JP2009182379A - Image reader and image processing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image reader with a function of reading an image of a transparent document, which is free from the reduction of throughput of image reading and does not require a high-level image processing means nor a large-capacity image work memory, to be mounted thereon. <P>SOLUTION: The image reader includes: a visible light source for illuminating a transparent document; an infrared light source for illuminating the transparent document; a line image sensor for reading an image of the transparent document; an imaging optical system for forming the image of transparent document illuminated by the visible light source or the infrared light source, on the line image sensor; a reading means for lighting the visible light source and the infrared light source alternately to read a visible line image and an infrared line image by the line image sensor; and an addition means for adding data showing the presence or the absence of defects obtained from pixel data of the infrared line image, to pixel data of the visible line image. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

The present invention relates to an image reading apparatus having an image reading function, and more particularly to an image reading function that has an image reading function of a transparent original image and reduces the influence on a defect due to dust or scratches on the transparent original to be read. The present invention relates to an apparatus and an image processing method.

  2. Description of the Related Art Conventionally, in an image reading apparatus that reads a transparent original image, the following is performed in order to reduce the influence on the image caused by a defect due to dust or scratches on the transparent original to be read. That is, as a line CCD sensor for visible light, an RGB sensor provided with a color separation filter for color separation into three primary colors RGB light is arranged line by line, and an IR cut filter for shielding IR light is provided. For invisible light, an infrared (IR light) line CCD sensor is arranged in parallel with the visible light image sensor.

  Then, the visible light source and the invisible light source are turned on at the same time, and the reading by the visible light and the reading by the invisible light are simultaneously performed, and the defective pixel is detected in detail without reducing the reading speed (for example, Patent Documents) 1).

  Alternatively, as means for separating visible light and invisible light, a dichroic mirror or the like is provided, infrared (IR) light that is invisible light is dispersed, and an image is read by an individual sensor.

  Further, when image processing is performed, it is determined that a pixel whose infrared component level of the image of the transparent original is equal to or lower than the first infrared level (threshold) is a defective pixel of the transparent original. That is, the defect infrared component level is detected, and (first infrared component level) / (defect infrared component level) is calculated based on the first infrared component level and the defect infrared component level. It is known that a correction coefficient is obtained, a visible component level of a visible component of a transparent original image is detected, a defect visible component level at a defect position of the transparent original is multiplied by a correction coefficient, and a corrected visible component level is calculated. (For example, refer to Patent Document 2).

Further, an image reading apparatus using an off-axial optical system that does not cause chromatic aberration is known (for example, see Patent Document 3).
JP 2003-110801 A JP-A-11-98370 JP 2002-335375 A

  In the conventional example, first, the visible light source is turned on, and the visible light image data of the transparent original is read. Then, the light source is switched, the invisible light source is turned on, and the invisible light image data of the transparent original is read. Or, reverse the procedure. Alternatively, both the visible light source and the invisible light source are turned on, and the visible image data and the invisible image data are read simultaneously by dedicated sensors. In any case, the invisible image data obtained from the image reading operation using the invisible light source detects the defect position on the transparent original due to dust or scratches, and the visible image data obtained from the image reading operation using the visible light source is converted into the visible image data. The image is corrected at the corresponding defect position.

  However, since the visible light image data and the invisible light image data obtained by the above two image reading operations are stored, a memory having a capacity approximately twice as large as that of the original image to be read is required. Usually, the above processing is performed by a host PC mounted with a large-capacity image memory or connected to the image reading apparatus.

  Furthermore, there is a problem that special parts such as a dedicated sensor for invisible light other than the above for visible light and a dichroic mirror are required.

  The present invention can detect defective pixels due to dust and scratches on a transparent document and perform image correction without using a special part such as a sensor equipped with an IR cut filter, a dichroic mirror, or a large-capacity image memory. It is an object of the present invention to provide an image reading apparatus that can be used.

The present invention provides an image reading apparatus having an image reading function for transparent originals, which does not need to be equipped with advanced image processing means and a large-capacity image work memory without reducing the image reading throughput. The purpose is to provide.

The present invention provides an image reading apparatus including a transparent original image reading unit, wherein a visible light source that illuminates the transparent original image, an invisible light source that illuminates the transparent original image, and the read transparent original image are electrically converted. An image sensor that converts the image signal into an image signal; an optical system that forms an image on the image sensor of an optical signal generated by the visible light source and the invisible light source illuminating the transparent original image; In addition, the image reading apparatus includes a data acquisition control unit that acquires visible image data and invisible image data by alternately turning on the visible light source and the invisible light source.

According to the present invention, since the defective pixel is corrected by referring to the image data of the previous line having no defect, there is an effect that it is not necessary to use a special part such as a sensor equipped with an IR cut filter or a dichroic mirror. Play.

  The best mode for carrying out the invention is the following embodiment.

  FIG. 1 is a diagram illustrating an image reading apparatus 100 that is Embodiment 1 of the present invention.

  The image reading apparatus 100 includes a main body 11 and a document pressing unit 12.

  The main body 11 of the image reading apparatus 100 is provided with a moving optical unit 3, an image reading apparatus control circuit 5, and a document table glass 7.

  The document pressing unit 12 is provided with a transparent document light source unit 2 and a transparent document light source unit drive circuit 4. The read original 6 is illuminated by the transparent original light source unit 2.

  FIG. 2 is a diagram showing details of the moving optical unit 3.

  The moving optical unit 3 includes a reflection original light source unit 31, a plurality of plane reflection mirrors M1, M2, and M3, off-axial reflection surfaces M11 to M12, and a line image sensor 32 that is an imaging unit and a line sensor. Has been. The off-axial reflecting surface is a reflecting surface using a curved surface whose surface normal at the intersection with the reference axis of the component surface is not on the reference axis when considering a reference axis along a ray passing through the center of the image and the center of the pupil. is there.

  The optical signal generated by the transmissive original light source unit 2 passes through the original table glass 7 and reaches the line image sensor 32 disposed in the movable optical unit 3.

  Further, in the scanner main body 11, driving means such as a pulse motor, an endless belt, a pulley, a gear train, and a guide rail (not shown) are provided for sliding the movable optical unit 3.

  The movable optical unit 3 is slidably mounted on the guide rail by the mounting means, and the mounting means is fixed to an endless belt. The image reading device control circuit 5 and the pulse motor are electrically connected by a cable (not shown). The image reading device control circuit 5 drives the pulse motor, and can move the moving optical unit 3 through the endless belt via the gear train and pulley.

  Note that there is provided reading means (not shown) for alternately turning on the visible light source and the infrared light source and reading the visible line image and the infrared line image with the line image sensor.

  Further, there is provided an adding means (not shown) for adding defect presence / absence data obtained from the pixel data of the infrared line image to the pixel data of the visible line image.

  Further, a line memory (not shown) is provided for storing one line output from the adding means.

  Then, correction means (not shown) for correcting the data of the pixel of interest to which the data having the defect is added using the change amount of the data of the pixel having no defect adjacent to one line stored in the line memory. ) Is provided. This correction means regards the pixel data corrected by the correction means as pixel data having no defect, and uses it for correcting a pixel having a defect in the next line. The correcting means is a means for executing the step of sequentially correcting the visible component data at the defect position based on the visible component data around the defect-free area and storing the data in the storage device.

  In this case, the surrounding visible component data without a defect is the visible component data at the same position one line before without a defect until the line with the defect identified first. Then, for the defect positions of the second and subsequent lines specified subsequently, the visible component data of the already corrected line is obtained. The “defect position after the second line specified continuously” is a line having a defect, which is the next line after the corrected line.

  Further, the visible component data around the defect-free surroundings is the visible component data one pixel before without any defect with respect to the pixel with the defect identified first, and the two or more pixels identified subsequently. For the defect position, the visible component data one pixel before is corrected.

  Note that the above-described embodiment may be applied to an apparatus having a configuration in which an original is moved relative to a fixed optical unit and read in a direction perpendicular to the line of the line image sensor.

  Next, a transparent original image reading operation of the image reading apparatus 100 will be briefly described.

  The reading operation in the image reading apparatus 100 is started by a read command command from the host device or a system in the image reading apparatus 100. The image reading device control circuit 5 controls the transmissive document light source unit 2 disposed in the document pressing section 12 to light up and reflects the transmitted light by a plurality of mirrors M1, M2, and M3. The reduction optical system type moving optical unit 3 reads an image for one line in the main scanning direction by forming an image on the line image sensor 32 through the off-axial reflecting surfaces M11 and M12. Further, the moving optical unit 3 may be a close-contact unit of the same size, in which a light source group, a rod lens, and a line image sensor are integrated, such as a CIS (Contact Image Sensor).

  In addition, the endless belt is driven by rotating the pulley with the power of the pulse motor by the gear train. As a result, the movable optical unit 3 fixed to the endless belt by the placing means moves on the guide rail in the sub-scanning direction indicated by the arrow Y.

  The image reading apparatus 100 repeats reading the line image in the main scanning direction while moving the movable optical unit 3 in the sub-scanning direction.

  The image reading apparatus 100 moves the moving optical unit 3 to the end position of the read image while performing the reading operation according to the content of the read command from the host device or the built-in system. As a result, the entire surface of the read document 6 placed on the document table glass 7 can be scanned. For the read image range specified by the host or built-in system, the pixel range to be adopted in the sensor output is defined in the main scanning direction, and the movement range of the optical unit is defined in the sub-scanning direction. It is defined by the control means described later on the substrate.

  FIG. 3 is a perspective view showing the light source unit 2 for a transparent document.

  The light source unit 2 for the transmissive document is disposed in the document pressing unit 12 and includes a light guide 21, a visible light source unit 22, an invisible light source unit 23 that is an infrared light source, and a transmissive document guide unit 24. These are electrically connected to the light source unit driving circuit 4 for the transmissive original via a cable (not shown).

  FIG. 4 is a diagram illustrating the timing at which the image reading device control circuit 5 controls the transparent original light source unit 2 to light up via the transparent original light source unit drive circuit 4 in the first embodiment.

  The synchronization signal SH is a synchronization signal for reading the line image in the main scanning direction. One line reading time is from one synchronization signal SH to the next synchronization signal SH. During this one-line reading time, the visible light light (LED_ON) and the invisible light lighting signal (IR_ON) are sequentially turned on (alternately lighted), so that the visible component light in the same reading line can be obtained. The visible image data and the non-visible image data of the light of the invisible component are obtained.

  Further, the imaging optical system using the off-axial reflecting surfaces M11 to M12 can form an image on the image sensor without generating chromatic aberration between the visible image and the invisible image. In other words, the refraction angle at the refracting surface of the refraction lens varies depending on the color, so chromatic aberration occurs. No chromatic aberration occurs.

  In addition, this enables reading without causing an error in the scanning of the reading position by the drive system twice, and can detect the defect position at high speed.

  Conventionally, there is also a transmissive original image reading apparatus that uses a monochromatic (R / G / B) LED as a visible light source and an IR-LED as a non-visible light source, and is configured with an equal magnification optical system to minimize the influence of chromatic aberration. . In this conventional example, on / off can be controlled for each reading line, the aberration due to the equal-magnification optical system can be reduced, and the detected defect position accuracy can be improved.

  However, in this case, the visible light source is R-G-B, then the non-visible light source IR is sequentially controlled so that the four light sources are turned on, and the visible image and the non-visible image are line-sequentially conveyed while the carriage or document is being conveyed. Read. As a result, at the time of 1200 dpi scan, the sub-scan feed is 25.4 mm ÷ 1200 dpi => 0.02 mm / pixel. This means that the reading position of the visible image and the non-visible image is shifted by a maximum of 0.01 mm. General garbage and hair are 0.05 mm (average of Western women) to 0.08 mm (average of Japanese women), cotton dust is 0.03 mm, and feathers and pollen are 0.05 mm. . This error cannot be ignored in order to detect such dust, and this error hinders detection of a highly accurate defect position.

  In the first embodiment, as shown in FIG. 4, the visible image reading line and the non-visible image reading line are controlled by the same line, so that the reading position shift as described above does not occur. As a result, the reading position error caused by the conventional light source lighting control and the line-sequential reading method can be eliminated in the first embodiment, and the defect position can be detected at high speed and with high accuracy. .

  A general RGB three-line sensor is used to perform color reading. The positional deviation of the three lines in the sub-scanning direction is adjusted by a configuration using a known line buffer. Here, infrared reading is performed by an R line sensor. In the following, for the sake of simplicity, only the R signal is described for the visible image data.

  FIG. 5 is a flowchart illustrating a series of operations for specifying a defect position in the first embodiment.

  In the image reading apparatus 100, an image reading operation is started by a transparent original image reading command from a connected host or a system built in the image reading apparatus 100 (S0). The command at this time is transparent original image reading. Upon receiving this command, the image reading apparatus 100 sequentially turns on the visible light source unit 22 and the invisible light source unit 23 according to the timing shown in FIG. The invisible image data is scanned (S1).

  Depending on the order of lighting, the visible image data (16 bits) obtained previously is stored for one line and buffered in order to overlap with the invisible image data of the same line obtained later (S2). ). Based on the invisible image data (16 bits) that is infrared image data obtained later, it is determined whether or not the read transparent original image includes defective pixels due to dust and scratches (S3). If there is a defective pixel, the error flag “true” (1 bit) indicating the defective image is turned ON in the data at the same position of the visible image data buffered in S2 (S4). If there is no defective pixel, the error flag is turned off.

  As a result, regardless of whether the error flag is ON / OFF (that is, regardless of the presence / absence of a defective pixel), the visible image data having 17 bit data / pixel is written to the image memory in the image reading apparatus 100 (S5). Thereafter, the defective pixel is corrected by executing the defective pixel correction described later (S6). It is determined whether the visible image lines corresponding to the line designated by the scan command have been read (S7). If reading of the designated line is not completed, the next line is scanned and a series of operations are repeated. When the designated final line has been scanned, the image reading operation is terminated (S8).

  The level of the invisible image data is detected, and the level of the detected invisible image data is compared with a preset parameter to detect the presence or absence of a defect in the visible image.

  By the above process, visible image data (17 bits) to which defective pixel position information is added can be obtained, and the defective pixel position is detected.

  FIG. 6 is a diagram illustrating an example of defective pixel correction.

  By checking the error flag of the visible image data (17 bits) to which the obtained defective pixel position information is added, it is determined whether or not the corresponding pixel is defective. If it is determined that the pixel is a defective pixel, the determined defective pixel X (i, j) can be expressed by the following equation. Here, i and j are coordinates on the image.

X (i, j) = X (i-1, j) + | X (i + 1, j-1) -X (i, j-1) |
X (i, j) is a defective pixel (target pixel), and X (i−1, j) is a line including the defective pixel, which is data one pixel before the defective pixel. is there. | X (i + 1, j−1) −X (i, j−1) | is the same position change amount one line before, that is, the line adjacent to the line one line before the defective pixel is included. , The amount of change between the data of the pixel having the same horizontal direction as the defective pixel and the data of the pixel to the right of the defective pixel.

  When correcting the defective pixel X (i, j), reference is made to the data of j−1 one line before in the sub-scanning Y direction that is not determined to be a defective pixel. Then, the difference (change amount) between the data of the pixel at the same position i in the main scanning X direction and the data of the pixel i + 1 after that pixel is added to the data of the pixel i-1 before the pixel in the X direction. It is assumed that the obtained pixel is a correction pixel of the defective pixel X (i, J). By correcting the defective pixel in this way, an image equivalent to the image one line before the defective pixel can be obtained.

  The corrected pixel data is regarded as image data having no defective pixel, and the entire defective portion existing on the document can be corrected by repeating the same correction as described above.

  Here, since the G and B line image sensors do not read the infrared image, no additional data is attached. By aligning the positional deviation of the three-line image sensor by a known method using a line memory, the G and B image data corresponding to the defect position of the R image data are similarly corrected for each color.

  As described above, at least one line of image memory is provided, and pixel data for defect correction is created based on image data of one line before without defective pixels, and correction is performed for each line by performing correction. be able to. By this correction, without reducing the throughput of image reading, and without installing a large-capacity image memory more than twice as large as the read image, the read transparent original image can be reduced to an image caused by defects such as dust and scratches. The influence can be corrected.

  In a conventional transmission original image reading apparatus, a cold cathode tube lamp (CCFL) is often used as a visible light source and an infrared light (IR) LED is used as an invisible light source. It is impossible to turn off the lights. Also, since CCFL light and infrared light are imaged on the line image sensor by the same lens, the difference in the emission wavelength becomes an aberration, which causes the detected defect position to shift. Therefore, for each of the visible image and the invisible image, after the reading operation is finished up to the end position of the read image, the entire image data is subjected to scaling processing, and then the defective pixel is specified from the image data. . In this case, since it is necessary to perform the image reading twice, an error occurs in the reading position by the drive system, and the detection of the highly accurate defect position is hindered.

  In Example 1, a white LED capable of high-speed lighting control is used as a visible light source, and an IR-LED (infrared light LED) capable of high-speed lighting control is used as a non-visible light source. In addition, the lighting-off is controlled respectively. Therefore, in Example 1, a visible image and a non-visible image can be simultaneously acquired for each reading line.

  According to the above embodiment, since the defective pixel is corrected by referring to the image data of the previous line having no defect, it is not necessary to use a special part such as a sensor equipped with an IR cut filter or a dichroic mirror.

According to the above embodiment, the error flag is added to the visible image data of the defective pixel after the position of the defective pixel at the infrared component level is specified. There is no need to use the image memory.

1 is a diagram illustrating an image reading apparatus 100 that is Embodiment 1 of the present invention. FIG. FIG. 3 is a diagram showing details of the moving optical unit 3. FIG. 6 is a perspective view showing a light source unit 2 for a transparent document. FIG. 6 is a diagram illustrating a timing at which the image reading device control circuit 5 controls the transparent document light source unit 2 to be turned on via the transparent document light source unit drive circuit 4 according to the first exemplary embodiment. 5 is a flowchart illustrating a series of operations for specifying a defect position in the first embodiment. It is a figure which shows an example of defect pixel correction | amendment.

Explanation of symbols

100: Image reading device,
11 ... Main body of the image reading apparatus 100,
12 ... Document holding portion of the image reading apparatus 100,
2 ... Light source unit for transparent original,
21 ... light guide,
22: Visible light source unit,
23 ... Invisible light source unit,
24 ... Transparent document guide unit,
3 ... Moving optical unit,
31 ... Light source unit for reflection original,
M1, M2, M3 ... reflective mirrors,
M11, M12 ... off-axial reflecting surface,
32. Image sensor,
4 ... Light source unit drive circuit for transparent original,
5 ... Image reading device control circuit,
6 ... Scanned document,
7: Document glass.

Claims (10)

An image reading apparatus that scans and reads a transparent original,
A visible light source for illuminating the transparent original;
An infrared light source for illuminating the transparent original;
A line image sensor for reading the image of the transparent original;
An imaging optical system that forms an image of the transmission original illuminated by the visible light source or the infrared light source on the line image sensor;
Reading means for alternately turning on the visible light source and the infrared light source and reading the visible line image and the infrared line image with the line image sensor;
An adding means for adding defect presence / absence data obtained from the pixel data of the infrared line image to the pixel data of the visible line image;
An image reading apparatus comprising: In claim 1,
The visible light source is a white LED that can be turned on at high speed. In claim 1,
An invisible light source is an infrared LED capable of high-speed lighting control. In claim 1,
An image reading apparatus, wherein the imaging optical system is formed of an off-axial reflecting surface. In claim 1,
An image reading apparatus, wherein the imaging optical system is an equal-magnification rod lens. In claim 1,
Scanning means for relatively moving the image of the transparent original and the line image sensor in a direction perpendicular to the line image sensor;
A line memory for storing one line of the output of the additional means;
Correction means for correcting the data of the pixel of interest to which the data having the defect is added using the change amount of the data of the pixel having no defect adjacent to one line stored in the line memory;
An image reading apparatus comprising: In claim 6,
The image reading apparatus characterized in that the correction means regards the pixel data corrected by the correction means as pixel data having no defect, and uses it for correcting a pixel having a defect in the next line. Decomposing a transparent original image into a visible component (R / G / B) and an invisible component (IR), and storing them in a storage device;
Detecting the level of the invisible component and storing it in a storage device;
Identifying the defect position of the read transparent original image and storing it in a storage device;
A step of sequentially correcting the visible component data at the defect position based on the visible component data around the defect-free area and storing it in a storage device;
An image processing method comprising: In claim 8,
Visible component data around the defect is the visible component data at the same position one line before the previous defect with respect to the line with the defect identified first, followed by the two identified lines An image processing method characterized in that the subsequent defect positions are visible component data of already corrected lines. In claim 8,
The surrounding visible component data having no defect is the visible component data one pixel before without any defect for the first defective pixel identified, and the defect after the second pixel identified subsequently. An image processing method characterized in that for a position, it is already corrected visible component data one pixel before.

Images