b) as illustrated. Each of the image sensors 36 (1) to 36 (b) corresponds to one pixel.
It should be noted that the reference “33” may be used when explaining common to all of the plurality of sensor chips, and similarly, the reference “36” is used when explaining common to all of the plurality of image sensors. May be used.

The image sensors 36 (1) to 36 (b) on the sensor chip 33 are arranged so as to be aligned with each other in a line on the straight line at the interval dw1. In addition, the image pickup elements located at the adjacent end portions of the two sensor chips 33 that are connected to or connected to each other are separated from each other by a distance dw2.
For example, the image sensor 36 (b) at the right end of the sensor chip 33 (1) at one end (the left end in FIG. 3) of the image sensor 31 and the left end of the second sensor chip 33 (2) from the one end of the image sensor 31. The distance of the image sensor 36 (1) is dw2.
Here, the sensor chip is arranged so that dw2 is approximately twice as large as dw1, and interpolation processing is performed assuming that there is one missing pixel between the sensor chips, so that a high-quality interpolation result can be obtained. However, the present invention is not necessarily limited to this, and the interpolation processing according to the distance may be performed by setting dw2 to be other than twice dw1. good.

The outline of the interpolation processing in the interpolation unit 4 will be described below with reference to FIGS. 2 and 3 and FIGS. 4 and 5 described above.
FIG. 4 illustrates the sensor chip 33 and the position of the image sensor 36 arranged on the sensor chip 33 in the main scanning direction (column direction), and pixel data corresponding to each position (pixel data output from the sensitivity correction unit 22). FIG. 4 is a diagram schematically illustrating a relationship between WH and pixel data (pixel data output from the interpolation unit 4) IP. In FIG. 4, a case where eight image sensors 36 are arranged in each sensor chip 33, that is, a case where b = 8 will be described.

  In the figure, WH (1) to WH (8) are pixel data corresponding to the image sensors 36 (1) to 36 (8) of the sensor chip 33 (1), and WH (10) to WH (17) are sensors. This is pixel data corresponding to the image sensors 36 (1) to 36 (8) of the chip 33 (2). Since the image sensor 36 (8) of the sensor chip 33 (1) and the image sensor 36 (1) of the sensor chip 33 (2) are arranged at an interval of dw2 larger than dw1, as described above, the pixel data It is assumed that there is no image sensor corresponding to WH (9), that is, the pixel corresponding to the pixel data WH (9) is missing in the image signal WH. For example, it is assumed that dw2 is twice dw1, and one pixel is missing between the image sensor 36 (8) of the sensor chip 33 (1) and the image sensor 36 (1) of the sensor chip 33 (2). deal with.

  The interpolation unit 4 outputs the pixel data IP (9) after interpolation by calculating the missing pixel data WH (9) by interpolation calculation. The interpolation unit 4 outputs the pixel data IP having no missing pixels by performing this interpolation process on all the missing pixel data between the sensor chips 33.

FIG. 5 is a diagram illustrating an operation when the interpolation unit 4 performs an interpolation process on a document that gradually becomes brighter from left to right in the main scanning direction (horizontal direction in the drawing) at the position indicated as “reading position”.
5A shows the above-described original, FIG. 5B shows the brightness of the original at the reading position shown in FIG. 5A, and FIG. 5C shows the sensor chip 33 for the original. FIG. 5D shows the positional relationship in the main scanning direction, and FIG. 5D shows the image signal output from the sensor chip 33 shown in FIG. 5C by the A / D conversion unit 2 and the characteristic correction unit 3. FIG. 5 is a diagram showing the brightness of the pixel data WH subjected to the A / D conversion and characteristic correction described above, and shows a state where the pixel data WH (9) and the pixel data WH (17) are missing as described above. . FIG. 5E is a diagram showing a result of the interpolation processing performed on the pixel data shown in FIG. 5D by the interpolation unit 4. The pixel data WH (9) and the pixel data WH ( 17) shows a state in which the data IP (9) and the data IP (17) are interpolated.
In FIG. 5, FIG. 7 and FIG. 8, which will be described later, and descriptions related thereto, the characteristic correction unit 3 corrects all pixels.

  As illustrated in FIGS. 5D and 5E, the interpolation unit 4 interpolates missing pixel data at positions (between the respective sensor chips 33) where the image sensor 36 is widely spaced.

  The interpolation unit 4 is configured as shown in FIG. 6, for example. The illustrated interpolation unit 4 includes a left / right average interpolation calculation circuit 41 (0), first to nth average maintenance interpolation calculation circuits 41 (1) to 41 (n), a selection signal generation unit 42, and an output circuit 43. And have. The selection signal generation unit 42 includes a management circuit 44 and a scoring circuit 45.

  The pixel data WH includes a left-right average interpolation calculation circuit 41 (0) and n average maintenance interpolation calculation circuits, that is, a first average maintenance interpolation calculation circuit 41 (1), a second average maintenance interpolation calculation circuit 41 (2), ... are input to the nth average preserving interpolation calculation circuit 41 (n) and the scoring circuit 45.

  The left / right average interpolation calculation circuit 41 (0) generates an average value of the values of adjacent pixels located on the left and right of the interpolation target pixel as interpolation data. Note that “left and right” here means that when a series of pixels are arranged in the horizontal direction, they are positioned to the left and right of the pixel to be interpolated. The thing located in "" corresponds to this.

Each of the first to n-th average maintenance interpolation calculation circuits 41 (1) to 41 (n) is located in the vicinity of the interpolation target pixel, and an average value of a plurality of pixels including the interpolation target pixel and the interpolation target pixel Interpolation is performed so that the average values of a plurality of pixels not including the pixel values are equal to each other, and pixel data of the interpolation target pixel is obtained. For example, an average value of pixel data of a set of pixels including a predetermined number of pixels including an interpolation target pixel is equal to an average value of pixel data of a set of pixels including a predetermined number of pixels not including the interpolation target pixel. Perform interpolation.
For example, a predetermined number of pixels including the interpolation target pixel and a predetermined number of pixels not including the interpolation target pixel may be selected so as to partially overlap each other, or may be selected so as not to overlap.
As data designating the predetermined number, parameters k = k1 to kn are given to the first average maintenance interpolation calculation circuit 41 (1) to the nth average maintenance interpolation calculation circuit 41 (n), respectively.

  Interpolation is performed by the left / right average interpolation calculation circuit 41 (0) and the average preserving interpolation calculation circuits 41 (1) to 41 (n), and the results are input to the scoring circuit 45 and the test interpolation is performed to obtain the optimum. Select an appropriate interpolation method and perform interpolation.

  In the test interpolation, interpolation data is generated by each interpolation circuit using a pixel (non-missing pixel) having a known pixel value (value of pixel data WH) located near the missing pixel as an interpolation target pixel. Interpolation data is compared with a known pixel value, and it is determined that an interpolation circuit that generates interpolation data having the smallest difference from the pixel value can achieve the best result, and an interpolation operation circuit is selected based on the determination result. . In this way, the interpolation unit 4 evaluates the interpolation result when interpolation is performed by each interpolation circuit, and performs interpolation of missing pixels using the interpolation calculation circuit that has produced the best interpolation result.

Next, the overall operation of the image reading apparatus will be described.
An image signal AS output from the contact image sensor 31 is input to the A / D converter 2. The contact image sensor 31 is controlled by a timing control signal from the reading control unit 5, so that the light source 35 is turned on and off, the image reading timing, and the output timing of the image signal AS are determined. The timing itself is not illustrated as AS is output at a predetermined timing.

  The sensor chips 33 (1) to 33 (a) provided in the contact image sensor 31 are configured such that a plurality of sensor chips are connected in a line in the main scanning direction according to the number of read pixels. The image signal AS is 1 by reading the values of the read pixels from the image sensor 36 (1) of the sensor chip 33 (1) to the image sensor 36 (b) of the sensor chip 33 (a) continuously or intermittently. A two-dimensional image signal is formed by repeatedly reading out the image signal of one line at a predetermined timing by forming a line image signal and further moving the contact image sensor 31 in the sub-scanning direction or moving the document. Form.

  This two-dimensional image signal is composed of a set of a plurality of pixels arranged in a matrix in the horizontal direction corresponding to the main scanning direction and the vertical direction corresponding to the sub-scanning direction.

The image signal AS output from the imaging unit 1 is sampled at the timing when the signal from each pixel is output by the A / D conversion unit 2, A / D converted, and digital data (pixel data) representing the value of each pixel. ) Become RD.
The pixel data RD output from the A / D conversion unit 2 is input to the characteristic correction unit 3, and correction data is generated in the correction data generation mode, and characteristic correction is performed in the pixel data reading mode. .

  As described above, the image reading apparatus shown in FIG. 1 can operate in the image reading mode, the black correction data generation mode, or the sensitivity correction data generation.

When generating black correction data, light is not input to each image sensor 36, and a signal output from the image sensor 36 at that time is input to the characteristic correction unit 3 via the A / D conversion unit 2. Since the image sensor 36 is shielded so as not to receive light from the outside, the above-described “state in which no light is input to each image sensor 36” can be created by not turning on the light source 35.
The characteristic correction unit 3 supplies the pixel data RD from the A / D conversion unit 2 to the black correction data generation unit 23 via the switching unit 11a.

  The black correction data generation unit 23 stores pixel data of several lines (m lines) (that is, pixel data of a predetermined number m lines up to the current line) RD in the data memory 25, and the predetermined number of lines for each pixel. The average value RDa (p) is calculated by the following formula (1). Here, p represents the number of one pixel. For example, the number is assigned in order from one end of one line including the missing pixel.

The average value RDa determined in this way is used as black correction data BC (p) for the pixel, that is, BC (p) = RDa (p)
Thus, the black correction data BC (p) is obtained and stored in the data memory 25. The black correction data BC (p) is stored in association with each pixel number p.

  When generating the sensitivity correction data, the light is reflected by the white surface and light is input to each image sensor 36, and the signal output from the image sensor 36 at that time is sent to the A / D converter 2. To the characteristic correction unit 3. As the “white surface”, for example, a white platen surface through which an original passes can be used.

The characteristic correction unit 3 supplies the pixel data RD from the A / D conversion unit 2 to the black correction unit 21 via the switching unit 11a, and the black correction unit 21 supplies the black correction supplied from the correction data generation unit 13. Black correction is performed using the data, and corrected pixel data BK obtained as a result is output.
That is, when pixel data RD (p) of each pixel is input from the A / D conversion unit 2 to the black correction unit 21 via the switching unit 11a, the black correction data BC (p) of the pixel is synchronized with this. Are read from the data memory 25 and supplied to the black correction unit 21.
The black correction unit 21 calculates the following equation (3) based on the pixel data RD (p) and the correction data BC (p) to obtain corrected pixel data BK (p).
BK (p) = RD (p) −BC (p) (3)
The corrected pixel data BK obtained in this way is supplied to the sensitivity correction data generation unit 24 via the switching unit 11b. The operation of the black correction unit 21 at this time is the same as the operation at the time of image reading, which will be described in detail later with reference to the drawings.

  The sensitivity correction data generation unit 24 stores several lines of pixel data BK in the data memory 25, and calculates an average value BKa (p) for several lines, for example, m lines, for each pixel by the following equation (2). . Here, p represents the number of the pixel. For example, the number is sequentially added from one end of one line including the missing pixel.

The average value BKa (p) thus determined is used as sensitivity correction data WC (p) for the pixel, that is, WC (p) = BK (p).
Thus, the sensitivity correction data WC (p) is obtained and stored in the data memory 25. Sensitivity correction data WC (p) is stored in association with each pixel number p.

After the correction data is determined as described above, the operation when the image to be corrected is read is as follows.
The pixel data RD from the A / D conversion unit 2 is supplied to the black correction unit 21 via the switching unit 11a, and the black correction unit 21 receives the correction data from the correction data generation unit 13 as in the case of the above-described sensitivity correction data generation. Black correction is performed using the supplied black correction data, and corrected pixel data BK obtained as a result is output.
That is, when pixel data RD (p) of each pixel is input from the A / D conversion unit 2 to the black correction unit 21 via the switching unit 11a, the black correction data BC (p) of the pixel is synchronized with this. Are read from the data memory 25 and supplied to the black correction unit 21.
The black correction unit 21 calculates the following equation (3) based on the pixel data RD (p) and the correction data BC (p) to obtain corrected pixel data BK (p).
BK (p) = RD (p) −BC (p) (3)

  FIG. 7 is a diagram for explaining the concept of the correction operation in the black correction unit 21. FIG. 7A shows a black read image in which black pixels continue in the main scanning direction (horizontal direction in the figure) at least at the position indicated as “reading position” on the original surface, and FIG. The brightness of the image at the “reading position” shown in FIG. 7A, FIG. 7C shows the positional relationship in the horizontal direction (main scanning direction) of the sensor chip corresponding to the read image, and FIG. An example of data RD indicating the brightness of the image read by the sensor chip, FIG. 7E, is the corrected pixel data BK output by the black correction unit 21.

  As shown in FIG. 7D, the image sensors on the sensor chip 33 have different data indicating black. The black correction unit 21 corrects this variation and works to uniformize data when a black image is read (or when light is not input to the image sensor 36) as shown in FIG. . That is, the black correction data generation unit 23 can generate the black correction data BC based on the pixel data RD output from the A / D conversion unit 2 when no light enters each image sensor 36, and this data Stored in the memory 25. The black correction data BC generated by the black correction data generation unit 23 is supplied to the black correction unit 21, and the black correction unit 21 performs the calculation of the above formula (3) using the black correction data BC. The black level of the image sensor 36 is corrected.

  Next, the corrected pixel data BK output from the black correction unit 21 is input to the sensitivity correction unit 22. The sensitivity correction unit 22 corrects variations in sensitivity of the image sensors 36 based on the input pixel data BK in accordance with sensitivity correction data WC generated by the sensitivity correction data generation unit 12. Here, the variation in sensitivity is the variation in the dynamic range of the output signal of each image sensor, that is, the level of the output signal when light is reflected from the light source 35 on the white surface and light is input to each image sensor 36. , It means a variation in the difference from the level of the output signal when no light is input. In this case, the influence of the variation in the amount of light emitted from the light source 35 is also included in the sensitivity variation described here.

When pixel data BK (p) subjected to black correction for each pixel is input from the black correction unit 21 to the sensitivity correction unit 22 via the switching unit 11b, the sensitivity correction data WC (p) for that pixel is synchronized with this. ) Is read from the data memory 25 and supplied to the sensitivity correction unit 22.
The sensitivity correction unit 22 calculates the following equation (4) based on the pixel data BK (p) and the sensitivity correction data WC (p) to obtain pixel data WH (p) whose sensitivity has been corrected.
WH (p) = BK (p) * TG / WC (p) (4)
Here, TG is a gradation value indicating white. For example, when the gradation value indicating white is represented by 8 bits, the TG is the maximum value 255.

  Note that the sensitivity correction data WC used in the calculation of Expression (4) is 1 bit more than the pixel data BK after black correction and the number of bits (8 bits) necessary to represent the gradation value TG indicating white. Although it is necessary to use a bit number (9 bits), it is only necessary to store 8-bit correction data and generate 9-bit correction data at the time of calculation.

  FIG. 8 is a diagram for explaining the concept of the correction operation in the sensitivity correction unit 22. FIG. 8A shows a white read image in which white pixels continue in the main scanning direction (horizontal direction in the drawing) at least at the position indicated as “reading position” on the document surface, and FIG. The brightness of the image at the reading position shown in FIG. 8A, FIG. 8C shows the positional relationship of the sensor chip in the horizontal direction (main scanning direction) with respect to the read image, and FIG. 8D shows the sensor chip. FIG. 8E shows the corrected pixel data WH output by the sensitivity correction unit 22 and indicates the brightness of the image read by BK.

  As shown in FIG. 8D, the image sensors 36 on the sensor chip 33 have different data indicating white. The sensitivity correction unit 22 corrects this variation and works to uniformize data when a white image is read as shown in FIG. That is, the sensitivity correction data generation unit 24 generates the sensitivity correction data WC based on the pixel data BK obtained by performing black correction processing on the pixel data RD output from the A / D conversion unit 2 when a white image serving as a reference is read. It can be generated and stored in the data memory 25. The sensitivity correction data WC generated by the sensitivity correction data generation unit 24 is supplied to the sensitivity correction unit 22, and the sensitivity correction unit 22 performs the calculation of the above equation (2) using the sensitivity correction data WC. The sensitivity of the image sensor 36 is corrected.

  In the above description, it is assumed that the number of pixels on each sensor chip is 8, but the actual sensor chip often has more than 8. In the following description, a case where there are more than 8, for example, 16 is assumed.

As described above, the case where sensitivity correction is performed for all pixels has been described with reference to FIG. 8, but in this embodiment, sensitivity correction is performed only for pixels in a specific region. This specific area is a pixel in the vicinity of the missing pixel, and is an area composed of a part or all of the pixels used in the interpolation calculation in the interpolation unit 4.
The interpolation unit 4 supplies a signal IPN indicating the number of pixels used for the interpolation calculation, for example, to the region specifying unit 14 of the characteristic correction unit 3. As used herein, “pixels used for interpolation calculation” means that when the interpolation unit 4 includes a plurality of interpolation calculation circuits 41 (0) to 41 (n) as shown in FIG. Therefore, the pixel used for the interpolation calculation corresponds to this, which performs the interpolation calculation using the largest number of pixels among the plurality of average preserving interpolation calculation circuits.

  The region designating unit 14 includes a region determining unit 14a, a region storage unit 14b, and a region inside / outside detecting unit 14c. The region determination unit 14a determines a specific region based on the signal IPN from the interpolation unit 4. The specific area is defined by the number of pixels as described above, for example. The specific area may be determined so as to include all of the pixels used for the interpolation calculation, or determined so as to include only a part of the pixels used for the interpolation calculation, for example, a predetermined ratio. Alternatively, only a predetermined number may be included.

  Data representing a specific area determined by the area determination unit 14a is stored in an area storage unit 14b configured by, for example, a register.

  The area inside / outside detection unit 14c receives a signal (a signal indicating which pixel data is being read) synchronized with the reading from the reading control unit 5, and the pixel data output from the A / D conversion unit 2 at that time Is the pixel in the specific area represented by the data stored in the area storage unit 14b. If the pixel is in the area, the value of the area inside / outside detection signal RU is set to “ If the pixel is outside the area, the value of the inside / outside area detection signal RU is set to “0”.

  The sensitivity correction data generation unit 24 generates sensitivity correction data based on the inside / outside region detection signal RU. That is, when the area inside / outside detection signal RU is “1”, the sensitivity correction data generation unit 24 generates sensitivity correction data, and when the area inside / outside detection signal RU is “0”, the sensitivity correction data generation unit 24. Does not generate sensitivity correction data.

  FIG. 9 is a diagram for explaining a concept of an operation in which sensitivity correction is performed only on i pixels (i i in the front and i in the back) before and after the missing pixel (i = 4 in FIG. 9) in the sensitivity correction unit 22. It is. FIG. 9A shows a white read image in which white pixels continue throughout the main scanning direction (horizontal direction in the figure) at least at the position indicated as “reading position” on the document surface, and FIG. The brightness of the image at the reading position shown in FIG. 9A, FIG. 9C shows the positional relationship of the sensor chip in the horizontal direction (main scanning direction) with respect to the read image, and FIG. 9D shows the sensor chip. The pixel data BK after black correction corresponding to the brightness of the image read by, FIG. 9E is the pixel data WH after correction output from the sensitivity correction unit 22.

  Since the pixels in the specific areas C1 to C5 (4 pixels each) perform sensitivity correction in units of pixels, the pixel values after sensitivity correction shown in FIG. 9E are equal to each other, and the maximum value (8 If the bit is “255”, the sensitivity correction is not performed in other regions, and therefore, the correction is not necessarily equal to each other, and the input (FIG. 9D) and output (FIG. 9D) of the sensitivity correction unit 22 are not necessarily equal to each other. 9 (e)) are equal to each other.

  Since correction data is generated only for pixels in a specific area and stored in the data memory, the relationship between the address of the area storing the correction data and the pixel number is not continuous, and the pixel number It is necessary to store the correspondence (mapping) between the address and the address separately.

  As described above, the generation and storage of sensitivity correction data is performed only for pixels in a specific area, for example, i pixels in the vicinity of missing pixels, and correction data is not generated for other pixels. Thus, the capacity of the data memory 25 can be reduced.

  In the above-described embodiment, the specific area is determined based on the signal IPN from the interpolation unit 4, but the outside, for example, another device connected to the image reading device, for example, the image reading device is a part. It may be specified from the control unit of the copying machine. In this case, as the region specifying unit, the one shown in FIG. 10 is used instead of the one shown in FIG. The area specifying unit includes an interface 14d, an area storage unit 14b, and an area inside / outside detection unit 14c. The area storage unit 14b and the area inside / outside detection unit 14c are the same as those shown in FIG. The interface unit 14d receives external data and stores it in the area storage unit 14b.

Embodiment 2. FIG.
In the first embodiment, sensitivity correction is performed only for pixels in a specific area, while black correction is performed for all pixels. In the second embodiment described below, black correction is performed. Similarly to the sensitivity correction, it is performed only for the pixels in the specific area.
The configuration of the apparatus in that case is as shown in FIG. That is, the apparatus shown in FIG. 11 is the same as the apparatus shown in FIG. 1, but the area inside / outside detection signal RU output from the area designating unit 14 generates not only the sensitivity correction data generation unit 24 but also the black correction data generation. Also supplied to the unit 23.

  The black correction data generation unit 23 generates correction data for the pixel signal output from the A / D conversion unit 2 and supplied via the switching unit 11a when the area inside / outside detection signal RU is “1”. The correction data is not generated for the pixel signal that is output from the A / D conversion unit 2 and supplied via the switching unit 11a when the intra-region detection signal RU is “0”.

FIG. 12 is a diagram for explaining a concept of an operation in which the black correction unit 21 performs black correction only on i pixels (i = front i and i on the back) before and after the missing pixel (i = 4 in FIG. 12). FIG. 12A is a black read image in which black pixels continue in the main scanning direction (horizontal direction in the figure) at least at the position labeled “reading position” on the document surface, FIG. FIG. 12B shows the brightness of the image at the reading position shown in FIG. 12A, FIG. 12C shows the positional relationship in the horizontal direction (main scanning direction) of the sensor chip corresponding to the read image, and FIG. d) is the pixel data RD output from the A / D conversion unit 2 corresponding to the brightness of the image read by the sensor chip, and FIG. 12E is the corrected pixel data output from the black correction unit 21. BK.
Since the pixels in the region C1 to C5 (4 pixels each) perform black correction on a pixel basis, the pixel values after black correction shown in FIG. 12E are equal to each other, and the minimum value of the gradation expression range, that is, Although it is zero, black correction is not performed in other regions, and therefore, they are not necessarily equal to each other, and the input (FIG. 12 (d)) and output (FIG. 12 (e)) of the black correction unit 21 are mutually equal. equal.

Embodiment 3 FIG.
In the first embodiment, the sensitivity correction is not performed at all for the pixels outside the specific region. However, in the third embodiment described below, the sensitivity of the pixels outside the specific region is set in units of a plurality of pixels. Make corrections. That is, for pixels outside the specific area, a plurality of adjacent pixels constitute one set (the pixels constituting each set do not constitute another set), and a common sensitivity correction is performed for this. Data is obtained, and sensitivity correction is performed on the set of pixels using the common sensitivity correction data.

In generating common sensitivity correction data for a plurality of pixels in each group, for example, the maximum of the average values of the pixel data BK calculated in the same manner as in the first embodiment for each of the plurality of pixels in each group. Are used as sensitivity correction data common to a plurality of pixels in the set.
In reading the pixel data, the data of the plurality of pixels is the same while being supplied from the A / D converter 2 to the sensitivity correction unit 22 via the switching unit 11a, the black correction unit 21, and the switching unit 11b. Sensitivity correction is performed using the correction data.

An example of the sensitivity correction data generation unit 24 and the sensitivity correction unit 22 that perform such processing is shown in FIG. The overall configuration of the apparatus is the same as in FIG.
The sensitivity correction data generation unit 70 shown in FIG. 13 includes an in-region sensitivity correction data generation unit 71 and an out-of-region sensitivity correction data generation unit 72. The in-region sensitivity correction data generation unit 71 has the same configuration and function as the sensitivity correction data generation unit 24 of FIG. The out-of-region sensitivity correction data generation unit 72 outputs the sensitivity correction data using the pixel data BK output from the black correction unit 21 and supplied via the switching unit 11b when the in-region / outside detection signal RU is “0”. Generate.

As shown in FIG. 14, for example, the out-of-region sensitivity correction data generation unit 72 includes an average value calculation unit 72a, an average value storage unit 72b, a maximum value selection unit 72c, and a set size storage unit 72d.
The set size storage unit 72d is configured by a register, for example, and stores the number nw of pixels constituting each set.

  Similar to the sensitivity correction data generation unit 24, the average value calculation unit 72a is configured to receive a predetermined number of lines (p) of each pixel (p) in a state where light from the light source is reflected by a white surface and light is incident on the image sensor. The average value BKa (p) of the pixel data BK over (m lines) is obtained by Expression (2). The average value storage unit 72b sequentially stores the average value BKa (p) obtained in this way.

The maximum value selection unit 72c selects the maximum one of the average values BKa (p) of the pixels constituting each set and stored in the average value storage unit 72b, and the common value of the pixels of the set is selected. Output as correction data WC.
In the process of reading the average value of each pixel constituting each set from the average value storage unit 72b and selecting the maximum value by the maximum value selection unit 72c, the number nw of pixels constituting each set is set to the set size storage unit 72d. Supplied from

Hereinafter, assuming that correction data is determined for each of two pixels, a case where the maximum of the average values BKa of the two pixels is determined as sensitivity correction data for each of the two pixels will be described in detail.
For example, for the pixel with pixel number p, the average value calculated in the same manner as in the first embodiment is BKa (p), and for the pixel with pixel number (p + 1), the average value calculated in the same manner as in the first embodiment is BKa (p + 1). ).
These maximum values MAX {BKa (p), BKa (p + 1)} are used as correction data WC (p, p + 1) common to the pixels p and (p + 1), that is,
WC (p, p + 1) = MAX {BKa (p), BKa (p + 1)}
WC (p, p + 1) is obtained by the above and stored in the data memory 25.

  At the time of image reading, the sensitivity correction unit 22 corrects both the data of the pixel p and the data of the pixel (p + 1) using the sensitivity correction data WC (p, p + 1). This is realized by the correction control unit 27 of the correction data generation unit 13 controlling the supply of correction data in this way. Therefore, the correction control unit 27 recognizes the pixel number corresponding to the pixel data currently input based on the timing signal from the reading control unit 5, and performs the above-described control based on this.

  FIG. 15 is a diagram for explaining the concept of an operation of performing sensitivity correction in units of two pixels for pixels other than the specific region in the sensitivity correction unit 22. In this case, it is assumed that one pair is constituted by two pixels (p-th pixel and p + 1-th pixel) adjacent to each other. FIG. 15A shows a white read image in which white pixels continue throughout the main scanning direction (horizontal direction in the figure) at least at the position indicated as “reading position” on the document surface, and FIG. The brightness of the image at the “reading position” shown in FIG. 15A, FIG. 15C shows the positional relationship of the sensor chip in the horizontal direction (main scanning direction) with respect to the read image, and FIG. The pixel data BK after black correction corresponding to the brightness of the image read by the sensor chip, FIG. 15E is the pixel data WH after correction output from the sensitivity correction unit 22. In the case of FIG. 15, the sensitivity correction is performed on the pixels in the specific areas C <b> 1 to C <b> 5 (each four pixels) in units of pixels, and in the other areas, the sensitivity correction is performed in units of 2 pixels as described above.

  In FIG. 15D, the larger average value over m lines of each set of pixel data BK output from the black correction unit 21 is the sensitivity correction data common to the two pixels of the set. . Therefore, unlike FIG. 8E, as shown in FIG. 15E, the data after correction of the pixels (representative pixels) corresponding to the average value BKa adopted as the sensitivity correction data WC is not uniform. , Which coincides with the corrected data of similar pixels (representative pixels) in other groups and becomes the maximum value of the gradation expression range (“255” if 8 bits), but correction of the other pixel in each group The subsequent data does not always match, and the corrected data has a difference corresponding to the difference in the correction data with the representative pixel of the same set.

  In such a configuration, the memory capacity for storing sensitivity correction data is larger than that of the first embodiment, but compared with the case where correction data is generated and stored individually for all pixels. , Memory capacity can be reduced. Further, as compared with the first embodiment, since correction is performed on pixels other than the specific region, an image closer to an image obtained when correction is performed on all pixels can be obtained.

  In the above embodiment, two adjacent pixels form one set, and common correction data is generated for the set of pixels. However, the number of pixels constituting each set nw May be other than 2, for example, limited by the available capacity of the data memory 25, the number of image sensors per sensor chip, the number of sensor chips constituting the image reading device, etc., but can be changed within the limits It is good. For example, as shown in FIG. 16, a set size storage unit 72d for storing the number of pixels constituting each set is connected to the outside of the apparatus via an interface unit 72e so that it can be rewritten from the outside. Also good.

  As described above, the third embodiment has been described as a modification to the first embodiment. However, the same modifications as those described in the third embodiment can be applied to the second embodiment.

Embodiment 4 FIG.
In the second embodiment, black correction is not performed at all for pixels outside the specific area. However, in the fourth embodiment described below, sensitivity correction in the third embodiment is performed on pixels outside the specific area. Similarly to the above, black correction is performed in units of a plurality of pixels. That is, for pixels outside the specific region, a plurality of adjacent pixels constitutes one set, common black correction data is obtained for this, and the common black correction data is used to determine the pixels of the set. For black correction.

In generating the common black correction data for the plurality of pixels in each group, for example, the minimum of the average values of the pixel data RD calculated in the same manner as in the second embodiment for each of the plurality of pixels in each group. Are used as common black correction data for a plurality of pixels in the set.
In reading the pixel data and generating the sensitivity correction data, the same correction data is supplied while the data of the plurality of pixels is supplied from the A / D conversion unit 2 to the black correction unit 21 via the switching unit 11a. Using this, black correction is performed.

An example of the black correction data generation unit 23 and the black correction unit 21 that perform such processing is shown in FIG. The overall configuration of the apparatus is the same as that shown in FIG.
The black correction data generation unit 60 illustrated in FIG. 17 includes an in-region black correction data generation unit 61 and an out-of-region black correction data generation unit 62. The in-region black correction data generation unit 61 has the same configuration and function as the black correction data generation unit 23 of FIG. The out-of-region black correction data generation unit 62 uses the pixel data RD that is output from the A / D conversion unit 2 and supplied via the switching unit 11a when the in-region / outside detection signal RU is “0”. Generate data.

For example, as shown in FIG. 18, the out-of-region sensitivity correction data generation unit 62 includes an average value calculation unit 62a, an average value storage unit 62b, a minimum value selection unit 62c, and a set size storage unit 62d.
The set size storage unit 62d is constituted by a register, for example, and stores the number nb of pixels constituting each set.

  Similar to the black correction data generation unit 23, the average value calculation unit 62a is an average value RDa (p) of the pixel data RD over a predetermined number of lines (m lines) of each pixel (p) in a state where light is not incident on the image sensor. Is obtained by the equation (1). The average value storage unit 62b sequentially stores the average value RDa (p) obtained in this way.

The minimum value selection unit 62c selects the minimum one of the average values RDa (p) of the pixels constituting each set and stored in the average value storage unit 62b, and the common value of the pixels of the set is selected. Output as correction data BC.
In the process of reading the average value of each pixel constituting each set from the average value storage unit 62b and selecting the minimum value by the minimum value selection unit 62c, the number nb of pixels constituting each set is determined by the set size storage unit 62d. Supplied from

Hereinafter, the case where the correction data is determined for each of the two pixels and the minimum of the average value RDa of the two pixels is determined as the black correction data for each of the two pixels will be described in detail.
For example, for the pixel with pixel number p, the average value calculated in the same manner as in the second embodiment is RDa (p), and for the pixel with pixel number (p + 1), the average value calculated in the same way as in the second embodiment is RDa (p + 1 ).
These minimum values MIN {RDa (p), RDa (p + 1)} are used as correction data BC (p, p + 1) common to the pixel p and the pixel (p + 1), that is,
BC (p, p + 1) = MIN {RDa (p), RDa (p + 1)}
BC (p, p + 1) is obtained by the above and stored in the data memory 25.

  At the time of image reading and sensitivity correction data generation, the black correction unit 21 corrects both the pixel p data and the pixel (p + 1) data using the black correction data BC (p, p + 1). This is realized by the correction control unit 27 of the correction data generation unit 13 controlling the supply of correction data in this way. Therefore, the correction control unit 27 recognizes the pixel number corresponding to the pixel data currently input based on the timing signal from the reading control unit 5, and performs the above-described control based on this.

  FIG. 19 is a diagram for explaining a concept of an operation in which black correction is performed in units of two pixels for pixels other than the specific region in the black correction unit 21. In this case, it is assumed that one pair is constituted by two pixels (p-th pixel and p + 1-th pixel) adjacent to each other. FIG. 19A shows a black read image in which black pixels continue in the main scanning direction (horizontal direction in the figure) at least at the position indicated as “reading position” on the document surface, and FIG. The brightness of the image at the “reading position” shown in FIG. 19A, FIG. 19C shows the positional relationship of the sensor chip in the horizontal direction (main scanning direction) with respect to the read image, and FIG. The pixel data RD output from the A / D conversion unit 2 corresponding to the brightness of the image read by the sensor chip, FIG. 19E is the corrected pixel data BK output from the black correction unit 21. In the case of FIG. 19, the pixels in the specific areas C1 to C5 (4 pixels each) are subjected to black correction in units of pixels, and in other areas, black correction is performed in units of 2 pixels as described above.

  In FIG. 19D, the smaller of the average value over m lines of each set of pixel data RD output from the A / D converter 2, the black correction data common to the two pixels of the set. It becomes. Therefore, unlike FIG. 7E, as shown in FIG. 19E, the data after correction of the pixel (representative pixel) corresponding to the average value RDa adopted as the black correction data BC is not uniform. , Which matches the corrected data of similar pixels (representative pixels) in other sets and becomes the minimum value (zero) of the gradation expression range, but the corrected data of the other pixels in each set does not match Not limited to this, the corrected data has a difference corresponding to the difference in correction data with the representative pixel of the same set.

  In such a configuration, the capacity of the memory for storing the black correction data is larger than that of the second embodiment, but compared with the case where the correction data is individually generated and stored for all the pixels. , Memory capacity can be reduced. Further, as compared with the second embodiment, since correction is performed on pixels other than the specific region, an image closer to an image obtained when correction is performed on all pixels can be obtained.

  In the above embodiment, two adjacent pixels constitute one set, and common correction data is generated for the set of pixels. However, the number of pixels constituting each set nb May be other than 2, for example, limited by the available capacity of the data memory 25, the number of image sensors per sensor chip, the number of sensor chips constituting the image reading device, etc., but can be changed within the limits It is good. For example, as shown in FIG. 20, a set size storage unit 62d for storing the number of pixels constituting each set is connected to the outside of the apparatus via an interface unit 62e and is configured to be rewritable from the outside. Also good.

  As described above, the fourth embodiment has been described as a modification of the second embodiment. However, the same modifications as those described in the fourth embodiment can be applied to the third embodiment.

  As described above, the case where the correction of the variation of the imaging element includes both the correction of the variation of the black level and the variation of the sensitivity has been described. However, the present invention corrects only one of them or collects them as one correction circuit. In this case, the image reading apparatus operates in either the image reading mode or the correction data generation mode. In the correction data generation mode, the correction data generation unit performs imaging. In the image reading mode, the correction data generation unit supplies the stored correction data to the correction calculation unit, and the correction calculation unit corrects the correction data. Using the correction data supplied from the generation unit, the pixel data output from the A / D conversion unit is corrected, and the corrected pixel data is output. To become.

  The characteristic correction unit 3 and the interpolation unit 4 have been described as being configured by hardware. However, these may be configured by software, that is, a programmed computer.

1 is a block diagram illustrating a configuration of an image reading apparatus according to a first embodiment. It is the schematic which shows the imaging part of FIG. It is the schematic which expands and shows a part of imaging part of FIG. It is a figure which shows the relationship between the image pick-up element and a missing pixel in the imaging part of FIG. It is a figure which shows the concept of the interpolation process of the missing pixel shown in FIG. It is a block diagram which shows the detail of the interpolation part of FIG. It is a figure which shows the operation | movement of the black correction | amendment of FIG. It is a figure which shows the operation | movement of the sensitivity correction | amendment of FIG. FIG. 10 is a diagram illustrating an operation of sensitivity correction in the first and second embodiments. FIG. 12 is a block diagram illustrating another configuration example of the area specifying unit of the characteristic correction unit according to the first and second embodiments. 6 is a block diagram illustrating an image reading apparatus according to a second embodiment. FIG. FIG. 10 is a diagram illustrating a black correction operation in the third and fourth embodiments. 10 is a block diagram illustrating a sensitivity correction data generation unit and a sensitivity correction unit according to Embodiment 3. FIG. It is a block diagram which shows the structure of the sensitivity correction data generation part 72 of FIG. FIG. 10 is a diagram illustrating sensitivity correction operation according to the third embodiment. It is a block diagram which shows the other example of the sensitivity correction data generation part 72 of FIG. FIG. 10 is a block diagram illustrating a black correction data generation unit and a sensitivity correction unit in the fourth embodiment. It is a block diagram which shows the black correction data generation part 62 of FIG. FIG. 10 is a diagram illustrating a black correction operation in the fourth embodiment. It is a block diagram which shows the other example of the black correction data generation part 62 of FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Image pick-up part, 2 A / D conversion part, 3 Characteristic correction part, 4 Interpolation part, 5 Reading control part, 11a, 11b Switching part, 12 Correction calculating part, 13 Correction data generation part, 14 Area designation part, 14a Area determination Unit, 14b region storage unit, 14c region inside / outside detection unit, 14d interface unit, 21 black correction unit, 22 sensitivity correction unit, 23 black correction data generation unit, 24 sensitivity correction data generation unit, 25 data memory, 27 control unit, 31 Contact image sensor, 33 sensor chip, 35 light source, 36 image sensor, 41 (0) left-right average interpolation calculation circuit, 41 (1) to 41 (n) average maintenance interpolation calculation circuit, 42 selection signal generation unit, 43 output circuit, 44 management circuit, 45 scoring circuit, 60 black correction data generation unit, 61 in-region black correction data generation unit, 62 out-of-region black correction data generation Parts, 70 sensitivity correction data generation unit, 71 region sensitivity correction data generation unit 72 outside the region sensitive black correction data generation unit.

Claims (22)

An imaging unit in which a plurality of sensor chips having a plurality of imaging elements are connected to each other and arranged in a row,
An A / D converter that converts an image signal output from the imaging unit into digital pixel data;
A characteristic correction unit for correcting variations in characteristics of the image sensor of the sensor chip;
An interpolation unit that interpolates the lack of pixel data that occurs at the connection parts of the plurality of sensor chips using pixels in the vicinity thereof,
The image reading apparatus according to claim 1, wherein the characteristic correction unit performs the correction on pixels in a specific region including part or all of pixels used for interpolation by the interpolation unit.   The image reading apparatus according to claim 1, wherein the characteristic correction unit corrects variation in characteristics of the image sensor not only for pixels in the specific area but also for other pixels. The characteristic correction unit includes:
For pixels in the specific area, correction data is obtained for each pixel, and correction is performed using the correction data.
For pixels other than the pixels in the specific region, common correction data is obtained for a set of pixels composed of a plurality of adjacent pixels, and correction is performed using the common correction data. The image reading apparatus according to claim 2.   The image reading apparatus according to claim 1, wherein the characteristic correction unit performs black level variation correction of the image sensor and sensitivity variation of the image sensor.   The image reading apparatus according to claim 4, wherein the characteristic correction unit corrects the black level variation of the image sensor not only for pixels in the specific area but also for other pixels.   The characteristic correction unit obtains correction data for correcting variations in black level for individual pixels as well as pixels in the specific region, and uses the correction data. The image reading apparatus according to claim 5, wherein black level variation is corrected. The characteristic correction unit includes:
For the pixels in the specific region, correction data for correcting the black level variation is obtained for each pixel, and the black level variation is corrected using the correction data.
For pixels other than the pixels in the specific region, correction data for correcting black level variation common to a plurality of adjacent pixels is obtained, and the black level is corrected using the common correction data. The image reading apparatus according to claim 5, wherein variation is corrected. The characteristic correction unit includes:
For pixels in the specific region, obtain correction data for correction of sensitivity variation for each pixel, perform correction of sensitivity variation using the correction data,
For pixels other than the pixels in the specific region, correction data for correcting sensitivity variations that is common to a plurality of adjacent pixels is obtained, and sensitivity variations using the common correction data are obtained. The image reading apparatus according to claim 5, wherein correction is performed.   In the interpolation unit, the pixel data missing at the connection part between the plurality of sensor chips is determined so that the average value of the plurality of pixels including the missing pixel is equal to the average value of the plurality of pixels not including the missing pixel. Are interpolated in order to obtain pixel data of missing pixels, evaluate the interpolation results when performing interpolation with each interpolation calculation circuit, and use the interpolation calculation circuit that produced the best interpolation result The image reading apparatus according to claim 1, wherein the missing pixel is interpolated.   The characteristic correction unit includes an area storage unit that stores data representing the specific area, and the data representing the specific area stored in the area storage unit can be rewritten from the outside. The image reading apparatus according to claim 1.   The characteristic correction unit has a set size storage unit that stores the number of pixels constituting the pixel set, and data representing the number of pixels constituting the set stored in the set size storage unit is external The image reading apparatus according to claim 1, wherein the image reading apparatus is rewritable. An imaging unit in which a plurality of sensor chips having a plurality of imaging elements are connected to each other and arranged in a row,
A signal processing method in an image reading apparatus including an A / D conversion unit that converts an image signal output from an imaging unit into digital pixel data,
A characteristic correction step for correcting variations in characteristics of the image sensor of the sensor chip;
An interpolation step of interpolating the lack of pixel data that occurs at the connection parts of the plurality of sensor chips using pixels in the vicinity thereof,
In the signal correction method, the characteristic correction step performs the correction on pixels in a specific region including part or all of the pixels used for interpolation in the interpolation step. The characteristic correction step includes
The signal processing method according to claim 12, wherein the correction of the variation in characteristics of the image sensor is performed not only on the pixels in the specific area but also on other pixels. The characteristic correction step includes
For pixels in the specific area, correction data is obtained for each pixel, and correction is performed using the correction data.
For pixels other than the pixels in the specific region, common correction data is obtained for a set of pixels composed of a plurality of adjacent pixels, and correction is performed using the common correction data. The signal processing method according to claim 13.   The signal processing method according to claim 12, wherein the characteristic correction step performs correction of variation in black level of the image sensor and correction of variation in sensitivity of the image sensor.   The signal processing method according to claim 15, wherein in the characteristic correction step, correction of black level variation of the image sensor is performed not only on pixels in the specific area but also on other pixels.   In the characteristic correction step, not only the pixels in the specific region but also other pixels are obtained correction data for correcting the black level variation for each pixel, and the correction data is used. The signal processing method according to claim 16, wherein the black level variation is corrected. The characteristic correction step includes
For the pixels in the specific region, correction data for correcting the black level variation is obtained for each pixel, and the black level variation is corrected using the correction data.
For pixels other than the pixels in the specific region, correction data for correcting black level variation common to a plurality of adjacent pixels is obtained, and the black level is corrected using the common correction data. The signal processing method according to claim 16, wherein variation is corrected. The characteristic correction step includes
For pixels in the specific region, obtain correction data for correction of sensitivity variation for each pixel, perform correction of sensitivity variation using the correction data,
For pixels other than the pixels in the specific region, correction data for correcting sensitivity variations that is common to a plurality of adjacent pixels is obtained, and sensitivity variations using the common correction data are obtained. The signal processing method according to claim 16, 17, or 18, wherein correction is performed.   In the interpolation step, pixel data missing at the connection parts of the plurality of sensor chips is determined such that an average value of a plurality of pixels including a missing pixel is equal to an average value of a plurality of pixels not including a missing pixel. Are interpolated in order to obtain pixel data of missing pixels, evaluate the interpolation results when performing interpolation with each interpolation calculation circuit, and use the interpolation calculation circuit that produced the best interpolation result The signal processing method according to claim 12, wherein interpolation of the missing pixels is performed.   13. The signal processing method according to claim 12, further comprising a step of supplying data representing the specific area from the outside of the image reading apparatus.   The signal processing method according to claim 12, further comprising a step of supplying data representing the number of pixels constituting the set of pixels from the outside of the image reading apparatus.

Images