JPH06296234A - Image reader - Google Patents

Abstract

(57) [Abstract] [PROBLEMS] To provide an image reading device which is a one-line format and which can obtain a high-accuracy, high-gradation, high-density reading output by smoothly connecting overlapping reading areas. . A beam splitting means 3 is provided which divides the focused light 2 from the imaging lens 1 into two parts with a substantially central portion in the main scanning direction as a boundary, and right and left of the original surface in the main scanning direction on each of the two divided optical paths. A / D for arranging line sensors 4 and 5 for detecting the image of
The conversion means 10 and 11 are provided, and the shading correction means 14 and 15 for performing the shading correction by the calculation of the digitally converted read output and the reference level are provided, and both the outputs corresponding to the same reading point on the document surface are provided at the subsequent stage. When the output of the line sensors 4 and 5 appears, the weighted calculation result of the output of each subsequent stage is read out, and when the output appears only on one line sensor (4 or 5), the output of the subsequent stage is read out. A calculation selecting means 20 is provided.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a field of a digital copying machine, an electronic file device, an image scanner as a computer peripheral, an input portion of a scanner device, etc., and particularly, it is read by using two line sensors for image reading. In this case, the present invention relates to an image reading apparatus capable of performing highly accurate and high-density reading by smoothly connecting the reading connection portions of these two line sensors.

[0002]

2. Description of the Related Art Conventionally, when a plurality of line sensors are used to divide an image in the main scanning direction to read an original,
It is necessary to connect the output signals of the respective read outputs so that the output format is the same as when one line sensor is used. It is necessary to make the joint portion inconspicuous in the output after such joint processing. Therefore, as an example of a device for performing a high-density image reading by using a plurality of line sensors to smooth the joint portion in the reading of the line sensor, for example, Japanese Patent Laid-Open No. 57-9168 discloses "Image Reading". Method ”. In this reading method, the image reading is performed by using two line sensors for each one scanning line while partially overlapping, and the two reading outputs are “weighted and added” according to the reading position, thereby performing the joining process. Is going.

[0003]

When an image is read using the two line sensors as described above, the output from the same read pixel during the duplicate reading generally differs due to the following factors. The first factor is
The difference in the output occurs due to the difference in the amount of light reaching the two line sensors (shading of the optical system). The second factor is that the output varies depending on the document position (accuracy of sensor mounting position) expected by the line sensor. As a third factor, a difference in output occurs due to a difference in characteristics of the line sensor itself (sensitivity, dark output, linearity, spectral sensitivity in the case of a color sensor, etc.). In the conventional example described above,
It is shown as a solution to the first factor, and is considered effective for the second and third factors. However, if higher gradations are desired, it must be said that the first to third factors are insufficiently addressed.

For example, in the shading of the optical system and the correction of the sensitivity variation among the pixels of the sensor, the reference surface (plate) of the document surface is read in advance and the reference level is made for one line, and the document reading output is normalized. Therefore, the gradation is generally improved, and the gradation is greatly improved. Therefore, even in the method of performing the joining process using a plurality of line sensors as in the above-described conventional example, the joining process as in the conventional example is not necessary if such a method of high gradation processing is used. However, the countermeasures against the second and third factors remain.

[0005]

According to a first aspect of the invention, the light transmitted or reflected by scanning the document surface is condensed by an imaging lens and is incident on a line sensor.
In an image reading apparatus that obtains image information by detecting an electric signal corresponding to the reflectance or the transmittance of the document as an image signal, the focused light from the imaging lens is used as a boundary at a substantially central portion in the main scanning direction. A light beam splitting means for splitting is provided, and line sensors for detecting left and right images in the main scanning direction of the document surface are arranged on the respective light paths divided by the light flux dividing means from the vicinity of the central portion in the main scanning direction. A / D conversion means for converting the output of these line sensors into a digital signal is provided, and shading correction means for performing shading correction by calculation of the read output digitally converted by these A / D conversion means and the reference level is provided. When an output corresponding to the same reading point on the document surface appears at the output of both the line sensors after the shading correction means. An operation selecting means to read an output of the subsequent stage is provided when the output only to the line sensor of the one and the weighting operation result read outputs of the respective subsequent stage appears.

According to the second aspect of the invention, the light transmitted or reflected by scanning the surface of the original is condensed by the imaging lens and is incident on the line sensor, and the light is reflected or transmitted according to the reflectance of the original. In an image reading apparatus that obtains image information by detecting an electric signal as an image signal, a light beam splitting unit is provided that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary. A line for detecting the left and right images of the document surface in the main scanning direction on the respective optical paths divided into two parts from the vicinity of the central portion in the main scanning direction by the means, and reading the odd-numbered and even-numbered read pixels by separate transfer sections. A sensor is provided, A / D conversion means for converting the output of these line sensors into a digital signal is provided, and a digitally converted read output is provided at the subsequent stage of these A / D conversion means. A shading correction unit that performs shading correction by calculation with a quasi level, a pixel output switching unit that converts an odd pixel output and an even pixel output into a single-row sequential signal corresponding to the read space, and the shading correction unit. When the output corresponding to the same reading point on the document surface placed in the subsequent stage appears in the outputs of both line sensors, the weighted calculation result of the output of each subsequent stage is set as the read output and the output is output to only one of the line sensors. When it appears, a calculation selecting means for making the output of the subsequent stage a read output is provided.

According to the third aspect of the present invention, the light transmitted or reflected by scanning the surface of the original is condensed by the imaging lens and is incident on the line sensor, and the light is reflected or transmitted according to the reflectance of the original. In an image reading apparatus that obtains image information by detecting an electric signal as an image signal, a light beam splitting unit is provided that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary. A line sensor having read pixel rows of a plurality of colors for detecting left and right images in the main scanning direction of the document surface is disposed on each of the optical paths divided into two parts from the vicinity of the central portion in the main scanning direction by the means. A / D conversion means for converting the output of the sensor into a digital signal is provided, and a shading is performed by a calculation of the digitally converted read output and the reference level at a stage subsequent to these A / D conversion means. Shading correction means for performing scanning correction, line delay means for correcting read line shifts in the sub-scanning direction of read pixel rows of a plurality of colors, and a plurality of colors placed after the shading correction means and after the line delay means. Of matrix scanning means for generating a new set of a plurality of signals from the output of the read pixel row, and the output corresponding to the same reading point on the document surface placed after the shading correction means of both line sensors. When the output appears, the weighted calculation result of the output of each subsequent stage is set as a read output, and when the output appears only at one of the line sensors, the output of the subsequent stage is set as a read output.

According to the fourth aspect of the present invention, the light transmitted or reflected by scanning the surface of the original is condensed by the image forming lens and is incident on the line sensor, and the light is reflected or transmitted according to the reflectance or the transmittance of the original. In an image reading apparatus that obtains image information by detecting an electric signal as an image signal, a light beam splitting unit is provided that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary. Line sensors for detecting left and right images of the document surface in the main scanning direction are disposed on the respective optical paths divided into two parts from the vicinity of the central portion in the main scanning direction by means, and the outputs of these line sensors are converted into digital signals. A / D converting means are provided, and when outputs corresponding to the same reading point on the document surface appear in the outputs of both line sensors in the subsequent stage of these A / D converting means, respectively. Provided is an operation selection means for making the addition result of the output of the subsequent stage the read output and the output of the subsequent stage the read output when the output appears only on the one line sensor, and the read output and the reference are provided at the stage subsequent to the operation selecting means. A shading correction means for performing shading correction by calculation with the level is provided.

According to the fifth aspect of the invention, the light transmitted or reflected by scanning the surface of the original is condensed by the imaging lens and is incident on the line sensor, and the light is reflected or transmitted according to the reflectance or the transmittance of the original. In an image reading apparatus that obtains image information by detecting an electric signal as an image signal, a light beam splitting unit is provided that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary. A line for detecting the left and right images of the document surface in the main scanning direction on the respective optical paths divided into two parts from the vicinity of the central portion in the main scanning direction by the means, and reading the odd-numbered and even-numbered read pixels by separate transfer sections. A sensor is provided, A / D conversion means for converting the output of these line sensors into a digital signal is provided, and an odd pixel output and an even pixel output are provided in the subsequent stage of these A / D conversion means. Pixel output switching means for converting into one column of sequential signals corresponding to the captured space is provided, and the output corresponding to the same reading point on the document surface is output from both of the line sensors after the pixel output switching means. If the result of addition of the output of each subsequent stage is a read output, and the output of only one of the line sensors appears as an output of the read after that, a calculation selecting means is provided. A shading correction means for performing shading correction by calculating the read output and the reference level is provided.

According to the sixth aspect of the present invention, the light transmitted or reflected by scanning the surface of the original is condensed by the imaging lens and is incident on the line sensor, and the light is reflected or transmitted according to the reflectance or the transmittance of the original. In an image reading apparatus that obtains image information by detecting an electric signal as an image signal, a light beam splitting unit is provided that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary. A line for detecting the left and right images of the document surface in the main scanning direction on the respective optical paths divided into two parts from the vicinity of the central portion in the main scanning direction by the means, and reading the odd-numbered and even-numbered read pixels by separate transfer sections. A sensor is provided, and A / D conversion means for converting the output of each of the odd-numbered pixel and the even-numbered pixel of these line sensors into a digital signal is provided,
When the output corresponding to the same reading point on the document surface appears in the output of both the line sensors in the subsequent stage of these A / D conversion means, the addition result of the output of each subsequent stage is set as the read output and only one of the line sensors is used. When an output appears at the output, a calculation selecting means for making the output of the subsequent stage a read output is provided, and a shading correcting means for performing a shading correction by the calculation of the read output and the reference level, and an odd pixel Pixel output switching means for converting the output and the even pixel output into a column of sequential signals corresponding to the read space is provided.

According to the seventh aspect of the invention, the light transmitted or reflected by scanning the surface of the original is condensed by the imaging lens and is incident on the line sensor, and the light is reflected or transmitted according to the reflectance or the transmittance of the original. In an image reading apparatus that obtains image information by detecting an electric signal as an image signal, a light beam splitting unit is provided that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary. A line sensor having read pixel rows of a plurality of colors for detecting left and right images in the main scanning direction of the document surface is disposed on each of the optical paths divided into two parts from the vicinity of the central portion in the main scanning direction by the means. A / D conversion means for converting the output of the sensor into a digital signal is provided, and a line for correcting the read line shift in the sub-scanning direction of the read pixel rows of a plurality of colors is provided after the A / D conversion means. When the output corresponding to the same reading point on the document surface appears in the outputs of both line sensors, the addition result of the output of each subsequent stage is used as the read output and the output appears only in one of the line sensors. In this case, there is provided a calculation selecting means for making the output of the subsequent stage the read output, and a shading correcting means arranged at the latter stage of these calculation selecting means for performing shading correction by the calculation of the read output and the reference level.

[0012]

According to the first aspect of the invention, the output of each line sensor is subjected to shading correction by calculation with the reference level for each pixel to perform shading of the optical system, and the influence of the sensitivity difference for each pixel of the line sensor. After the correction, the weighting calculation is performed on the overlapping reading area according to the read pixel position, so that it is possible to smoothly connect the output difference caused by the difference in the document position expected by the line sensor.

According to a second aspect of the present invention, each line sensor reads out the odd-numbered pixel and the even-numbered pixel separately, and performs shading correction by calculation with the reference level for each pixel, thereby performing an optical system. After correcting the influence of the difference in sensitivity of each pixel of the line sensor by converting the output of the odd pixel and the output of the even pixel into a single-row sequential signal corresponding to the read space, the overlapping read area By performing weighting calculation according to the read pixel position, it is possible to smoothly connect the output differences due to the difference in the document positions expected by the line sensor.

According to the third aspect of the invention, the output of each line sensor is subjected to shading correction by calculation with the reference level for each color pixel to perform shading of the optical system, and the difference in sensitivity between the read pixel rows of the line sensor. After correcting the influence of the above, the reading position of each pixel row in the sub-scanning direction is aligned by the line delay means, and a new set of signals is generated from the aligned signals, and then the overlapping reading region is weighted according to the reading pixel position. By performing the calculation, the influence of the output difference between the line sensors due to the difference in spectral sensitivity is made extremely small, and the remaining characteristic difference and the output difference caused by the difference in the document position expected by these line sensors are connected smoothly. It becomes possible.

According to another aspect of the present invention, the output of each line sensor is added to the same reading position to obtain a line of line output corresponding to the amount of light before the light beam division, and at the same time, the original position estimated by the sensor is estimated. After smoothing the output difference caused by the difference between the pixels, the shading correction is performed by the calculation with the reference level for each pixel to perform the shading of the optical system to correct the influence of the sensitivity difference for each pixel of the line sensor. Is possible.

According to a fifth aspect of the present invention, each line sensor reads out the odd-numbered pixel and the even-numbered pixel separately, and the output of the odd-numbered pixel and the output of the even-numbered pixel correspond to the read space. Then, the signals from each line sensor are converted into one line of sequential signals, and the output of each line sensor is added together at the same reading position to obtain one line of line output corresponding to the amount of light before beam splitting. After smoothly connecting the output difference caused by the above, by performing the shading correction by the calculation with the reference level for each pixel, it is possible to correct the shading correction of the optical system and the influence of the sensitivity difference of each line sensor. .

According to the sixth aspect of the invention, each line sensor separately reads the odd-numbered pixel and the even-numbered pixel, and outputs the output of the odd-numbered pixel and the output of the even-numbered pixel at the same reading position. After the process of adding about, the line output of one row of pixels that are scattered one by one corresponding to the light amount before the light flux division is obtained, and at the same time, the output difference caused by the difference in the document position that the line sensor expects is connected smoothly, Shading correction is performed by calculation with the reference level for each pixel to correct the effect of the shading of the optical system and the sensitivity difference of each pixel of the line sensor, and the output of the odd pixel and the output of the even pixel correspond to the read space. It becomes possible to convert into a single column of sequential signals.

According to the seventh aspect of the present invention, the reading position of the output of each pixel column of each line sensor in the sub-scanning direction is aligned by the line delay means, and the processing of adding the line sensor outputs for the same reading position is performed. Then, after obtaining a line output corresponding to the amount of light before splitting the light flux and at the same time smoothly connecting the output difference between the sensors due to the difference in the document position expected by the line sensor and the difference in spectral sensitivity to the pixel, It is possible to perform shading correction by calculation with the reference level for each pixel, and to correct the influence of the shading of the optical system and the sensitivity difference for each pixel of the line sensor.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the invention according to claim 1 is shown in FIGS.
It will be described based on. In this embodiment, the light transmitted or reflected by scanning the original surface is condensed by an imaging lens and is incident on a line sensor, and an electric signal corresponding to the reflectance or the transmittance of the original is used as an image signal. The present invention relates to an image reading device that obtains image information by detecting it.
FIG. 1 is a block diagram showing the internal structure of the image reading apparatus. On the optical path of the focused light 2 from the imaging lens 1, a light beam splitting mirror 3 is disposed as a light beam splitting unit that splits the focused light 2 into two with a substantially central portion in the main scanning direction (not shown) as a boundary. ing. Line sensors 4 and 5 for detecting left and right images in the main scanning direction of a document surface (not shown) are arranged on the respective optical paths divided by the light beam separating mirror 3 from the vicinity of the central portion in the main scanning direction. Has been done. After the line sensors 4 and 5, sample and hold circuits (SH circuits) 6 and 7, analog signal processing circuits 8 and 9 including an amplifier and a level shift circuit, and A / D conversion means for converting into digital signals are provided. A / D conversion circuits 10 and 11, dark output correction circuits 12 and 13, and shading correction circuits 14 and 15 as shading correction means for performing shading correction by calculation of a digitally converted read output and a reference level are sequentially connected. There is.
The dark output correction circuits 12 and 13 are correction memories 16 and 17, respectively.
The shading correction circuits 14 and 15 include correction memories 18 and 19. An operation selection circuit 20 as an operation selection unit is connected to a stage subsequent to the shading correction circuits 14 and 15. The operation selection circuit 20 is a FIFO (First In First Out).
It comprises a memory 21, a LIFO (Last In First Out) memory 22, multiplication circuits 23 and 24, an addition circuit 25, coefficient memories 26 and 27, and a switching circuit 28. The switching circuit 28 is connected to the output terminal 29. In this case, the calculation selection circuit 20 outputs the outputs of both line sensors 4 corresponding to the same reading point on the document surface.
When it appears in the output of No. 5, the weighting calculation result of the output of each subsequent stage is set as the read output, and when the output appears in only one of the line sensors, the output of the subsequent stage is set as the read output.

The operation of this apparatus having the above structure will be described. First, the light from the document surface is formed by the imaging lens 1
To be focused light 2, which is then incident on the light beam separation mirror 3 to be split in the main scanning direction. At this time, the light from the vicinity of the center of the document surface in the main scanning direction is split and incident on both line sensors 4 and 5. On the other hand, the light from both ends of the document surface in the main scanning direction respectively enter the line sensors 4 and 5 separately. FIGS. 2A and 2B show a light amount distribution showing how the light amounts are separated and the respective light amounts to the line sensors 4 and 5 become. In FIG. 2A, the area A on the horizontal axis indicates the light receiving range (installation range) of one line sensor 5, the area B indicates the light receiving range (installation range) of the other line sensor 4, and the vertical axis indicates the line. Sensor 4,
5 shows the amount of light detected. With this light quantity distribution 30,
It can be seen that the light from the center in the main scanning direction is exactly divided in half by the light beam separation mirror 3. Also, FIG.
(B) shows the light amount distribution 31 when the distance between the imaging lens 1 and the light beam separation mirror 3 is increased. In this case, in the overlapping range C of the overlapping reading area, the light quantity becomes zero, but by adding the light quantities of both, the original flat light quantity distribution over one line can be reproduced.

Then, the respective outputs of the line sensors 4 and 5 are guided to the SH circuits 6 and 7, the transfer clock component of the line sensors 4 and 5 is removed, and the analog signal processing circuit 8 and
At 9, the amplitude and DC level suitable for the A / D conversion circuits 10 and 11 are converted, and at the A / D conversion circuits 10 and 11, they are converted into digital signals. In the dark output correction circuits 12 and 13, the dark output levels stored in advance for each pixel are corrected in the correction memories 16 and 1.
7 is read out, and the digital signal converted thereby is corrected for each pixel. Next, the shading correction circuit 14,
In 15, by calculating the reference data stored in advance from the correction memories 18 and 19 for each pixel,
The brightness distribution of the illumination system and the imaging system, the sensitivity variations among the pixels of the line sensors 4 and 5 are corrected. At this time, the influence of the light amount distribution (shading) caused by the light beam splitting mirror 3 is also corrected, and outputs can be obtained from both the two line sensors 4 and 5 in the overlapping reading area. By performing each process up to this shading correction, the first factor (difference in the amount of light reaching each line sensor) and the third factor (difference in the characteristics of the line sensor itself) as described in the above-mentioned conventional technique are performed. Of the sensitivity variation among each pixel)
Although the problem due to the above can be solved, the problem due to the second factor (difference in document position expected by the line sensor) still remains. In order to remove this second factor, the shading-corrected data is sent to the operation selection circuit 20.

Therefore, the operation of the operation selection circuit 20 will be described. Here, processing for smoothly connecting the duplicate reading areas is performed. FIG. 3 shows the LIFO 22, F
The state of the write timing 32 and the read timing 33 of the IFO 21 is shown. Basically, LIFO
The data order is reversed by reading 22 and the signal of the overlapping reading area is brought to the end portion of the original as a signal from the end portion to the central portion of the document. The reading of the FIFO memory 21 is started in synchronization with the reading timing of the overlapping reading area so that the signals of the same reading position of the document are generated at the same time. The multiplying circuits 23 and 24, the coefficient memories 26 and 27, and the adding circuit 25 are used to perform weighted addition according to the reading position in the overlapping reading area. Figure 4 (a)
4B shows an example of values stored in the coefficient memories 26 and 27, and FIG. 4A shows the coefficient memory 26.
(B) shows the coefficient memory 27. In this case, FIG.
2B, the waveforms 34a and 35a of solid lines are shown when the light amount distribution of FIG. 2A is obtained, and the waveforms 34b and 35b of broken lines are shown when the light amount distribution of FIG. 2B is obtained. Then, finally, the signal of the portion other than the overlapped reading area and the signal of the portion of the overlapped reading area are selected by switching by the switching circuit 28, so that the output terminal 29 outputs the one-line format which is smoothly connected. Read output can be obtained. The reading directions of the line sensors 4 and 5 in this embodiment are such that the center of the document surface is read first (that is, in the opposite directions), but the reading method is not limited to this. Absent.

As described above, the outputs of the line sensors 4 and 5 are shading-corrected by the calculation with the reference level of each pixel to shade the optical system, and the influence of the sensitivity difference of each pixel of the line sensors 4 and 5 is influenced. After correcting the
By performing a weighting calculation according to the read pixel position with respect to the overlapping read area, it is possible to smoothly connect the output difference caused by the difference between the document positions expected by the line sensors 4 and 5, whereby the one-line format is obtained. And
Highly accurate, high gradation, and high-density reading output can be realized comprehensively.

Next, an embodiment of the invention described in claim 2 will be described with reference to FIG. The description of the same parts as those of the invention according to claim 1 will be omitted, and the same reference numerals will be used for the same parts (provided that numbers with lowercase letters a and b are the same as the numbers. For example, 6
The names of a and 6b are SH circuits which are the names of 6).

Here, the line sensors 4 and 5 of the type in which the odd (Odd) th pixel and the even (Even) th pixel are read out by different transfer units are used. Also, the shading correction circuit 1
4a and 14b and shading correction circuits 15a and 15
In the subsequent stage of b, multiplexers 36 and 37 are connected as pixel output switching means for converting the odd pixel output and the even pixel output into one column of sequential signals corresponding to the read space.

As described above, in this embodiment, the line sensors 4 and 5 separately read out the odd-numbered pixels and the even-numbered pixels, and the SH circuits 6a, 6b, 7a and 7b output the shading correction circuits 14a, 14b and 15a, respectively. After the shading of the optical system and the correction of sensitivity variations among the pixels of the line sensors 4 and 5 are performed up to 15b, the multiplexer 3
The signal is converted into a single column of sequential signals corresponding to the spaces read by the reference numerals 6 and 37, and the operation selection circuit 20 is used for the overlapping read area.
In, the weighting calculation is performed according to the read pixel position. As a result, it is possible to smoothly connect the output difference caused by the difference between the document positions expected by the line sensors 4 and 5, so that the one-line format is realized and the high accuracy is achieved.
It is possible to obtain high gradation and high density read output.

The multiplexers 36 and 37 can be provided in front of the shading correction circuits 14a, 14b, 15a and 15b. In this case, the shading correction circuits 14a and 14b form one block, and the shading correction circuits 15a and 15b form one block. Further, the multiplexers 36 and 37 can be provided after the operation selection circuit 20. In this case, the calculation selection circuit 20
Has two systems, one block to which inputs are supplied from the shading correction circuits 14a and 14b and one block to which inputs are supplied from the shading correction circuits 15a and 15b. The configuration is such that selection switching is performed.

Next, an embodiment of the invention described in claim 3 will be described with reference to FIG. The description of the same parts as those in the first and second aspects of the present invention will be omitted, and the same parts will be denoted by the same reference numerals (provided that the numbers with R, G, B are the same as the numbers. For example, 6
The names of R, 6G, and 6B are SH circuits which are the names of 6).

Here, R is used as the line sensors 4 and 5.
Color reading is performed using color line sensors of three colors (red), G (green), and B (blue). Further, FIFO memories 38R, 38G, 39B and 39G as line delay means for correcting the read line shift of the read pixel columns of the respective colors in the sub-scanning direction are connected to the subsequent stages of the shading correction circuits 14R, 14G, 15B and 15G. ing. Matrix circuits 40 and 41 are connected to the subsequent stages of the FIFO memories 38R, 38G, 39B and 39G as matrix calculation means for generating a new set of a plurality of signals from the output of the read pixel column of each color. The operation selection circuits 20R, 20G, and
20B is connected.

Thus, after shading correction is performed, the read line shift in the sub-scanning direction is corrected for the signal at the same reading position in the sub-scanning direction using the FIFO memories 38R, 38G, 39B and 39G. Align. Then, the matrix circuits 40 and 41 specify the spectroscopic reading characteristics of the line sensors 4 and 5 to a certain value (standard).
To the spectral reading characteristic of the calculation selection circuit 20.
By performing the overlapping processing of the overlapping reading areas by R, 20G, and 20B, it is possible to realize a good and smooth connection regarding the difference in spectral sensitivity.

As described above, in color reading,
The effect of the output difference between the line sensors 4 and 5 due to the difference in spectral sensitivity should be minimized, and the remaining characteristic difference and the output difference resulting from the difference in the document positions expected by these line sensors 4 and 5 should be connected smoothly. Thus, it is possible to obtain a read output having high accuracy, high gradation, high density, and uniform color characteristics.

The shading correction circuits 14R, 1
4G, 14B, 15R, 15G, 15B, FIFO memories 38R, 38G, 39B, 39G, matrix circuits 40, 41, and operation selection circuits 20R, 20G, 20B.
The processing order may be changed as follows. That is, the shading correction circuits 14R, 14G, 14B,
The operation selection circuits 20R, 15R, 15G, 15B are provided in the subsequent stages.
Conditions for providing 20G, 20B and shading correction circuits 14R, 14G, 14B, 15R, 15G, 1
5B, the conditions for providing the matrix circuits 40 and 41 in the subsequent stage, and the FIFO memories 38R, 38G, 39B and 39.
The processing order may be changed as long as the condition for providing the matrix circuits 40 and 41 in the latter stage of G is satisfied.

Next, an embodiment of the invention described in claim 4 will be described with reference to FIG. The description of the same parts as those in the first to third aspects of the present invention is omitted, and the same parts are designated by the same reference numerals.

Here, the line sensors 4 and 5 to the dark output correction circuits 12 and 13 have the same circuit configuration as that of FIG. 1 described above, but the configuration of the subsequent stage is changed. That is, when the output corresponding to the same reading point on the document surface appears in the output of both line sensors 4 and 5 after the dark output correction circuits 12 and 13, the addition result of the output of each subsequent stage is set as the read output. When an output appears only in the line sensor (4 or 5), an operation selection circuit 42 as an operation selection means for connecting the output of the subsequent stage to the read output is connected, and the read output and the reference level are provided in the subsequent stage of the operation selection circuit 42. The shading correction circuit 43 as a shading correction means for performing shading correction by the calculation of is connected. In this case, the operation selection circuit 42 uses the FIFO memory 2
1, a LIFO memory 22, an adder circuit 25, and a switching circuit 28. A correction memory 44 is connected to the shading correction circuit 43.

The difference between this configuration and the configuration of FIG. 1 is that the shading correction is performed after the joining process in the overlap reading area. In such a circuit configuration, in particular, the light amount distribution to the line sensors 4 and 5 is shown in FIG.
It becomes effective when it becomes as shown in (b). That is, first, in the connection processing in the operation selection circuit 42,
The write timing and the read timing of the FIFO memory 21 and the LIFO memory 22 are the same as those in FIG. In the adder circuit 25, the FIFO memory 21 and
By adding both outputs from the LIFO memory 22 to a level equivalent to the amount of light before the light beam division, the effect of shading due to the division can be restored to the original form. In terms of the smoothness of the connection, the light amount distributions for both sensors in the overlapping portion directly contribute to the reading position, which is a good connection. After that, the shading correction circuit 43 performs shading correction on the output from the calculation selection circuit 42 to correct the brightness unevenness of the illumination system, the lens system, and the like and the sensitivity unevenness of each pixel of the line sensors 4 and 5. It can be performed.

As described above, the calculation selection circuit 42 performs the addition processing for the same reading position to smoothly connect the output difference caused by the difference in the document positions expected by the line sensors 4 and 5, and then the shading correction circuit 43.
By performing shading correction of the optical system and correcting the influence of the sensitivity difference for each pixel of the line sensor, it is possible to obtain a one-line format and high-precision, high-gradation, high-density read output. .

Next, an embodiment of the invention described in claim 5 will be described with reference to FIG. The description of the same parts as those in the first to fourth aspects of the present invention is omitted, and the same parts are designated by the same reference numerals.

Here, the line sensors 4 and 5 are of a type in which the odd and even pixels of the pixels are read by different transfer units, and the SH circuits 6a, 6b, 7a and 7b to the dark output correction circuit 12a, 12b, 13a and 13b are shown in FIG.
Of the dark output correction circuit 12a,
12b, 13a and 13b are followed by a multiplexer 36,
37 is connected, and further, the multiplexers 36 and 37 are connected.
The operation selection circuit 42 and the shading correction circuit 43 used in the circuit of FIG. 7 are connected in the subsequent stage.

The difference between this structure and the structure shown in FIG. 7 is that the multiplexers 36 and 37 convert the signals into one column of sequential signals, and then perform the connecting process of the overlapped read regions. That is,
The line sensors 4 and 5 separately read the odd-numbered pixels and the even-numbered pixels, and the multiplexers 36 and 37 are used for the output of the odd-numbered pixels and the output of the even-numbered pixels corresponding to the read space. After being converted into one column of sequential signals, each of the line sensors 4, 4
The output of No. 5 is added at the same reading position, a line output corresponding to the light amount before the light beam division is obtained, and at the same time, a process for smoothly connecting the output difference caused by the difference in the document position expected by the line sensor is performed. . After that, the shading correction circuit 43 is used to perform shading correction by calculation with the reference level for each pixel, and the shading correction of the optical system and the influence of the sensitivity difference of each line sensor are corrected to obtain the one-line format. Moreover, high-accuracy, high-gradation, high-density read output can be obtained.

Next, an embodiment of the invention described in claim 6 will be described with reference to FIG. The description of the same parts as those in the first to fifth aspects of the present invention is omitted, and the same parts are denoted by the same reference numerals.

Here, the line sensors 4 and 5 are of the type in which the odd and even pixels of pixels are read out by separate transfer units, and the SH circuits 6a, 6b, 7a and 7b to the dark output correction circuit 12a, 12b, 13a, and 13b are shown in FIG.
Of the dark output correction circuit 12a,
Operation selection circuits 42a and 42b for separately processing odd-numbered pixels and even-numbered pixels are connected to the subsequent stages of 12b, 13a, and 13b, and shading correction circuits 43a and 43b are connected to the subsequent stages of the operation selection circuits 42a and 42b. A multiplexer 45 is connected as pixel output switching means for converting into a column of sequential signals corresponding to the space read after the shading correction circuits 43a and 43b.

The difference between this configuration and the configuration of FIG. 8 is that the overlapping reading areas are separately calculated, selected, and shading-corrected by using the odd-numbered pixels and the even-numbered pixels, and then the space finally read is corresponded. It is adapted to be converted into a one-row sequential signal. That is, the line sensors 4 and 5 separately read the odd-numbered pixels and the even-numbered pixels, and the operation selection circuits 42a and 42b are used to output the odd-numbered pixels and the even-numbered pixels at the same reading position. Is added to obtain the line output of one row of pixels that are scattered one by one corresponding to the amount of light before the light beam division, and at the same time smooth the output difference caused by the difference in the document positions that the line sensors 4 and 5 can see. Perform processing to connect to. Thereafter, shading correction circuits 43a and 43b are used to perform shading correction by calculation with a reference level for each pixel, and shading of the optical system and line sensors 4 and 5 are performed.
Of the sensitivity difference of each pixel of the
By converting the output of the odd-numbered pixel and the output of the even-numbered pixel into a one-row sequential signal corresponding to the read space, a one-line format is achieved, and high-precision, high-gradation, high-density reading is performed. You can get the output.

The shading correction circuits 43a, 4
3b and the multiplexer 45 may be connected so that the processing order is reversed. In this case, the shading correction circuits 43a and 43b are represented by one block.

Next, an embodiment of the invention described in claim 7 will be described with reference to FIG. The description of the same parts as those in the first to sixth aspects of the present invention will be omitted, and the same reference numerals will be used for the same parts.

Here, as the line sensors 4 and 5,
Color reading is performed using the three color line sensors R, G, B, and the SH circuits 6R, 6G, 6B, 7R, 7G,
7B to dark output correction circuits 12R, 12G, 12B, 13
The circuit configuration up to R, 13G, and 13B is the same as the circuit configuration of FIG. 8, and the dark output correction circuits 12R, 12G, 12B, and 13 are used.
A FIFO memory 38R, 3 is provided at the rear stage of the R, 13G, 13B.
8G, 39B, 39G are connected, and this FIFO memory 3
8R, 38G, 39B and 39G are provided at the subsequent stage of the operation selection circuit 4
2R, 42G, 42B are connected to each other, and the operation selection circuit 42
Shading correction circuit 4 is provided in the subsequent stage of R, 42G, 42B.
3R, 43G, 43B are connected.

The difference between this configuration and the configuration of FIG. 6 is that the reading line shift of the reading pixels of each color in the sub-scanning direction is corrected, and the shading correction is performed after the joining process in the overlapping reading area. Is. In this case, since the reading directions of the line sensors 4 and 5 are opposite to each other, the reading position shift in the sub-scanning direction is opposite in color in the line sensors 4 and 5. Therefore, the FIFO memories 38R, 38
G, 39B, and 39G are used to adjust the reading position shift in the sub-scanning direction, and then the result is guided to the calculation selection circuits 42R, 42G, and 42B.

In this way, the reading positions of the outputs of the respective pixel columns of the line sensors 4 and 5 are aligned in the sub-scanning direction, and the outputs of the line sensors 4 and 5 are added to each other at the same reading position. A line of line outputs corresponding to the light amount of 1 is obtained, and at the same time, a process for smoothly connecting the output difference between the sensors due to the difference in the document positions expected by the line sensors 4 and 5 and the difference in spectral sensitivity is performed. After that, the shading correction is performed by the calculation with the reference level for each pixel, and the shading of the optical system and the line sensor 4,
By correcting the influence of the sensitivity difference for each pixel of 5, it is possible to obtain a read output with high accuracy, high gradation, high density, and smooth color connection.

The FIFO memories 38R, 38G, 3
9B and 39G are operation selection circuits 42R, 42G and 42B.
Or the shading correction circuits 43R, 43
It may be in the latter stage of G and 43B.

[0049]

According to the first aspect of the present invention, the light transmitted or reflected by scanning the surface of the original is condensed by the imaging lens and is incident on the line sensor, and the reflectance or the transmittance of the original is adjusted. In an image reading apparatus that obtains image information by detecting a corresponding electric signal as an image signal, a light beam splitting unit is provided that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary. In principle, the accuracy difference caused by the lens, such as a magnification error when two lenses are used, is removed, and then the original is placed on each optical path divided into two from the vicinity of the central portion in the main scanning direction by the light beam splitting means. Line sensors for detecting left and right images of the surface in the main scanning direction are provided, and A / D conversion means for converting the outputs of these line sensors into digital signals are provided.
Shading correction means for performing shading correction by calculating the read output digitally converted by the conversion means and the reference level is provided, and the output corresponding to the same reading point on the document surface is provided at both stages of the line sensor after the shading correction means. In the case where the output of each of the latter stage outputs the weighted calculation result as a read output and the output of only one of the line sensors appears, the output of the latter stage becomes the read output, so that the calculation selecting means is provided. Shading correction is performed on the output of the line sensor by calculation with the reference level for each pixel to perform the shading of the optical system to correct the influence of the sensitivity difference of each pixel of the line sensor, and then the overlapping reading area is set to the reading pixel position. Output due to difference in document position expected by line sensor It can be coupled also smooth, thereby, becomes 1 line form, and, overall high-precision, high gradation, in which it is possible to realize a high-density read.

According to the second aspect of the present invention, the light transmitted or reflected by scanning the surface of the original is condensed by the imaging lens and is incident on the line sensor, and the light is reflected or transmitted according to the reflectance or the transmittance of the original. In an image reading apparatus that obtains image information by detecting an electric signal as an image signal, a light beam splitting unit that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary is provided. In principle, the accuracy difference caused by the lens, such as the magnification error, is removed, and the light beam splitting means splits the document surface into two on the respective optical paths from the vicinity of the central portion in the main scanning direction. Line sensors that detect the left and right images in the main scanning direction and read out the odd and even numbers of the read pixels by separate transfer units are provided, and the outputs of these line sensors are converted into digital signals. That A / D provided converting means, these A
A shading correction unit that performs shading correction by a calculation of a digitally converted read output and a reference level, and an odd-pixel output and an even-pixel output corresponding to the read space are arranged in a row after the / D conversion unit. Pixel output switching means for converting into a signal and weighting of the output of each subsequent stage when an output corresponding to the same reading point of the document surface placed at the subsequent stage of the shading correction means appears in the outputs of both line sensors. Since the calculation result is used as the read output and the output of the subsequent stage is used as the read output when the output appears only on the one line sensor, each line sensor has an odd-numbered pixel and an even-numbered pixel. Pixels are read separately, and shading correction is performed by calculation with the reference level for each pixel to determine the shade of the optical system. After correcting the influence of the sensitivity difference of each pixel of the line sensor and converting the output of the odd pixel and the output of the even pixel into one column of sequential signals corresponding to the read space, It is possible to smoothly connect the output difference caused by the difference in the document position, which is expected by the line sensor, by performing the weighting calculation according to the read pixel position, whereby the one-line format is achieved, and the overall accuracy is high. High gradation and high density reading can be realized.

According to a third aspect of the present invention, the light transmitted or reflected by scanning the surface of the original is condensed by the imaging lens and is incident on the line sensor, and the light is reflected or transmitted in accordance with the reflectance of the original. In an image reading apparatus that obtains image information by detecting an electric signal as an image signal, a light beam splitting unit that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary is provided. In principle, the accuracy difference caused by the lens, such as the magnification error, is removed, and the light beam splitting means splits the document surface into two on the respective optical paths from the vicinity of the central portion in the main scanning direction. A line sensor having read pixel rows of a plurality of colors for detecting left and right images in the main scanning direction is provided, and A / D conversion means for converting outputs of the line sensors into digital signals is provided. A shading correction unit that performs shading correction by calculating a digitally converted read output and a reference level, and a line delay that corrects a read line shift in a sub-scanning direction of read pixel columns of a plurality of colors are provided in a subsequent stage of the A / D conversion unit. Means, a matrix operation means for generating a new set of a plurality of signals from the outputs of the read pixel rows of a plurality of colors, which is placed after the shading correction means and the line delay means, and a matrix operation means provided after the shading correction means. When the output corresponding to the same reading point on the document surface appears in the outputs of both line sensors, the weighted calculation result of the output of each subsequent stage is used as the read output, and the output appears only in one of the line sensors. Since the output of the subsequent stage is used as the read output, the output of each line sensor is set to the pixel of each color. The shading of the optical system is corrected by calculating the shading correction by the calculation with the reference level, and the influence of the sensitivity difference for each reading pixel row of the line sensor is corrected, and the reading position of each pixel row in the sub-scanning direction is aligned by the line delay means. After generating a new set of signals from the aligned signals, weighting calculation is performed according to the read pixel position in the overlapping read area, and the influence of the output difference between the line sensors due to the difference in spectral sensitivity is minimized and left. It is possible to smoothly connect the difference in characteristics and the difference in output caused by the difference in the document positions expected by these line sensors, which enables high-accuracy, high-gradation, high-density, and color-matched reading. The output can be obtained.

According to a fourth aspect of the present invention, the light transmitted or reflected by scanning the surface of the original is condensed by the imaging lens and is incident on the line sensor, and the light is reflected or transmitted according to the reflectance or the transmittance of the original. In an image reading apparatus that obtains image information by detecting an electric signal as an image signal, a light beam splitting unit that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary is provided. In principle, the accuracy difference caused by the lens, such as the magnification error, is removed, and the light beam splitting means splits the document surface into two on the respective optical paths from the vicinity of the central portion in the main scanning direction. Line sensors for detecting the left and right images in the main scanning direction are provided, and A / D conversion means for converting the output of these line sensors into digital signals are provided. When the output corresponding to the same reading point on the document surface appears in the outputs of both line sensors, the addition result of the output of each subsequent stage is taken as the read output and the output of the subsequent stage is output when the output appears only in one of the line sensors. Is provided as a read output, and shading correction means for performing shading correction by the calculation of the read output and the reference level is provided at the subsequent stage of the calculation selection means, so that the output of each line sensor is the same read position. Then, a line output corresponding to the amount of light before the light beam division is obtained, and at the same time, the output difference caused by the difference in the document position that the sensor expects is smoothly connected, and then the calculation with the reference level for each pixel is performed. The shading correction can be performed to correct the effect of the optical system shading and the sensitivity difference of each pixel of the line sensor. The became 1 line form, and high-precision, high gradation, in which it is possible to realize a high-density read.

According to a fifth aspect of the present invention, the light transmitted or reflected by scanning the surface of the original is condensed by the imaging lens and is incident on the line sensor, and the light is reflected or transmitted depending on the reflectance or the transmittance of the original. In an image reading apparatus that obtains image information by detecting an electric signal as an image signal, a light beam splitting unit that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary is provided. In principle, the accuracy difference caused by the lens, such as the magnification error, is removed, and the light beam splitting means splits the document surface into two on the respective optical paths from the vicinity of the central portion in the main scanning direction. Line sensors that detect the left and right images in the main scanning direction and read out the odd and even numbers of the read pixels by separate transfer units are provided, and the outputs of these line sensors are converted into digital signals. That A / D provided converting means, these A
Pixel output switching means for converting the odd pixel output and the even pixel output into a single-row sequential signal corresponding to the read space is provided in the subsequent stage of the / D conversion means, and the original surface of the document is provided in the subsequent stage of the pixel output switching means. When the output corresponding to the same reading point appears in the output of both line sensors, the addition result of the output of each subsequent stage is set as the read output, and when the output appears only in one of the line sensors, the output of the subsequent stage is read out. Since the shading correction means for performing shading correction by the calculation of the read output and the reference level is provided at the subsequent stage of this calculation selection means, each line sensor has an odd-numbered pixel and an even-numbered pixel. Pixels are read separately, and the output of these odd pixels and the output of even pixels are converted into a column of sequential signals corresponding to the read space and each line is read. After the sensor output is added at the same reading position, a line output corresponding to the amount of light before beam splitting is obtained and at the same time the output difference caused by the difference in the document position that the line sensor expects is connected smoothly, and then the pixel Shading correction can be performed by calculation with the reference level for each, and it is possible to correct the effect of the shading correction of the optical system and the sensitivity difference of each line sensor.
This is a one-line format and can realize high-accuracy, high-gradation, high-density reading.

According to a sixth aspect of the present invention, the light transmitted or reflected by scanning the surface of the document is condensed by the imaging lens and is incident on the line sensor, and the light is reflected or transmitted according to the reflectivity or the transmittance of the document. In an image reading apparatus that obtains image information by detecting an electric signal as an image signal, a light beam splitting unit that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary is provided. In principle, the accuracy difference caused by the lens, such as the magnification error, is removed, and the light beam splitting means splits the document surface into two on the respective optical paths from the vicinity of the central portion in the main scanning direction. A line sensor is provided which detects each of the left and right images in the main scanning direction and reads out the odd-numbered and even-numbered read pixels by separate transfer units. A / D conversion means for converting each output of the pixel into a digital signal is provided, and an output corresponding to the same reading point of the document surface appears in the output of both line sensors after the A / D conversion means. Is provided with a calculation selection means for making the addition result of the output of each subsequent stage as a read output and the output of the subsequent stage as a read output when the output appears only at one of the line sensors, and the reading selection is provided at the stage subsequent to this calculation selection means. Since the shading correction means for performing the shading correction by the calculation of the output and the reference level and the pixel output switching means for converting the odd pixel output and the even pixel output into the one-row sequential signal corresponding to the read space are provided. , Each line sensor reads the odd-numbered pixel and the even-numbered pixel separately, and the output of the odd-numbered pixel and the output of the even-numbered pixel are the same. The output of the taking position is added together to obtain the line output of one row of pixels, which corresponds to the amount of light before the light flux division, one pixel at a time, and at the same time the output difference caused by the difference in the document position seen by the line sensor is smoothed. After connecting, shading correction was performed by calculation with the reference level for each pixel to correct the effect of the shading of the optical system and the sensitivity difference of each pixel of the line sensor, and the output of the odd pixel and the output of the even pixel were read. It is possible to realize high-accuracy, high-gradation, high-density reading in the one-line format by converting into a column of sequential signals corresponding to the space.

According to a seventh aspect of the present invention, the light transmitted or reflected by scanning the surface of the original is condensed by the imaging lens and is incident on the line sensor, and the light is reflected or transmitted according to the reflectance or the transmittance of the original. In an image reading apparatus that obtains image information by detecting an electric signal as an image signal, a light beam splitting unit that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary is provided. In principle, the accuracy difference caused by the lens, such as the magnification error, is removed, and the light beam splitting means splits the document surface into two on the respective optical paths from the vicinity of the central portion in the main scanning direction. A line sensor having read pixel rows of a plurality of colors for detecting left and right images in the main scanning direction is provided, and A / D conversion means for converting outputs of the line sensors into digital signals is provided. After the A / D conversion means, a line delay means for correcting read line shifts in the sub-scanning direction of read pixel rows of a plurality of colors and an output corresponding to the same read point on the document surface are output from both line sensors. And the output of the latter stage is read out, and the output of the latter stage is read out when the output appears only on one of the line sensors. Since the shading correction means for performing the shading correction by the calculation of the read output and the reference level thus obtained is provided, the reading position in the sub-scanning direction of the output of each pixel column of each line sensor is aligned by the line delay means.
The line sensor outputs are added together for the same reading position to obtain a line output corresponding to the amount of light before beam splitting, and at the same time, the output difference and the spectral sensitivity difference due to the difference in the document position that the line sensor expects. After smoothly connecting the output difference between the sensors, the shading correction is performed by the calculation with the reference level for each pixel to correct the effect of the shading of the optical system and the sensitivity difference of each pixel of the line sensor, thereby the 1-line format It is possible to realize high-precision, high-gradation, and high-density reading.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an image reading apparatus according to an embodiment of the present invention.

FIG. 2 is a characteristic diagram showing a light amount distribution detected by two line sensors.

FIG. 3 is a schematic diagram showing write timing and read timing with respect to a FIFO memory and a LIFO memory.

FIG. 4 is a schematic diagram showing a storage distribution in a coefficient memory.

FIG. 5 is a block diagram showing a configuration of an image reading apparatus which is an embodiment of the invention described in claim 2.

FIG. 6 is a block diagram showing a configuration of an image reading apparatus which is an embodiment of the invention according to claim 3;

FIG. 7 is a block diagram showing a configuration of an image reading apparatus which is an embodiment of the invention described in claim 4.

FIG. 8 is a block diagram showing a configuration of an image reading apparatus according to an embodiment of the present invention.

FIG. 9 is a block diagram showing a configuration of an image reading apparatus which is an embodiment of the invention described in claim 6;

FIG. 10 is a block diagram showing the arrangement of an image reading apparatus according to an embodiment of the present invention.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Imaging lens 2 Focused light 3 Light flux dividing means 4,5 Line sensor 10, 10a, 10b, 10R, 10G, 10B A
/ D conversion means 11, 11a, 11b, 11R, 11G, 11B A
/ D conversion means 14, 14a, 14b, 14R, 14G, 14B Shading correction means 15, 15a, 15b, 15R, 15G, 15B Shading correction means 20 Calculation selection means 36, 37 Pixel output switching means 38R, 38G, 39B, 39G Line delay means 40, 41 Matrix calculation means 42, 42a, 42b Calculation selection means 43, 43a, 43b Shading correction means 45 Pixel output switching means

Claims (7)

[Claims] 1. A light transmitted or reflected by scanning a document surface is condensed by an imaging lens and is incident on a line sensor, and an electric signal corresponding to the reflectance or the transmittance of the document is used as an image signal. In the image reading apparatus that obtains image information by detecting, a light beam splitting unit that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary is provided. A / D conversion means for arranging line sensors for detecting left and right images of the document surface in the main scanning direction on respective optical paths divided from the vicinity of the central portion and converting outputs of these line sensors into digital signals. And shading correction means for performing shading correction by calculating the read output digitally converted by these A / D conversion means and the reference level. When the output corresponding to the same reading point on the document surface appears in the output of both the line sensors in the subsequent stage of the shading correction means, the weighted calculation result of the output in each subsequent stage is set as the read output and is output only to the one line sensor. The image reading apparatus is provided with an operation selecting unit that outputs the output of the subsequent stage as a read output when the error occurs. 2. The light transmitted or reflected by scanning the surface of the original is condensed by an imaging lens and is incident on a line sensor, and an electric signal corresponding to the reflectance or the transmittance of the original is used as an image signal. In the image reading apparatus that obtains image information by detecting, a light beam splitting unit that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary is provided. A line sensor is provided on each of the two optical paths divided from the vicinity of the central portion to detect the left and right images of the document surface in the main scanning direction and read out the odd-numbered and even-numbered read pixels by separate transfer units. A / D conversion means for converting the output of these line sensors into a digital signal is provided, and the digitally converted read output and the reference level are calculated after the A / D conversion means. Shading correction means for performing shading correction, pixel output switching means for converting the odd pixel output and the even pixel output into one column of sequential signals corresponding to the read space, and the shading correction means which is provided after the shading correction means. When the output corresponding to the same reading point on the document surface appears in the outputs of both line sensors, the weighting calculation result of the output of each subsequent stage is taken as the read output, and when the output appears only in one of the line sensors, An image reading apparatus comprising: an operation selecting unit that outputs an output as a read output. 3. The light transmitted or reflected by scanning the surface of the original is condensed by an imaging lens and is incident on a line sensor, and an electric signal corresponding to the reflectance or the transmittance of the original is used as an image signal. In the image reading apparatus that obtains image information by detecting, a light beam splitting unit that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary is provided. Line sensors having read pixel rows of a plurality of colors for detecting left and right images of the document surface in the main scanning direction are provided on respective optical paths divided from the vicinity of the central portion, and the outputs of these line sensors are digital signals. A shading correction is performed by providing an A / D conversion means for converting to a digital signal, and performing shading correction by calculating the digitally converted read output and the reference level after the A / D conversion means. From the output of the read pixel row of a plurality of colors to the stage subsequent to the shading corrector and the stage of the line delay means. Matrix calculation means for generating a new set of a plurality of signals and output of each of the latter stages when an output corresponding to the same reading point of the document surface appears in the output of both of the line sensors after the shading correction means. An image reading apparatus comprising: a weighting calculation result as a read output, and an operation selecting unit that makes an output of the subsequent stage a read output when the output appears only on one of the line sensors. 4. The light transmitted or reflected by scanning the surface of the original is condensed by an imaging lens and made incident on a line sensor, and an electric signal corresponding to the reflectance or the transmittance of the original is used as an image signal. In the image reading apparatus that obtains image information by detecting, a light beam splitting unit that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary is provided. A / D conversion means for arranging line sensors for detecting left and right images of the document surface in the main scanning direction on respective optical paths divided from the vicinity of the central portion and converting outputs of these line sensors into digital signals. Is provided, and when the output corresponding to the same reading point on the document surface appears in the output of both of the line sensors in the subsequent stage of these A / D conversion means, the addition result of the output of each subsequent stage Is provided as a read output, and when an output appears only on one of the line sensors, a calculation selecting means is provided for making the output of the subsequent stage a read output, and shading correction is performed by calculation of the read output and the reference level at a stage subsequent to the calculation selecting means. An image reading apparatus comprising a shading correction means for performing the following. 5. The light transmitted or reflected by scanning the surface of the original is condensed by an imaging lens and is incident on a line sensor, and an electric signal corresponding to the reflectance or the transmittance of the original is used as an image signal. In the image reading apparatus that obtains image information by detecting, a light beam splitting unit that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary is provided. A line sensor is provided on each of the two optical paths divided from the vicinity of the central portion to detect the left and right images of the document surface in the main scanning direction and read out the odd-numbered and even-numbered read pixels by separate transfer units. A / D conversion means for converting the output of these line sensors into a digital signal is provided, and the odd pixel output and the even pixel output correspond to the read space after the A / D conversion means. Pixel output switching means for converting into one column of sequential signals is provided, and when outputs corresponding to the same reading point of the document surface appear in the outputs of both line sensors, the pixel output switching means is provided in the rear stage of the pixel output switching means. Provided is a calculation selection means for making the addition result of the outputs a read output and the output of the subsequent stage as the read output when the output appears only on one of the line sensors, and the read output and the reference level are provided at the subsequent stage of the calculation selecting means. An image reading apparatus comprising shading correction means for performing shading correction by calculation. 6. The light transmitted or reflected by scanning the surface of the original is condensed by an imaging lens and made incident on a line sensor, and an electric signal corresponding to the reflectance or the transmittance of the original is used as an image signal. In the image reading apparatus that obtains image information by detecting, a light beam splitting unit that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary is provided. A line sensor is provided on each of the two optical paths divided from the vicinity of the central portion to detect the left and right images of the document surface in the main scanning direction and read out the odd-numbered and even-numbered read pixels by separate transfer units. A / D conversion means for converting the output of each of the odd-numbered pixel and the even-numbered pixel of these line sensors into a digital signal is provided, and the document surface of the document is provided at the subsequent stage of these A / D conversion means. When the output corresponding to the same reading point appears in the output of both line sensors, the addition result of the output of each subsequent stage is set as the read output, and when the output appears only in one of the line sensors, the output of the subsequent stage is read out. And a shading correction unit for performing shading correction by the calculation of the read output and the reference level, and an odd pixel output and an even pixel output corresponding to the read space. An image reading apparatus comprising: a pixel output switching unit that converts the signal into a single-row sequential signal. 7. The light transmitted or reflected by scanning the original surface is condensed by an imaging lens and is incident on a line sensor, and an electric signal corresponding to the reflectance or the transmittance of the original is used as an image signal. In the image reading apparatus that obtains image information by detecting, a light beam splitting unit that splits the focused light from the imaging lens into two with a substantially central portion in the main scanning direction as a boundary is provided. Line sensors having read pixel rows of a plurality of colors for detecting left and right images of the document surface in the main scanning direction are provided on respective optical paths divided from the vicinity of the central portion, and the outputs of these line sensors are digital signals. A / D conversion means for converting into a line is provided, and a line delay means for correcting a read line shift in the sub-scanning direction of read pixel rows of a plurality of colors is provided at a stage subsequent to the A / D conversion means, and the original document. When the output corresponding to the same reading point on the surface appears in the outputs of both the line sensors, the addition result of the output of each subsequent stage is set as the read output, and the output of the subsequent stage is output when the output appears only in one of the line sensors. An image reading apparatus comprising: calculation selecting means for reading output; and shading correcting means arranged after the calculation selecting means for shading correction by calculation of the reading output and a reference level.