JPH09298628A - Image reader - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily demonstrate a base removal read function even in the case of reading the original of a high base density by using a photoelectric conversion unit for colors capable of outputting two colors. SOLUTION: At the time of actual original read, the output of line sensors for the respective colors is digitized by A/D conversion means 22, 23 and 24 by using a reference value set based on the peak value of the output obtained from a peak holding means 25 corresponding to the line sensor for an optional specified color G as the representative data of an original density. Thus, a base removal function is demonstrated, and even when the original of the high base density is read in the photoelectric conversion unit for the colors, read with the clear contrast of a base part and a character part is performed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image reading apparatus such as an image scanner for reading image data from an original, a digital copying machine or a scanner for a facsimile.

[0002]

2. Description of the Related Art At present, image reading devices such as image scanners are used to generate binary or multivalued read data from various originals. Such an image reading apparatus generally has a contact glass, and a document placed on the contact glass is read and scanned by the sub-scanning mechanism. Such a sub-scanning mechanism has a line sensor in which a large number of photoelectric conversion elements are arranged in an array in the main scanning direction, and the line sensor and the document are relatively moved in the sub-scanning direction. Then, the line sensor sequentially reads the image data line by line from the document, so that the read data of the document is generated by the large number of one-dimensional line data.

After that, various correction processes are generally performed on the read data. For example, proper correction data is required for shading correction and density correction. For this reason, in a general image reading apparatus, a white white reference plate, which serves as a reference for image reading, is juxtaposed on the contact glass, and the white reference plate is read and scanned prior to the original to correct the read data. Shading correction is performed as data.

However, when the correction process is executed based on the read data of the white reference plate in this manner, most of the read data of a document whose background portion is not white, such as a newspaper, is gray and inappropriate. . This is generally not preferred as background stains, so an image reading apparatus that sets the AE mode as a switchable operation mode has been put into practical use. The AE function in the AE mode is a function of removing the background by changing the output level of the image signal according to the background (background) of the read document.

For example, in the image reading apparatus disclosed in Japanese Patent Laid-Open No. 3-201774, the read data of the entire area of the original is stored in a memory to detect the maximum density and the minimum density, and the read data is obtained based on these. Correct the density. In such an image reading apparatus, even when an original having a gray background is scanned for scanning, the density of the read data is corrected so that the background becomes white, so that good read data can be generated.

Further, at present, an image reading apparatus which reads black and red from an original and provides two-color printing is put into practical use. In such an image reading apparatus, a color photoelectric conversion unit in which three line sensors for individually reading RGB components are separately arranged in the sub-scanning direction is actually used, and the read image is converted into RGB components. Disassemble and read and scan. Since read data of red and black is generated from the read data of the RGB components, it is possible to print two colors of black and red based on a black-and-white original document in which red is added.

[0007]

An image reading apparatus such as that disclosed in Japanese Patent Laid-Open No. 3-201774 that realizes the AE mode as described above can generate black and white read data from a gray original in a good condition. However, since it is necessary to store the read data of the entire area of the original document in the memory, a large-capacity memory is required, which is not preferable. Although a prescanning method is also disclosed for detecting the maximum density and the minimum density without using a memory, the processing time is increased because press scanning is required.

Particularly, in order to realize the above-mentioned AE mode in an image reading apparatus which reads a document using a color photoelectric conversion unit and outputs two-color data of black and red, RGB components are required. A memory is required for each of the three read data. For this reason, such an image reading apparatus does not implement the AE mode, and when a red document is written on a document with a dark background, the background part is recognized as a part of the color image when read by the color photoelectric conversion unit. As a result, the background dirt is generated.

[0009]

According to the first aspect of the present invention,
A photoelectric conversion unit in which a plurality of line sensors for reading different color component information are spaced apart in the sub-scanning direction, a sub-scanning mechanism for relatively moving the photoelectric conversion unit in the sub-scanning direction with respect to a document, and the photoelectric conversion unit. The peak value of the output obtained from the reference white plate arranged in the sub-scanning range and the line sensor for an arbitrary specific color in the photoelectric conversion unit is held until the reading of the document by all the line sensors is completed. The peak hold means controlled as described above and the output of the line sensor for each color using the reference voltage set in advance for each line sensor in consideration of the tint of the reference white plate when reading the reference white plate. Outputs data for shading correction after digitization, and for the specific color held by the peak hold means when reading the original. And a, and A / D converting means for outputting an image data by digitizing the output of the line sensor for each color using the reference value determined based on the peak value of the output.

Therefore, by reading the reference white plate while sub-scanning the photoelectric conversion unit, shading correction data for each line sensor can be obtained.
At this time, the A / D conversion means digitizes the output of the line sensor for each color using the reference voltage set in advance for each line sensor in consideration of the tint of the reference white plate. The shading correction data can be obtained without disturbing the color balance. On the other hand, at the time of actual document reading, the output of the line sensor for each color is digitized by using the reference value set based on the peak value of the output obtained from the line sensor for any specific color as the representative data of the document density. Therefore, the background removal function can be exerted without disturbing the color balance, and even if the original having a high background density is read by the color photoelectric conversion unit, the contrast between the background portion and the character portion can be clearly read. For this reason also, if the peak value of the output obtained from the line sensor for an arbitrary specific color in the photoelectric conversion unit is held by the peak hold means until the reading of the document by all line sensors is completed, The output can be processed without any problems, and image memory for all images and prescanning are not required.

According to a second aspect of the present invention, in the first aspect of the invention, the peak detection period of the output for the specific color by the peak hold means is defined as the period during which the line sensor for the specific color exists in the original reading area. I am trying. Therefore, the peak value for the specific color can be detected without error, and the operation according to the first aspect of the invention can be surely obtained.

According to a third aspect of the present invention, in the first or second aspect of the present invention, the A / D conversion means uses a reference value determined based on the output for the specific color held by the peak hold means. The background removal reading mode that digitizes the output of the line sensor for each color and the non-background removal reading mode that digitizes the output of the line sensor for each color by using the reference value with a preset fixed value It is switchable. Therefore, depending on the user's needs, there are cases in which it is desired to read the original as the non-background removal reading mode. In this case, the reference value of the A / D conversion means is simply switched so that the reference value of a fixed value is used. Therefore, it is possible to freely select the background removal reading mode and the non-background removal reading mode.

According to a fourth aspect of the present invention, in the first, second or third aspect of the present invention, a data generating means for generating two-color image data from the read data of the document by the plurality of line sensors is further provided. Therefore, for example, when a manuscript having a red circle on a newspaper is read and printed in two colors, it is possible to obtain a good two-color image in which the contrast between the background portion, the character portion, and the red circle portion is clear.

[0014]

An embodiment of the present invention will be described with reference to the drawings. In the image reading apparatus according to the present embodiment, as shown in FIG. 2, the contact glass 2 and the reference white plate 3 are arranged side by side in the sub-scanning direction while being positioned on the upper surface of the main body housing 1. The contact glass 2 is for placing the original 4 to be read, and is provided with an openable / closable original pressure plate or an automatic original conveying device. The reference white plate 3 is for acquiring data for shading correction, and is a substantially white member elongated in the main scanning direction and having uniform density.

Below the reference white plate 3 or the contact glass 2, a reading optical system is arranged. The reading optical system guides the light source 5 that illuminates the reading surface at an angle with respect to the reference white plate 3 or the contact glass 2 surface and the light reflected by the reference white plate 3 or the original 4 to the photoelectric conversion unit 6. 1 to 3 mirrors 7, 8 and 9, an image forming lens 10, and the photoelectric conversion unit 6. Here, the light source 5 and the first mirror 7 are the first traveling body, and the second and third mirrors 8 and 9 are the second traveling body, and the first traveling body is constituted by the sub-scanning mechanism including the motor 11 serving as a drive source. And the second traveling body are configured to sub-scan in the same direction at a speed ratio of 2: 1. The photoelectric conversion unit 6 outputs a voltage proportional to the amount of incident light and passes it to the image processing unit 12 as image data.

The photoelectric conversion unit 6 is for color, and as shown in FIG. 3, line sensors 13, 14, 15 for RGB three colors are arranged in the sub-scanning direction with a space therebetween. Individual line sensors 13, 14, 15
Is an array of many CCD elements in the main scanning direction.
On each surface, for example, a color filter (not shown) that individually transmits the RGB components of the light beam is mounted.
As a result, each of the line sensors 13, 14, 15 has a function of individually reading the RGB components of the light beam.

Next, an example of the configuration of an electrical system mainly including the image processing section 12 will be described with reference to FIG. First, in each line sensor 13, 14, 15 of the photoelectric conversion unit 6, the sample and hold circuits (S / H) 16, 17, 18 and the amplifier 1 are provided.
9, 20, 21 and A / D converter 2 serving as A / D conversion means
2, 23, and 24 are individually connected. Here, the sample hold circuits 16, 17, and 18 are for removing extra signal components such as reset noise.
Further, the amplifiers 19, 20, 21 have variable gain amounts, and the A / D converter 22,
Output to 23 and 24.

In the present embodiment, the line sensor 1 for green G is used as an arbitrary specific color among the three colors.
A peak hold circuit 25 functioning as a peak hold means is connected to 4 via the sample hold circuit 17 and the amplifier 20. An amplifier 26 having a predetermined gain is connected to the output side of the peak hold circuit 25. An image processing block 27 is connected to the control terminal of the peak hold circuit 25, and control signals AEMD-G and P from this image processing block 27 are connected.
The operation is controlled according to HLD-G as shown in Table 1.

[0019]

[Table 1]

On the output side of the peak hold circuit 25, each A / D converter 22,
Switch circuits 28, 29, 30 for switching reference voltages (reference values) of 23, 24 are connected. These switch circuits 28, 29, 30 have a fixed voltage V ₀ , a color-specific fixed voltage Vr, Vb, Vg preset for each color, and a peak value Vp based on the peak hold circuit 25.
Various types of voltage values are selectively input, and control signals WTGT-R, G, B and AE from the image processing block 27 are input.
Switching control is performed as shown in Table 2 according to MD-R, G, and B.

[0021]

[Table 2]

The fixed voltages Vr, Vb,
Vg is a voltage preset in consideration of the tint of the reference white plate 3. That is, when an ideally white document is read, the RGB data after shading correction must have their color balance. Since the shading correction is generally obtained by dividing the original read data by the read data of the reference white plate 3, when the reference white plate 3 itself has a tint, the read data of the reference white plate 3 has the same reference value. If digitized by the A / D converters 22, 23, and 24, the color balance is lost after shading correction. Therefore, in the present embodiment, the color-specific fixed voltages Vr, Vb, V set in consideration of the tint of the reference white plate 3 so that the color balance is not lost after shading correction.
g is used as the reference value.

Therefore, each of the A / D converters 22, 23, 24 in the present embodiment uses each of the line sensors 13, 14, 1 by using the reference value selected by switching at each operation time.
It has a function to convert the data input from the 5 side into digital data.

Further, the image processing block 27 performs well-known image processing such as black level correction, shading correction, MTF correction and the like, and uses any one of the R, G and B signals for the target color in the original 4. For example, the red color is recognized and is output to the writing block 31 as two-color data consisting of black writing image data and red writing image data. A CPU 32 and a memory 33 are connected to the image processing block 27. The CPU 32 is an image processing block 27.
The write and write blocks 31 are subjected to control processing such as read, write mode setting, and timing signal transmission, and the memory 3
3 is used to store data during image processing such as shading correction.

The operation control in the AE mode (background removal reading mode) in such a configuration will be described with reference to the time chart shown in FIG. First, when a read request is issued, the CPU 32 sends a read mode (AE mode, two-color output mode,
(Variable mode, etc.) and set the first light source 5 etc.
Move the running body to the reading position of the reference white plate 3, and use XLEA.
The D signal is sent to the image processing block 27. As a result, when the reference white plate 3 is read, the image processing block 27 displays X
After receiving the LEAD signal, the A / D converters 22, 23,
The AEMD-R, G, B signals are all set to "L" level so that the reference values of 24 become the color-specific fixed voltages Vr, Vb, Vg, respectively, and the WTGT-R, G, B signals are sequentially set to "H" level. To do. Next, in order to determine the gains of the amplifiers 19, 20, and 21, the reference white plate 3 is read for one line or several lines with a predetermined gain. The image processing block 27 resets the gains of the amplifiers 19, 20, and 21 based on the outputs of the A / D converters 22, 23, and 24 at this time. This function is known as the auto gain control function, but this function is not essential for implementing the present invention.

In this state, the reference white plate 3 is read while moving the reading optical system in the sub-scanning direction, and the A / D converters 22, 23, 24 are read from the line sensors 13, 14, 15 respectively.
The image processing block 27 captures the output obtained through the above.
During this operation, the WTGT-R, G, B signals are sequentially returned to the "L" level at the timing when the first traveling body passes through the reference white plate 3, and the data acquired during this period is R, G, Each pixel of each color B is averaged and stored in the memory 33 as shading correction data in a form in which random noise is removed.

The WTGT-R, G, B signals are sequentially at "H" level at intervals (timings) A, B, respectively.
It is switched to "L" level. This is because the photoelectric conversion unit 6 is used in order to effectively use the width of the narrow reference white plate 3.
This is because the reading is started in the order in which the line sensors 13, 14 and 15 entered the area of the reference white plate 3, and the reading ends when the line sensors 13, 14 and 15 are outside the area. Here, the number of lines corresponding to the distance between the line sensors 13 and 14 at the same-magnification reading is X, the number of lines corresponding to the distance between the line sensors 14 and 15 at the same-magnification reading is Y, and the scaling factor is N (for example, 25 to 25). A) is X * N / 100 (line),
B is a value of (X + Y) * N / 100 (line).

Next, the image of the original 4 is actually read. CP
U32 moves the first running body to the document area, and
The XSCAN signal is transmitted to the image processing block 27. The image processing block 27 receives the XSCAN signal to read the original, and the A / D converters 22 and 2 are read.
The WTGT-R, G, B signals are set to "L" level and the AEMD-R, G, B signals are sequentially set to "H" so that the reference values of 3, 24 become the peak values obtained from the peak hold circuit 25. "Set to level. Again, AEM
R, G, and B signals are sequentially switched to the “H” level at intervals (timings) A and B, respectively.
This is because G and B (line sensors 13, 14, and 15) start reading from the same position on the original 4, and the values of A and B are as described above.

Here, the period during which R, G, and B read the original 4 is about the time length of XSCAN being at "L" level after AEMD-R, G, and B are switched to "H" level. . However, since the AEMD-G enables the peak hold circuit 25 until the line sensor 15 for B, which is the most delayed, finishes reading the document 4,
It is controlled to remain at the "H" level. On the other hand, regarding the most preceding R line sensor 13, it is sufficient that the AEMD-R is at the “H” level only during the period of reading the document of its own, but in the present embodiment, according to the AEMD-G. is there.

The PHLD-G signal for peak hold is
During the period in which the G line sensor 14 reads the original 4, the line sensor 14 is turned on for an arbitrary period for each line to perform the peak detection operation.
In this way, the original 4 is transferred to the line sensors 13, 14, 15
The R, G, B data read by the A / D converters 22 and 2
After being digitized by using the reference value determined based on the peak value of the output obtained from the line sensor 14 by 3, 24 and taken into the image processing block 27, the image of known black level correction, shading correction, etc. It is processed and output to the write block 31.

As described above, according to this embodiment, since the peak value of the output of the G line sensor 14 is used as the reference value as the representative data of the document density during the document reading, the document is quantized. Even if 4 has a high background density, the background can be removed. Also,
When reading the reference white plate 3, the fixed voltages Vr, Vb, V for each color are read.
Color-balanced shading correction data is acquired using g, and quantization is performed using the peak value of the peak hold circuit 25 that is common to each color when the original 4 is read, so it is obtained after shading correction. The color balance of the received image signal is not lost. It should be noted that the arbitrary specific color is not limited to green G, but if the specific color is green G, the recognition is closest to that of human vision, and therefore the background removal reading with less unnaturalness is performed.

On the other hand, when the non-background removal reading mode is designated, the control is performed as shown in the time chart of FIG. First, if there is a read request in this mode, C
The PU 32 and the image processing block 27 output signals such as XREAD at the same timing as in the background removal reading mode, obtain shading correction data based on the reference white plate 3, and store the shading correction data in the memory 33.

Next, the process moves to reading the original 4. CPU
Reference numeral 32 moves the first traveling body to the original area and transmits an XSCAN signal to the image processing block 27. After receiving the XSCAN signal, the image processing block 27 sets the WTGT-R, G, B so that the reference values of the A / D converters 22, 23, 24 become the common fixed voltage V _0.
The signal and the AEMD-R, G, B signals are set to "L" level. Also in this case, each line sensor 13, 14, 15
The period during which the original 4 is read is the same as that in the background removal reading mode. In this way, the line sensors 13, 14, 1
The R, G, and B signals obtained by reading the original 4 with the A / D converter 2
2, 23, and 24, the fixed voltage V ₀ is quantized as a reference value, the image is quantized by the image processing block 27, and then, known black level correction, shading correction, and the like are processed and output to the writing block 31. It In this case, the fixed voltage V ₀
Since the image data is actually output with the density of the original 4 as it is seen, the color balance of R, G, and B is not lost even after the shading correction.

When the image processing block 27 recognizes a target color (for example, red) in the original image when the image is read in the background remover reading mode, R, which is obtained from the line sensors 13 and 14, Using the G signal, the data generating means compares the magnitude relationship and the like, and the original 4
This is performed by recognizing the target color image therein and transmitting the black writing image data and the red writing image data to the writing block 31.

[0035]

According to the first aspect of the present invention, the reference white plate is read while the photoelectric conversion unit is sub-scanned to obtain the shading correction data for each line sensor. The D conversion means digitizes the output of the line sensor for each color by using the reference voltage preset for each line sensor in consideration of the tint of the reference white plate, and thus the color balance between the line sensors is lost. It is possible to obtain shading correction data that is unique, and at the time of actual document reading, the reference value set based on the peak value of the output obtained from the line sensor for any specific color is used as the representative data of the document density. Since the output of the line sensor for each color is digitized, it is possible to exert the background removal function without disturbing the color balance, Be read dark originals background density in the photoelectric conversion unit for error can be performed clear reading of the contrast between the background portion and the character portion,
For this reason also, if the peak value of the output obtained from the line sensor for an arbitrary specific color in the photoelectric conversion unit is held by the peak hold means until the reading of the document by all line sensors is completed, The output can be processed without any trouble, and it can be realized without requiring an image memory for all images or prescanning.

According to a second aspect of the present invention, in the first aspect of the invention, the peak detection period of the output for the specific color by the peak hold means is set in the reading area of the original by the line sensor for the specific color. The peak value for the specific color can be detected without error, and the effect of the invention according to claim 1 can be reliably exhibited.

According to the invention of claim 3, in the invention of claim 1 or 2, the A / D conversion means is set to a reference value determined based on the output for the specific color held by the peak hold means. The background removal reading mode that digitizes the output of the line sensor for each color using it, and the non-background removal reading mode that digitizes the output of the line sensor for each color using the reference value with a preset fixed value Since the values are switchable, there are cases where the user needs to read the original in the non-background removal reading mode, but in this case, the A / D conversion means uses a reference value that is a fixed value. It is possible to respond by simply switching the reference value, and it is possible to freely select the background removal reading mode and the non-background removal reading mode.

According to the invention of claim 4, claim 1,
In the invention described in item 2 or 3, since the image forming device further includes data generating means for generating image data of two colors from the read data of the original document by a plurality of line sensors, for example, by reading an original document in which a red circle is marked on newspaper, In the case of color printing, a good two-color image can be obtained in which the contrast between the background portion, the character portion and the red circle portion is clear.

[Brief description of drawings]

FIG. 1 is a block diagram mainly showing an image processing unit according to an embodiment of the present invention.

FIG. 2 is a front view showing a schematic configuration of an image reading apparatus.

FIG. 3 is a schematic front view showing the arrangement of line sensors in the photoelectric conversion unit.

FIG. 4 is a time chart showing an operation control example in a background removal reading mode.

FIG. 5 is a time chart showing an operation control example in a non-background removal reading mode.

[Explanation of symbols]

 3 Reference white plate 4 Original 6 Photoelectric conversion unit 13-15 Line sensor 22-24 A / D conversion means 25 Peak hold means

Claims (4)

[Claims] 1. A photoelectric conversion unit in which a plurality of line sensors for reading different color component information are spaced apart in the sub-scanning direction, and a sub-scanning mechanism for moving the photoelectric conversion unit relative to the original in the sub-scanning direction. , A reference white plate arranged in a sub-scanning range by the photoelectric conversion unit and a peak value of an output obtained from a line sensor for an arbitrary specific color in the photoelectric conversion unit, Peak holding means that is controlled to hold until the end of reading, and when reading the reference white plate, a reference voltage preset for each line sensor is used in consideration of the tint of the reference white plate for each color. The shading correction data is output by digitizing the output of the line sensor, and the peak hold means holds the data when reading the document. A / D conversion means for digitizing the output of the line sensor for each color using the reference value determined based on the peak value of the output for the specified color and outputting the image data. Image reading device. 2. The image reading according to claim 1, wherein the peak detection period of the output for the specific color by the peak hold means is a period in which the line sensor for the specific color exists in the original reading area. apparatus. 3. A background removal reading mode in which the A / D conversion means digitizes the output of the line sensor for each color using a reference value determined based on the output for the specific color held by the peak hold means. 3. The reference value is switchable between a non-background removal reading mode in which the output of the line sensor for each color is digitized using a reference value based on a preset fixed value. Image reading device. 4. The image reading apparatus according to claim 1, further comprising a data generation unit that generates image data of two colors from the read data of the original document by a plurality of line sensors.