JP7515134B2 - IMAGE PROCESSING APPARATUS, IMAGE PROCESSING METHOD, IMAGE PROCESSING PROGRAM, AND IMAGE READING APPARATUS - Google Patents

Description

The present invention relates to an image processing device, an image processing method, an image processing program, and an image reading device that process read image data obtained by reading an object in a line.

Typical systems that have traditionally been used in surface inspection equipment that mainly uses a light source in the visible range are line sensors, contact image sensors, or a combination of a scanning optical system using a laser beam and a light receiving optical system including a photoelectric conversion element (photomultiplier, avalanche photodiode, CCD sensor, CMOS sensor, etc.) and a light guide, and most of these are reflective types that receive reflected light or fluorescence from scratches, unevenness, defects, missing parts, attached foreign matter, etc. on the inspection object (read object). In addition, examples of inspection objects include paper products, food, and for industrial use, resin molded objects such as films or sheets.

In contrast, in a transmission type where the light receiving system and the illumination system are arranged opposite each other with the object to be inspected in between, the objects to be inspected are often transparent, thin, and highly transmittant. However, there are very few systems that can detect foreign objects, scratches, defects, and other objects contained in thick objects to be inspected.

X-ray inspection equipment, which is renowned as a non-destructive testing device with excellent transparency, uses X-rays, which are a type of radiation, so it is necessary to set up a radiation control area and control the amount of radiation exposure to people. In other words, there are high hurdles in deciding where to install them. Moreover, because they are large and heavy, it is not easy to add them to a factory's existing production line. In addition, because they are expensive, it is not possible to install many inspection points.

Furthermore, the excellent transparency of X-rays themselves is a drawback for X-ray inspection equipment, as they can penetrate even foreign objects, defects, and scratches, making it often impossible to distinguish between them.

International Publication No. 2015/186566

Patent Document 1 discloses an example of a conventional image reading device equipped with a line light source and a line sensor. In conventional transmission type image reading devices, like X-ray inspection devices, light emitted from the line light source is scattered due to foreign objects, defects, scratches, etc. contained in the inspection object.

In particular, when the object to be inspected is thick, and the object is scanned in a line by an image reader and a data format image showing the object is generated, the contour of the object area, which is the area on the image that corresponds to the object, tends to become blurred. The object area is specifically the area of pixels that make up the object on the image.

The inventors of the present application focused on the blurring of the contours of the target area on the image as described above, and after extensive investigation, came to the conclusion that while the blurring of the contours of the target area on the image caused by stray light is eliminated in the sub-scanning direction, the blurring is not eliminated in the main scanning direction, which is why clear output image data cannot be obtained.

The present invention has been made in consideration of the above-mentioned situation, and aims to provide an image processing device, an image processing method, an image processing program, and an image reading device that can generate output image data with clear contours of the target area by applying image processing to multiple read image data obtained by reading an object in a line.

The image processing device according to the present invention is an image processing device that performs image processing on multiple read image data obtained by reading the same object in lines along main scanning directions that extend in different directions, and the image processing device includes a spectrum generation processing unit, a comparison processing unit, a synthetic spectrum generation processing unit, and an image generation processing unit. The spectrum generation processing unit generates multiple spectral data by performing Fourier transform processing on each of the multiple read image data. The comparison processing unit compares each component corresponding to the same frequency in the multiple spectral data. The synthetic spectrum generation processing unit generates synthetic spectral data by selecting one of the components corresponding to the same frequency in the multiple spectral data based on the comparison result by the comparison processing unit. The image generation processing unit generates output image data by performing inverse Fourier transform processing on the synthetic spectral data.

With this configuration, it is possible to generate output image data with clear contours of the area corresponding to the target object.

The synthetic spectrum generation processing unit may select the largest component from among the components corresponding to the same frequency of the multiple spectral data.

With this configuration, it is possible to generate output image data in which the contours of the area corresponding to the object are clearer. It is also believed that with this configuration, the greater the number of read image data, the clearer the contours of the area corresponding to the object can be generated.

The image processing device may further include a correction processing unit that performs a correction process on at least one of the plurality of read image data before Fourier transform processing is performed so that the area corresponding to the object is positioned approximately identically on the image in each of the plurality of read image data. In this case, the spectrum generation processing unit may perform a Fourier transform process on each of the plurality of read image data that has been subjected to the correction process by the correction processing unit.

With this configuration, the area corresponding to the object can be positioned approximately identically in each scanned image data.

The correction processing unit may correct the inclination or magnification of an area in the image of at least one of the plurality of read image data before the Fourier transform processing is performed, which corresponds to the object.

With this configuration, by correcting the inclination or magnification of the area corresponding to the object in the scanned image data, the area corresponding to the object can be positioned approximately identically in each scanned image data.

The image reading device according to the present invention comprises a plurality of line sensors and the image processing device. The plurality of line sensors are each capable of reading an image in a line shape along a main scanning direction, and the main scanning direction extends in a different direction from the sub-scanning direction. The image processing device performs image processing on a plurality of read image data obtained by reading the same object with the plurality of line sensors.

With this configuration, multiple scanned image data can be generated simply by moving the object or multiple line sensors in the sub-scanning direction, i.e., with just one operation. In other words, with this configuration, multiple scanned image data can be obtained efficiently.

The plurality of line sensors may include two line sensors arranged so that their main scanning directions are perpendicular to each other.

With this configuration, multiple scanned image data can be obtained efficiently using two line sensors.

The two line sensors may be arranged so that their respective main scanning directions intersect at an angle of 45° with respect to the sub-scanning direction.

With this configuration, multiple scanned image data can be obtained more efficiently using two line sensors.

The multiple line sensors may be arranged in a direction intersecting the sub-scanning direction.

With this configuration, the object can be divided and read by multiple line sensors arranged in a line in a direction intersecting the sub-scanning direction. This makes it possible to narrow the area in which the line sensors are arranged in the sub-scanning direction, thereby saving space.

The image reading device of the present invention comprises at least one line sensor and the image processing device. The at least one line sensor has an elongated shape along a longitudinal direction intersecting with the sub-scanning direction. The image processing device performs image processing on a plurality of read image data obtained by reading the same object with the at least one line sensor. The at least one line sensor includes a plurality of first image pickup elements inclined with respect to the longitudinal direction and arranged side by side in the longitudinal direction, and a plurality of second image pickup elements inclined at an angle different from that of the first image pickup elements with respect to the longitudinal direction and arranged side by side in the longitudinal direction.

With this configuration, the placement area of each line sensor in the sub-scanning direction can be narrowed, thereby saving space.

The at least one line sensor may include two line sensors, one of which may have the plurality of first imaging elements arranged in the longitudinal direction and inclined with respect to the longitudinal direction, and the other of which may have the plurality of second imaging elements arranged in the longitudinal direction and inclined at an angle different from that of the first imaging elements.

With this configuration, the arrangement area of each line sensor in the sub-scanning direction can be narrowed, thereby saving space, compared to a configuration in which multiple first imaging elements or multiple second imaging elements are arranged side by side along the longitudinal direction of each line sensor.

In one of the at least one line sensor, the multiple first imaging elements inclined with respect to the longitudinal direction are arranged side by side in the longitudinal direction, and the multiple second imaging elements inclined with respect to the longitudinal direction at an angle different from that of the first imaging elements are arranged side by side in the longitudinal direction.

With this configuration, compared to a configuration in which two line sensors are provided with multiple first imaging elements or multiple second imaging elements arranged side by side along the longitudinal direction, the two line sensors can be integrated and the arrangement area of the line sensors in the sub-scanning direction can be narrowed, thereby saving space.

The image reading device of the present invention comprises at least one line sensor and the image processing device. The at least one line sensor is capable of reading an image in a line shape along the main scanning direction. The image processing device performs image processing on multiple read image data obtained by reading the same object with the at least one line sensor. The line sensor has a light source and a slit member. The light source extends in the main scanning direction and irradiates light onto the object. The slit member extends in the main scanning direction and is disposed between the object and the light source.

With this configuration, it is possible to suppress blurring of the image contours caused by light circulating around the object in the sub-scanning direction, thereby generating output image data with clearer contours in the area corresponding to the object.

The image reading device of the present invention comprises at least one line sensor and the image processing device. The at least one line sensor is capable of reading an image in a line shape along the main scanning direction. The image processing device performs image processing on multiple read image data obtained by reading the same object with the at least one line sensor. The line sensor has a line laser light source. The line laser light source irradiates light onto the object.

With this configuration, by using a line laser light source as a line sensor, it is possible to suppress blurring of the image contours caused by light that wraps around the object in the sub-scanning direction, so that output image data can be generated with clearer contours in the area corresponding to the object without increasing the number of components.

The image reading device of the present invention comprises at least one line sensor, the image processing device, and a plurality of conveying devices. The at least one line sensor is capable of reading an image in a line shape along a main scanning direction. The image processing device performs image processing on a plurality of read image data obtained by reading the same object with the at least one line sensor. The plurality of conveying devices convey the object along a conveying path extending in a direction intersecting the main scanning direction. The line sensor has a light source. The light source irradiates light onto the object. The plurality of conveying devices are arranged along the conveying path with gaps between each other, and the light source irradiates light onto the object conveyed along the conveying path through the gap, or causes the light irradiated onto the object conveyed along the conveying path to enter the gap.

With this configuration, it is possible to prevent the light emitted from the light source from being absorbed or reflected before it reaches the object, thereby preventing the contours of the area corresponding to the object from becoming blurred in the read image data.

The image reading device of the present invention comprises at least one line sensor, the image processing device, and a conveying device. The at least one line sensor is capable of reading an image in a line shape along a main scanning direction. The image processing device performs image processing on multiple read image data obtained by reading the same object with the at least one line sensor. The conveying device conveys the object along a conveying path extending in a direction intersecting the main scanning direction. The line sensor has a light source. The light source irradiates light onto the object. The conveying device includes a conveying surface that forms the conveying path, and the conveying surface is optically transparent, so that light from the light source passes through the conveying surface and is irradiated onto the object.

With this configuration, it is possible to prevent the light emitted from the light source from being absorbed or reflected before it reaches the object, thereby preventing the contours of the area corresponding to the object from becoming blurred in the read image data.

The image processing method according to the present invention is an image processing method that performs image processing on a plurality of read image data obtained by reading the same object in lines along main scanning directions that extend in different directions, and the image processing method includes a spectrum generation step, a comparison step, a synthetic spectrum generation step, and an image generation step. The spectrum generation step generates a plurality of spectral data by performing Fourier transform processing on each of the plurality of read image data. The comparison step compares each component corresponding to the same frequency of the plurality of spectral data. The synthetic spectrum generation processing step generates synthetic spectrum data by selecting one of the components corresponding to the same frequency of the plurality of spectral data based on the comparison result of the comparison step. The image generation step generates output image data by performing inverse Fourier transform processing on the synthetic spectrum data.

With this configuration, an image processing method can be provided that is capable of generating output image data with clear contours of the area corresponding to the object.

The image processing program according to the present invention is an image processing program that performs image processing on multiple read image data obtained by reading the same object in lines along main scanning directions that extend in different directions, and the image processing program causes a computer to execute a spectrum generation step, a comparison step, a synthetic spectrum generation step, and an image generation step. The spectrum generation step generates multiple spectral data by performing Fourier transform processing on each of the multiple read image data. The comparison step compares each component corresponding to the same frequency of the multiple spectral data. The synthetic spectrum generation processing step generates synthetic spectral data by selecting one of the components corresponding to the same frequency of the multiple spectral data based on the comparison result of the comparison step. The image generation step generates output image data by performing inverse Fourier transform processing on the synthetic spectral data.

With this configuration, it is possible to provide an image processing program capable of generating output image data with clear contours of the area corresponding to the object.

According to the present invention, by applying image processing to multiple scanned image data obtained by scanning an object in a line, it is possible to generate output image data with clear contours of the area corresponding to the object.

2 is a block diagram showing an example of an electrical configuration of the image reading apparatus according to the first embodiment; FIG. FIG. 2 is a cross-sectional view illustrating an example of an image reading unit according to the first embodiment. FIG. 2 is an exploded perspective view illustrating an example of the appearance of a line light source in the image reading unit according to the first embodiment. FIG. 4 is a diagram showing an example of a read image according to the first embodiment; FIG. 11 is a diagram showing another example of the read image in the first embodiment. FIG. 4 is a diagram illustrating an example of an output image according to the first embodiment. FIG. 4 is a diagram illustrating an example of a memory map of a RAM in the first embodiment. 1 is a block diagram showing an example of a functional configuration of an image reading apparatus according to a first embodiment; 4 is a flowchart showing an example of image processing by a CPU of the image reading apparatus according to the first embodiment; FIG. 11 is a schematic diagram illustrating an example of the periphery of a plurality of image reading units according to a second embodiment. FIG. 11 is a diagram for explaining an example of a correction process according to a second embodiment. FIG. 11 is a diagram illustrating an example of a memory map of a RAM according to the second embodiment. FIG. 11 is a block diagram showing an example of a functional configuration of an image reading apparatus according to a second embodiment. FIG. 11 is a flowchart showing an example of image processing by a CPU of the image reading apparatus according to the second embodiment. FIG. 11 is a schematic diagram for explaining the arrangement of an image reading unit according to a second embodiment. FIG. 11 is a schematic diagram for explaining the arrangement of an image reading unit according to a second embodiment. FIG. 11 is a schematic diagram for explaining the arrangement of an image reading unit according to a second embodiment. FIG. 11 is a schematic diagram for explaining the arrangement of an image reading unit according to a second embodiment. FIG. 11 is a schematic diagram for explaining the arrangement of an image reading unit according to a second embodiment. FIG. 13 is a schematic diagram for explaining the arrangement of an image reading unit according to a third embodiment. FIG. 13 is a schematic diagram for explaining a configuration of an image reading unit according to a third embodiment. 13 is a schematic diagram for explaining a modified example of the configuration of the image reading unit of the third embodiment. FIG. FIG. 13 is a cross-sectional view illustrating a portion of an example of an image reading unit according to a fourth embodiment. FIG. 13 is a schematic diagram for explaining a configuration around an image reading unit according to a fifth embodiment. 13 is a schematic diagram for explaining a modified example of the configuration around the image reading unit of the fifth embodiment. FIG.

1. First embodiment Fig. 1 is a block diagram showing an example of an electrical configuration of an image reading device 10 according to a first embodiment. As shown in Fig. 1, the image reading device 10 includes a control unit 20 and an image reading unit 28, which are electrically connected to each other via a bus 30.

The control unit 20 also includes a CPU (Central Processing Unit) 22, a RAM (Random Access Memory) 24, and a memory unit 26.

The CPU 22 is responsible for the overall control of the image reading device 10. The RAM 24 is used as a work area and buffer area for the CPU 22.

The memory unit 26 is the main memory of the image reading device 10, and is implemented by a hard disk drive (HDD) and a non-volatile memory such as an electrically erasable programmable read only memory (EEPROM). The memory unit 26 may also be configured to include a RAM 24.

The memory unit 26 stores data regarding the control program used by the CPU 22 to control the operation of each component of the image reading device 10, data regarding various images, and execution data required to execute the control program.

The image reading unit 28 includes at least an image sensor, and reads an object in a line shape along the main scanning direction while scanning in the sub-scanning direction. The image sensor is a sensor for reading an image by converting light into an electrical output through photoelectric conversion.

There is no particular limit to the type or number of image sensors, as long as they are capable of reading an object in a line along the main scanning direction while scanning in the sub-scanning direction.

As the image reading unit 28, for example, a general-purpose line sensor, a general-purpose camera, or a CIS (Contact Image Sensor) can be used. In the first embodiment, a case where a CIS is used as the image reading unit 28 will be described as an example.

When a general-purpose camera such as an area camera, i.e., a camera equipped with an area sensor, is used as the image reading unit 28, only one row of the area sensor is used. The viewing angle of the camera is preferably narrow. Furthermore, it is preferable that the emission width and emission angle of the light source are also narrow. Furthermore, it is preferable that the emission angle of the light source is close to 0.

Note that the electrical configuration of the image reading device 10 in the first embodiment is an example. For example, the image reading device 10 may be provided with multiple image reading units 28.

Figure 2 is a cross-sectional view showing an example of the image reading unit 28 of the first embodiment. Figure 3 is an exploded perspective view showing a schematic example of the appearance of the line light source 44 (44A, 44B) in the image reading unit 28 of the first embodiment. As shown in Figure 2, the image reading unit 28 includes two housings 42 (42A, 42B) arranged opposite each other across the focal plane 40, and various components provided in the housings 42.

Each housing 42 includes a line light source 44 (44A, 44B) for illuminating an object on the focal plane 40. The line light source 44 is a component that emits light toward a reading target (inspection target) T on the focal plane 40. In FIG. 2, the light emitted from the line light source 44A toward the focal plane 40 is indicated by L1, and the light emitted from the line light source 44B toward the focal plane 40 is indicated by L2.

3, the line light source 44 includes a transparent light guide 46 extending along the longitudinal direction D, a first light source unit 48 provided on one end surface in the longitudinal direction D, a second light source unit 50 provided on the other end surface in the longitudinal direction D, and a cover member 51 for holding each outer side surface of the light guide 46. The line light source 44 is not limited to a guide type in which light is guided by the light guide 46. In addition, in FIG. 3, the first light source unit 48 and the second light source unit 50 are illustrated spaced apart from the light guide 46.

Furthermore, the linear light source 44 has a light diffusion pattern P. The light diffusion pattern P is formed on the bottom surface 52 of the light guide 46 so as to extend along the longitudinal direction D. When light is incident on the light guide 46 from the first light source unit 48 and the second light source unit 50, the light diffusion pattern P diffuses and refracts the light traveling through the light guide 46, causing it to exit from the light exit side surface 54 of the light guide 46.

In addition, the light exit side surface 54 is formed into a smooth convex curve facing outward to provide the light-gathering effect of a lens.

Also, as shown in FIG. 2, a substrate 56 for fixing the line-shaped light source 44A is provided in the housing 42A, and a substrate 58B for fixing the line-shaped light source 44B is provided in the housing 42B.

3, the first light source unit 48 and the second light source unit 50 are provided with terminals 60. The line light source 44A is fixed to the board 56 by inserting the terminal 60 into the board 56 and soldering it. The line light source 44B is fixed to the board 58B by inserting the terminal 60 into the board 58B and soldering it.

2, a lens array 62 is provided within the housing 42A. The lens array 62 is an optical element that forms an image of light reflected by or transmitted through the object to be read T on a light receiving unit 64 (described later), and may be a rod lens array such as a SELFOC lens array (registered trademark: manufactured by Nippon Sheet Glass Co., Ltd.).

Furthermore, an ultraviolet light blocking filter (UV cut filter) 67 that blocks ultraviolet light from entering the light receiving unit 64 is provided at any position between the focal plane 40 and the light receiving unit 64. In addition, a color filter 68 that passes visible light in a specific wavelength range is provided between the light receiving unit 64 and the ultraviolet light blocking filter 67.

The light receiving unit 64 includes an image sensor and is mounted on a substrate 58A inside the housing 42A. Each housing 42 is provided with a protective glass 66 (66A, 66B) to protect the linear light source 44 from dust and scratches during use.

In the image reading unit 28 configured in this manner, the light L1 emitted from the linear light source 44A becomes a linear illumination light and illuminates the reading target T, and the reflected light from the reading target T is guided to the light receiving unit 64. The light L2 emitted from the linear light source 44B becomes a linear illumination light and passes through the reading target T, and the transmitted light that passes through the reading target T is guided to the light receiving unit 64. An example of the reading target T is food, but is not limited to this. For example, the reading target T can be food (in addition to solids, liquids, etc., any item can be used even if it has a characteristic density (bread, oil, etc.), color (ketchup, curry, etc.) or shape (noodles, etc.), resin molded products such as films or sheets, textile products such as fabrics, clothes, or nonwoven fabrics, and paper products such as sheets, rolls, or structures (envelopes or boxes). The object to be read may be the read target T itself, or may be a part that is contained in, mixed with, or covered by the read target T and is read at the same time. Examples of such parts include metal pieces, resin pieces (plastic or rubber, etc.), mineral pieces (stone or sand, etc.), animal or plant pieces (branches, leaves, bones, insects, etc.), and composites thereof, as well as structural defects (scratches, irregularities, air bubbles, attached foreign matter, etc.) inside the read target T.

Therefore, the image reading unit 28 can read the object T in a line along the main scanning direction (Y direction in FIG. 2) while scanning in the sub-scanning direction (X direction in FIG. 2). However, the linear light source 44A may be omitted so that only the transmitted light from the object T is received by the light receiving unit 64.

Although not shown in the figure, a structure for electrically connecting to the control unit 20, such as a connector, is provided on the substrate 58 (58A, 58B).

In the first embodiment, the image reading unit 28 is used to read the same object in lines along main scanning directions that extend in different directions, thereby generating multiple read image data. For example, on the focal plane 40 of the fixed image reading unit 28, the object can be read along the main scanning directions that extend in different directions for the same object by transporting the object in multiple different directions that intersect with the Y direction and reading the object along the main scanning direction during each transport. In addition, the image reading unit 28 may be moved relative to an object that is stopped on the focal plane 40.

In this specification, image data refers to an image in a data format. Therefore, in this specification, for example, read image data may be described as a read image, and a read image may be described as read image data. This also applies to spectra, which will be described later.

In the first embodiment, when multiple pieces of read image data are generated, a Fourier transform process is performed on each of the multiple pieces of read image data. In addition, when a Fourier transform process is performed on each of the multiple pieces of read image data, multiple pieces of spectral data are generated.

Specifically, in the Fourier transform process, a two-dimensional Fourier transform is performed on the scanned image. Furthermore, the spectral data is a frequency spectrum in a data format. A frequency spectrum is a spectrum having components corresponding to each frequency. Furthermore, specifically, the frequency spectrum is a spatial frequency spectrum.

In the first embodiment, the frequency spectrum corresponds to any one of an amplitude spectrum, a phase spectrum, and a power spectrum. For example, if the frequency spectrum corresponds to an amplitude spectrum, the component corresponding to the frequency indicates the magnitude of the amplitude. It is preferable to use an amplitude spectrum as the frequency spectrum.

In the first embodiment, when multiple pieces of spectral data are generated, the pieces of spectral data are compared with each other. Specifically, components corresponding to the same frequency in the multiple pieces of spectral data are compared with each other.

When the components corresponding to the same frequency are compared, one of the components corresponding to the same frequency in the multiple spectral data is selected as the component to be combined based on the comparison result. In the first embodiment, the largest component among the components corresponding to the same frequency in the multiple spectral data is selected as the component to be combined.

When synthesis components are selected as components corresponding to each frequency, synthetic spectrum data is generated based on those synthesis components. The synthetic spectrum data is a synthetic spectrum in data format. The synthetic spectrum is a spectrum in which the components corresponding to each frequency are used as synthesis components.

In the first embodiment, when the synthetic spectral data is generated, an inverse Fourier transform process is performed on the synthetic spectral data. When the synthetic spectral data is subjected to an inverse Fourier transform process, output image data is generated. Note that in the inverse Fourier transform process, a two-dimensional inverse Fourier transform is performed on the synthetic spectral data.

Figure 4 is a diagram showing an example of a read image in the first embodiment. Figure 5 is a diagram showing another example of a read image in the first embodiment. The read images shown in Figures 4 and 5 are images generated by reading the same read object T in a line. Furthermore, when the read images shown in Figures 4 and 5 are generated, the main scanning direction and the sub-scanning direction of the image reading unit 28 are orthogonal. Furthermore, the sub-scanning direction of the image reading unit 28 when the read image shown in Figure 4 is generated is orthogonal to the sub-scanning direction of the image reading unit 28 when the read image shown in Figure 5 is generated. Specifically, in Figure 4, the up-down direction is the main scanning direction and the left-right direction is the sub-scanning direction, and in Figure 5, the left-right direction is the main scanning direction and the up-down direction is the sub-scanning direction.

When the scanned images shown in Figures 4 and 5 are generated, the environment in which the object is scanned, specifically the main scanning direction, is different, so that in these scanned images, the blurred parts of the contour of the object area, which is the area on the image that corresponds to the object, are different. In other words, the clear parts of the contour of the object area are different.

The target area is specifically the area of pixels that make up the object on the image. In other words, the target area can also be said to be the object on the image.

Figure 6 is a diagram showing an example of an output image in the first embodiment. The output image in Figure 6 is an image corresponding to the read images shown in Figures 4 and 5. In the output image, the target area is displayed as if it were a combination of the clear parts of the target area in each read image. In other words, the contour of the target area is displayed clearly in the output image.

Note that Figure 6 shows an example of an output image corresponding to two scanned images, but the more scanned images there are, the clearer the entire contour of the target area will be displayed in the output image.

In the first embodiment, for example, when g _θ is the scanned image data with a blurred contour of the target area, g _θ is expressed by the following formula (1). f is the scanned image data with a clear contour of the target area. h _θ is a PSF (Point Spread Function). In other words, h _θ is a factor caused by blurring of the contour of the target area. In addition, the PSF depends on the angle of the main scanning direction.

When the Fourier transform process is applied to both sides of the formula (1) as shown in the formula (2), the formula (3) is derived. Therefore, in the formula (3), G _θ , F, and H _θ are g _θ , f, and h _θ that have been subjected to the Fourier transform process. That is, G _θ is spectrum data.

As described above, when the synthetic spectral data is generated, the largest component is selected from among the components corresponding to the same frequency of the multiple spectral data. Therefore, the synthetic spectral data is generated according to formula (4). That is, in formula (4), the left side is the synthetic spectral data, and the right side is each synthesis component.

Since the synthetic spectral data is subjected to an inverse Fourier transform process, the inverse Fourier transform process is applied to equation (4) as shown in equation (5). That is, in equation (5), the left side is the output image data.

Figure 7 is a diagram showing an example of a memory map 200 of the RAM 24 in the first embodiment. As shown in Figure 7, the RAM 24 includes a program area 201 and a data area 202, and the program area 201 stores a control program that has been previously read out from the memory unit 26.

The control programs include an image reading program 201a, a spectrum generation program 201b, a comparison program 201c, a synthetic spectrum generation program 201d, and an image generation program 201e, etc.

The image reading program 201a is a program for controlling the image reading unit 28 to generate read image data 202a.

The spectrum generation program 201b is a program for generating spectrum data 202b by performing Fourier transform processing on the read image data 202a generated by the image reading program 201a.

The comparison program 201c is a program for comparing components corresponding to the same frequency in multiple spectral data 202b generated by the spectral generation program 201b.

The synthetic spectrum generation program 201d is a program for selecting the largest component from among the components corresponding to the same frequency in the multiple spectral data 202b based on the comparison results by the comparison program 201c, and generating synthetic spectrum data 202c.

The image generation program 201e is a program for generating output image data 202d by performing inverse Fourier transform processing on the synthetic spectral data 202c.

Although not shown in the figure, the program area 201 also stores control programs other than the image reading program 201a and the spectrum generation program 201b.

The data area 202 stores data that has been previously read out from the memory unit 26. In the example shown in FIG. 7, the data area 202 stores read image data 202a, spectral data 202b, composite spectral data 202c, output image data 202d, etc.

The read image data 202a is data corresponding to a read image. In addition, multiple read image data 202a may be stored in the data area 202.

The spectral data 202b is data corresponding to a frequency spectrum. In addition, multiple pieces of spectral data 202b may be stored in the data area 202.

The synthetic spectrum data 202c is data corresponding to the synthetic spectrum. The output image data 202d is data corresponding to the output image.

In addition, the data area 202 may store, for example, execution data and may include timers (counters) and registers necessary for executing the control program.

8 is a block diagram showing an example of the functional configuration of the image reading device 10 of the first embodiment. In the control unit 20 including the CPU 22, the CPU 22 executes the image reading program 201a, whereby the control unit 20 functions as an image reading processing unit 90 that controls the image reading unit 28 to generate read image data 202a.

In addition, when the CPU 22 executes the spectrum generation program 201b, the control unit 20 functions as a spectrum generation processing unit 92 that performs Fourier transform processing on the read image data 202a and generates spectrum data 202b.

Furthermore, when the CPU 22 executes the comparison program 201c, the control unit 20 functions as a comparison processing unit 94 that compares each component corresponding to the same frequency in multiple spectral data 202b.

Furthermore, when the CPU 22 executes the synthetic spectrum generation program 201d, the control unit 20 functions as a synthetic spectrum generation processing unit 96 that selects the largest component from among the components corresponding to the same frequency in the multiple spectral data 202b based on the comparison result of the comparison processing unit 94, and generates synthetic spectral data.

In addition, when the CPU 22 executes the image generation program 201e, the control unit 20 functions as an image generation processing unit 98 that performs inverse Fourier transform processing on the synthetic spectral data 202c and generates output image data 202d.

In the example shown in FIG. 8, for example, when the image reading processing unit 90 generates a plurality of read image data 202a, the spectrum generation processing unit 92 generates a plurality of spectrum data 202b. Furthermore, when the plurality of spectrum data 202b are generated, the comparison processing unit 94 compares each component corresponding to the same frequency in the plurality of spectrum data 202b, and the synthetic spectrum generation processing unit 96 selects the largest component from among the components corresponding to the same frequency in the plurality of spectrum data 202b based on the comparison result of the comparison processing unit 94, and generates synthetic spectrum data 202c. Furthermore, when the synthetic spectrum data 202c is generated, the image generation processing unit 98 generates output image data 202d.

Figure 9 is a flow diagram showing an example of image processing by the CPU 22 of the image reading device 10 of the first embodiment. The CPU 22 starts image processing, for example, when multiple read image data 202a are generated.

In step S1, Fourier transform processing is performed on each scanned image data 202a to generate multiple spectral data 202b.

In step S2, each component corresponding to the same frequency in the multiple spectral data 202b is compared.

In step S3, based on the comparison results of each component corresponding to the same frequency, the largest component is selected from among the components corresponding to the same frequency in the multiple spectral data 202b, and synthetic spectral data 202c is generated.

In step S4, an inverse Fourier transform process is performed on the synthetic spectrum data 202c to generate output image data 202d. Also, when the output image data 202d is generated in step S4, the image processing ends.

According to the first embodiment, output image data with clear contours of the target area can be generated.

Furthermore, according to the first embodiment, it is believed that the greater the number of read image data, the clearer the contours of the target area will be in the output image data that is generated.

In addition, the configuration is not limited to one in which the largest component is selected as the component to be combined from among the components corresponding to the same frequency of multiple spectral data, and the component to be combined may be selected as appropriate.

For example, among the multiple spectral data, for components corresponding to some frequencies, the largest component may be selected as the component to be synthesized, and for components corresponding to other frequencies, an appropriate component to be synthesized may be selected.

Furthermore, when a component to be combined is selected from each component corresponding to the same frequency of multiple spectral data, any component between the smallest component and the largest component may be selected as the component to be combined.

2. Second embodiment In the second embodiment, a target object is read using a plurality of image reading units 28. In the second embodiment, before the Fourier transform process is applied to the plurality of read image data, a process is applied to position the target area in approximately the same position in each read image data. Note that the series of processes after the Fourier transform process are the same as those in the first embodiment, and therefore a duplicated description will be omitted.

In the second embodiment, the main scanning directions of the multiple image reading units 28 are arranged so that they intersect with each other. Also, each image reading unit 28 is arranged so that its main scanning direction intersects with the sub-scanning direction at a predetermined angle. However, the sub-scanning direction of each image reading unit 28 in the second embodiment (the transport direction of the object, or the movement direction of each image reading unit 28) is the same direction and intersects with all of the image reading units 28.

Figure 10 is a schematic diagram showing an example of the periphery of the multiple image reading units 28 in the second embodiment. Note that Figure 10 is also a schematic diagram showing an example of the periphery of the multiple image reading units 28 when viewed from a direction perpendicular to the focal plane 40.

In the example shown in Figure 10, the main scanning directions of image reading unit 28A and image reading unit 28B are perpendicular to each other, and image reading unit 28A and image reading unit 28B are arranged so that the main scanning direction intersects with the sub-scanning direction at an angle of 45°. In other words, in the example shown in Figure 10, the main scanning direction and sub-scanning direction of each image reading unit 28 are not perpendicular to each other.

In addition, the sub-scanning direction of each image reading unit 28 is the same direction, and the sub-scanning direction of each image reading unit 28 intersects with image reading unit 28A and image reading unit 28B. However, the angle between the sub-scanning direction and the main scanning direction of image reading unit 28 is not limited to 45°. In addition, the angle between the main scanning directions of image reading units 28 is not limited to 90°.

In addition, in the second embodiment, the multiple image reading units 28 are arranged so that their main scanning directions extend in different directions relative to the sub-scanning direction and their main scanning directions intersect at one point.

In the second embodiment, read image data is generated based on an image reading unit 28 in which the main scanning direction and the sub-scanning direction are not perpendicular to each other, and thus read image data in which the target area is tilted is generated.

In the second embodiment, a correction process is performed on multiple read image data before Fourier transform processing is performed, so that the target area is positioned approximately identically on the image in each of the multiple read image data.

The correction process includes a tilt correction process for correcting the tilt of the target area and a magnification correction process for correcting the magnification of the target area, and is performed, for example, in the order of tilt correction process and magnification correction process. The correction process will be described with reference to FIG. 11. FIG. 11 is a diagram for explaining an example of the correction process of the second embodiment. FIG. 11 also shows a read image corresponding to image reading unit 28A and a read image corresponding to image reading unit 28B. In the correction process, first, the read image is scanned and pixel coordinates are obtained.

The target area is tilted at an angle that corresponds to the relative speed between the image reading unit 28 and the target object when reading the target object. The faster the relative speed, the greater the tilt of the target area.

In the tilt correction process, the tilt of the target area is calculated based on the pixel coordinates, and the pixels are shifted vertically according to that tilt. In this way, in the tilt correction process, the tilt of the target area is corrected as the pixels are shifted. Note that a well-known method is used to calculate the tilt of the target area from the pixel coordinates.

In the magnification correction process, the target area is reduced in the left-right direction. The correction magnification, which is the magnification by which the target area is reduced, is not particularly limited as long as the target area is positioned approximately the same on each read image. For example, the correction magnification may be set based on the inclination of the target area before inclination correction. Note that pixel interpolation and the like associated with adjusting the magnification of the target area are well known, so a detailed explanation will be omitted. However, in the magnification correction process, the target area may be enlarged rather than reduced.

In addition, since the target area can be positioned approximately identically in each scanned image data by just the tilt correction process, the magnification correction process does not need to be performed. In other words, in the correction process, at least the tilt correction process may be performed out of the tilt correction process and the magnification correction process.

In addition, the correction process may use a known method to correct at least the tilt of the target area. For example, the correction process may use a known method to correct the tilt and magnification of an object, such as a character, figure, or illustration, based on the positions (coordinates) of pixels that make up the object on the image.

Furthermore, in the second embodiment, the main scanning directions of image reading unit 28A and image reading unit 28B are arranged perpendicular to each other, but the number of image reading units 28 may be three or more as long as the main scanning directions of the multiple image reading units 28 extend in different directions relative to the sub-scanning direction and the main scanning directions intersect at one point. Also, depending on the angle and number of image reading units 28 when they are arranged, the number of read image data to be subjected to correction processing may be at least one.

Figure 12 is a diagram showing an example of a memory map 200 of RAM 24 in the second embodiment. In the second embodiment, the control program includes a correction program 201f. The correction program 201f also includes a tilt correction program 201g and a magnification correction program 201h.

The correction program 201f is a program for performing correction processing on the read image data 202a generated by the image reading program 201a and generating corrected image data 202e.

The tilt correction program 202g is a program for performing tilt correction processing on the read image data 202a generated by the image reading program 201a. The magnification correction program 201h is a program for performing magnification correction processing on the read image data 202a generated by the image reading program 201a.

In the second embodiment, correction image data 202e is stored in the data area 202. The correction image data 202e is data corresponding to a correction image, which is a read image that has been subjected to a correction process. Note that the correction image data 202e may include multiple pieces of data corresponding to the correction image.

Furthermore, in the second embodiment, the spectrum generation program 201b is a program for generating spectrum data 202b by performing Fourier transform processing on the read image data 202a generated by the image reading program 201a, specifically, the corrected image data 202e.

Figure 13 is a block diagram showing an example of the functional configuration of the image reading device 10 of the second embodiment. In the second embodiment, the CPU 22 executes the correction program 201f to function as a correction processing unit 100 that performs correction processing on the read image data 202a and generates corrected image data 202e. Also, in the second embodiment, the CPU 22 executes the spectrum generation program 201b to function as a spectrum generation processing unit 92 that performs Fourier transform processing on the corrected image data 202e and generates spectrum data 202b.

Figure 14 is a flow diagram showing an example of image processing by the CPU 22 of the image reading device 10 of the second embodiment. In the second embodiment, when multiple read image data 202a are generated, multiple corrected image data 202e are generated in step S1. Note that steps S2 to S5 correspond to steps S1 to S4 in the flow diagram shown in Figure 9, and therefore repeated explanations will be omitted.

In the second embodiment, as described above, the number of image reading units 28 may be three or more, and depending on the angle and number of image reading units 28 when they are arranged, the number of read image data to which correction processing is applied may be at least one. Therefore, according to the second embodiment, even if the target area is tilted in at least some of the multiple read image data, the target area can be positioned approximately the same in each read image data. In other words, the image reading units 28 can be arranged so that the main scanning direction and the sub-scanning direction are not perpendicular to each other.

Furthermore, according to the second embodiment, multiple read image data can be generated simply by moving the object or multiple image reading units 28 in the sub-scanning direction, that is, with just one operation. Therefore, according to the second embodiment, multiple read image data can be obtained efficiently. However, in this case, the main scanning direction and the sub-scanning direction of each image reading unit 28 are not orthogonal.

In the second embodiment, the multiple image reading units 28 may be arranged in a direction intersecting the sub-scanning direction, as shown in Figures 15 to 19. Specifically, the main scanning direction extends in a different direction from the sub-scanning direction, and a group of image reading units 28 arranged in the sub-scanning direction may be arranged in a direction perpendicular to the sub-scanning direction.

Figures 15 to 19 are schematic diagrams for explaining the arrangement of image reading unit 28 in the second embodiment. In the example shown in Figure 15, image reading unit 28A and image reading unit 28B are arranged so that their main scanning directions are perpendicular to each other and intersect with the sub-scanning direction at an angle of 45°. Furthermore, image reading unit 28A and image reading unit 28B are arranged symmetrically with respect to an axis (axis of symmetry) that is a direction perpendicular to the sub-scanning direction. Furthermore, image reading unit 28C and image reading unit 28D are arranged in the same manner as image reading unit 28A and image reading unit 28B.

In addition, the axis of symmetry for image reading unit 28A and image reading unit 28B is the same as the axis of symmetry for image reading unit 28C and image reading unit 28D. The group including image reading unit 28C and image reading unit 28D is arranged side by side with the group including image reading unit 28A and image reading unit 28B in a direction perpendicular to the sub-scanning direction.

In the example shown in Fig. 15, image reading units 28A and 28C are arranged so that their main scanning directions are parallel to each other, and further arranged so that they partially overlap each other in the sub-scanning direction. This also applies to image reading units 28B and 28D.

In the example shown in Figure 16, unlike the example shown in Figure 15, image reading units 28A and 28C are arranged so that their main scanning directions intersect with each other, and are further arranged so that they are symmetrical with respect to the sub-scanning direction. In other words, image reading units 28A and 28C do not partially overlap with each other in the sub-scanning direction. The same is true for image reading units 28B and 28D.

In the example shown in Figure 17, unlike the example shown in Figure 16, the axis of symmetry for image reading units 28A and 28B is different from the axis of symmetry for image reading units 28C and 28D. In other words, image reading units 28A and 28C are asymmetric, and image reading units 28B and 28D are also asymmetric. Furthermore, image reading units 28A and 28C partially overlap each other in the sub-scanning direction. The same is true for image reading units 28B and 28D.

In the example shown in Figure 18, unlike the example shown in Figure 16, image reading units 28A and 28C are asymmetric and partially overlap each other in the sub-scanning direction. The same is true for image reading units 28B and 28D.

In the example shown in FIG. 19, unlike the example shown in FIG. 16, an image reading unit 28E is provided in which the main scanning direction is perpendicular to the sub-scanning direction. The image reading unit 28E is provided along the axis of symmetry related to the image reading unit 28A and the image reading unit 28B. In the example shown in FIG. 19, the image reading unit 28A and the image reading unit 28B are arranged so that the main scanning direction intersects with the sub-scanning direction at an angle of 60°. Similarly, the image reading unit 28C and the image reading unit 28D are arranged so that the main scanning direction intersects with the sub-scanning direction at an angle of 60°. As a result, the image reading unit 28A, the image reading unit 28C, and the image reading unit 28E are arranged in an equilateral triangle, and the image reading unit 28B, the image reading unit 28D, and the image reading unit 28E are also arranged in an equilateral triangle.

As in the examples of Figures 15 to 19, when a group of image reading units 28 are arranged in a direction intersecting the sub-scanning direction, even if the length of each image reading unit 28 in the main scanning direction is short, the object to be read T can be divided and read by multiple image reading units 28 arranged in a direction intersecting the sub-scanning direction.

This narrows the area in the sub-scanning direction where the image reading units 28 are placed, thereby saving space. In particular, in the examples of Figures 15 to 19, each image reading unit 28 has the same length, and it is only necessary to prepare and appropriately place multiple identical image reading units 28, thereby realizing space saving at low cost. Note that when a group of image reading units 28 is placed side by side in a direction intersecting the sub-scanning direction in this way, the corrected read images may be synthesized as necessary.

In the first and second embodiments, the image pickup elements 32 (32A, 32B) (see Figures 21 and 22) described later are arranged side by side in the longitudinal direction of the image reading unit 28 without being inclined relative to the longitudinal direction of the image reading unit 28, so that the main scanning direction of the image reading unit 28 coincides with the longitudinal direction of the image reading unit 28.

3. Third embodiment The third embodiment is similar to the first and second embodiments, except that the configuration and arrangement of the image reading unit 28 are changed. Fig. 20 is a schematic diagram for explaining the arrangement of the image reading unit 28 of the third embodiment. Fig. 21 is a schematic diagram for explaining the configuration of the image reading unit 28 of the third embodiment. Note that Fig. 21 is also an enlarged view of the periphery of the image reading unit 28 in Fig. 20.

In the third embodiment, two image reading units 28 are used. Each of the image reading units 28 has an elongated shape along a longitudinal direction that intersects with the sub-scanning direction. In the example shown in FIG. 21, each of the image reading units 28 is arranged so that the longitudinal direction is perpendicular to the sub-scanning direction.

In the third embodiment, the first image sensor 32A, which is inclined with respect to the longitudinal direction of the image sensor 28A, is arranged in the longitudinal direction of the image sensor 28A in one of the two image sensor units 28. The second image sensor 32B, which is inclined at an angle different from the first image sensor 32A with respect to the longitudinal direction of the image sensor 28B, is arranged in the longitudinal direction of the image sensor 28B in the other image sensor unit 28B. The first image sensor 32A and the second image sensor 32B are each a light receiving IC chip, and are configured by arranging a plurality of photoelectric conversion elements such as photodiodes in a straight line.

21, the first imaging element 32A is inclined at an angle of +45° with respect to the longitudinal direction of the image reading unit 28A, and the second imaging element 32B is inclined at an angle of -45° with respect to the longitudinal direction of the image reading unit 28B. The first imaging element 32A and the second imaging element 32B are also arranged symmetrically with respect to an axis perpendicular to the sub-scanning direction. In other words, in the example shown in FIG. 21, the main scanning directions of each image reading unit 28 are perpendicular to each other.

In this way, when the first image sensor 32A and the second image sensor 32B are arranged, even if the two image sensors 28 are arranged in parallel, two read image data can be obtained by the two image sensors 28 having different main scanning directions. Therefore, compared to a configuration in which multiple first image sensors 32A or multiple second image sensors 32B are arranged side by side along the longitudinal direction of each image sensor 28 as shown in FIG. 10, the arrangement area of each image sensor 28 in the sub-scanning direction can be narrowed, thereby saving space.

Figure 22 is a schematic diagram for explaining a modified configuration of the image reading unit 28 of the third embodiment. In this modified example, only one image reading unit 28 is used. In this case, on the same substrate in one image reading unit 28, a plurality of first image pickup elements 32A are arranged in parallel in the longitudinal direction of the image reading unit 28, and a plurality of second image pickup elements 32B are arranged in parallel in the longitudinal direction of the image reading unit 28.

22, the first imaging element 32A is inclined at an angle of +45° with respect to the longitudinal direction of the image reading unit 28A, and the second imaging element 32B is inclined at an angle of -45° with respect to the longitudinal direction of the image reading unit 28B. Furthermore, the first imaging element 32A and the second imaging element 32B are arranged symmetrically with respect to an axis perpendicular to the sub-scanning direction. Therefore, in the example shown in FIG. 22, the image reading unit 28 has two main scanning directions, and the main scanning directions are perpendicular to each other.

In this way, even when a first image capturing element 32A and a second image capturing element 32B are arranged in one image reading unit 28, the arrangement area of the image reading unit 28 in the sub-scanning direction can be narrowed, thereby saving space, compared to a configuration in which multiple first image capturing elements 32A or multiple second image capturing elements 32B are arranged side by side along the longitudinal direction of each image reading unit 28 as shown in Figure 10.

4. Fourth Embodiment The fourth embodiment is similar to the first and second embodiments except for a change in the configuration of the image reading unit 28. That is, in the fourth embodiment, the longitudinal direction of the image reading unit 28 coincides with the main scanning direction.

Figure 23 is a cross-sectional view showing an example of a portion of the image reading unit 28 of the fourth embodiment. Also, Figure 23 is a cross-sectional view of the image reading unit 28 when viewed from the main scanning direction. As shown in Figure 23, the image reading unit 28 of the fourth embodiment has a slit member 34. The slit member 34 extends in the main scanning direction and is disposed between the reading object T and the linear light source 44B. The slit member 34 has a long and thin slit formed along the main scanning direction, and only light that passes through the slit is irradiated onto the reading object T. The width of the slit perpendicular to the main scanning direction may be approximately equal to or less than the width of each photoelectric conversion element of the imaging element.

The linear light source 44B may be composed of an LED or a halogen lamp, and may emit light in the near-infrared region of 750 nm to 2500 nm. In this case, the slit member 34 is preferably formed of a material and shape that is resistant to high-output near-infrared light and has excellent dimensional stability. The slit member 34 may also be movable in a direction to move closer to or away from the reading object T.

The slit member 34 blocks a portion of the light L2 irradiated to the reading target T in a direction perpendicular to the main scanning direction, thereby suppressing blurring of the image contours caused by light that wraps around the reading target T in the sub-scanning direction. This makes it possible to generate output image data with clearer contours in the area corresponding to the reading target T.

Although not shown in the figure, a general-purpose line laser light source may be used as the line-shaped light source 44B. The line laser light source includes a laser light source and a lens, and irradiates laser light in a line shape. The width perpendicular to the main scanning direction of the line-shaped laser light irradiated from the line laser light source may be approximately equal to or less than the width of each photoelectric conversion element of the imaging element. The line laser light source may emit light in the near-infrared range of 750 nm to 2500 nm. An example of a line laser light source is a telecentric laser light source.

When a line laser light source is used as the line light source 44B, the radial spread of light is suppressed in a direction perpendicular to the main scanning direction, so that the same effect as when the slit member 34 is provided can be obtained. In other words, by using a line laser light source as the line light source 44B, blurring of the contours on the image caused by light that wraps around the reading target T in the sub-scanning direction can be suppressed, so that output image data with clearer contours of the area corresponding to the reading target T can be generated without increasing the number of components.

5. Fifth embodiment In the fifth embodiment, a specific configuration of the conveying device 36 will be described. Note that, although the image reading device 10 includes a conveying device in the first, second, third, and fourth embodiments, a specific description thereof will be omitted.

The conveying device 36 conveys the object T to be read along a conveying path extending in a direction intersecting the main scanning direction of the image reading unit 28. When the fifth embodiment is combined with the fourth embodiment, the conveying device 36 conveys the object T to be read along a conveying path extending in a direction intersecting the longitudinal direction of the image reading unit 28.

Figure 24 is a schematic diagram for explaining the configuration around the image reading unit 28 of the fifth embodiment. In the example shown in Figure 24, multiple conveying devices 36 are provided, and these conveying devices 36 are arranged with gaps between each other along the conveying path. Furthermore, the multiple conveying devices 36 convey the reading target T along the conveying path extending in a direction perpendicular to the main scanning direction. A belt conveyor is used as the conveying device 36, but is not limited to this.

Furthermore, the linear light source 44B of the image reading unit 28 irradiates light to the object T to be read transported along the transport path through the gap between adjacent transport devices 36. However, in a configuration in which the object T to be read is transported closer to the linear light source 44B than the transport device 36, the light irradiated to the object T to be read transported along the transport path may enter the gap between adjacent transport devices 36.

In this way, when the linear light source 44B is disposed between the transport devices 36, the light irradiated from the linear light source 44B can be prevented from being absorbed or reflected before it reaches the reading target T, and therefore, the outline of the area corresponding to the reading target T in the read image data can be prevented from becoming blurred. In the example of FIG. 24, two transport devices 36 are provided, but three or more transport devices 36 may be provided and a linear light source 44B may be disposed between each of the adjacent transport devices 36.

Figure 25 is a schematic diagram for explaining a modified example of the configuration around the image reading unit 28 of the fifth embodiment. In this example, light irradiated from the linear light source 44B of the image reading unit 28 passes through the conveying device 36. Specifically, light irradiated from the linear light source 44B of the image reading unit 28 passes through the belt of the conveying device 36, which is composed of a belt conveyor. Therefore, since the belt is formed of a transparent material, the conveying surface 36A forming the conveying path has light transmissivity, and light from the linear light source 44B passes through the conveying surface 36A and is irradiated to the reading object T. In this case, the image reading unit 28 is arranged so that the conveying surface 36A is located between the linear light source 44B and the reading object T.

In this way, if the conveying surface 36A has optical transparency, it is possible to prevent the light irradiated from the linear light source 44B from being absorbed or reflected before it reaches the reading target T, and therefore it is possible to prevent the outline of the area corresponding to the reading target T from becoming blurred in the read image data. Note that, although the linear light source 44B is arranged outside the conveying device 36 (below the belt) in FIG. 25, it is not limited to this, and it may be arranged inside the conveying device 36 (inside the belt).

Furthermore, the specific configurations etc. described in the above-mentioned embodiments are merely examples and can be modified as appropriate depending on the actual product. For example, the image reading unit 28 in the image reading device 10 may be omitted so that it functions as an image processing device. In this case, however, an external image reading unit capable of functioning as the image reading unit 28 in each embodiment is communicatively connected to the image processing device. Also, for example, the image reading device 10 described in each embodiment may be implemented as an inspection device for detecting foreign objects, scratches, defects, etc.

Furthermore, the order in which each step in the flow chart shown in the above embodiment is processed can be changed as appropriate, provided that the same results are obtained.

In addition, ZNCC (Zero-mean Normalized Cross-Correlation) can be used as an evaluation index for the similarity of output image data to scanned image data with clear contours of the target area.

10 Image reading device 20 Control unit 22 CPU
24 RAM
26 Storage unit 28 Image reading unit 32 Image pickup element 34 Slit member 36 Conveyor device 44 Light source 90 Image processing unit 92 Spectrum generation processing unit 94 Comparison processing unit 96 Composite spectrum generation processing unit 98 Image generation processing unit 100 Correction processing unit 202a Read image data 202b Spectrum data 202c Composite spectrum data 202d Output image data 202e Corrected image data

Claims (17)

An image processing device that performs image processing on a plurality of scanned image data obtained by scanning a single object in lines along main scanning directions each extending in a different direction, the image processing device comprising:
a spectrum generation processing unit that generates a plurality of spectrum data by performing a Fourier transform process on each of the plurality of read image data;
a comparison processing unit that compares components corresponding to the same frequency of the plurality of spectrum data with each other;
a synthetic spectrum generation processing unit that generates synthetic spectrum data by selecting one component from among the components corresponding to the same frequency of the plurality of spectral data based on a comparison result by the comparison processing unit;
an image generation processing unit that generates output image data by performing an inverse Fourier transform process on the synthetic spectral data. The image processing device according to claim 1, wherein the synthetic spectrum generation processing unit selects the largest component from among the components corresponding to the same frequency of the plurality of spectral data. a correction processing unit that performs a correction process on at least one of the plurality of read image data before the Fourier transform process is performed so that an area corresponding to the object is located in the same position on an image in each of the plurality of read image data,
The image processing apparatus according to claim 1 , wherein the spectrum generating section performs a Fourier transform process on each of the plurality of read image data that have been subjected to the correction process by the correction section. The image processing device according to claim 3, wherein the correction processing unit corrects the inclination or magnification of an area in at least one of the plurality of read image data before the Fourier transform process is performed, the area corresponding to the object on the image in the read image data. a plurality of line sensors each capable of reading an image in a line shape along a main scanning direction, the main scanning direction extending in a direction different from the sub-scanning direction;
5. An image reading apparatus comprising: the image processing apparatus according to claim 1, which performs image processing on a plurality of read image data obtained by reading the same object with the plurality of line sensors. The image reading device according to claim 5, wherein the plurality of line sensors includes two line sensors arranged so that their main scanning directions are perpendicular to each other. The image reading device according to claim 6, wherein the two line sensors are arranged so that their main scanning directions intersect at an angle of 45° with respect to the sub-scanning direction. The image reading device according to any one of claims 5 to 7, wherein the plurality of line sensors are arranged in a direction intersecting the sub-scanning direction. At least one line sensor having an elongated shape along a longitudinal direction intersecting with a sub-scanning direction;
and an image processing device according to any one of claims 1 to 4, which performs image processing on a plurality of read image data obtained by reading the same object with the at least one line sensor;
An image reading device, wherein the at least one line sensor includes a plurality of first image pickup elements inclined with respect to the longitudinal direction and arranged in a row in the longitudinal direction, and a plurality of second image pickup elements inclined with respect to the longitudinal direction at an angle different from that of the first image pickup elements and arranged in a row in the longitudinal direction. the at least one line sensor includes two line sensors;
One of the two line sensors has the plurality of first image pickup elements arranged in the longitudinal direction and inclined with respect to the longitudinal direction,
10. The image reading device according to claim 9, wherein the other of the two line sensors has the plurality of second image pickup elements arranged in the longitudinal direction and inclined at an angle different from that of the first image pickup elements. The image reading device according to claim 9, wherein one of the at least one line sensor has the plurality of first image pickup elements arranged in the longitudinal direction and inclined with respect to the longitudinal direction, and the plurality of second image pickup elements arranged in the longitudinal direction and inclined at an angle different from that of the first image pickup elements. At least one line sensor capable of reading an image in a line shape along a main scanning direction;
and an image processing device according to any one of claims 1 to 4, which performs image processing on a plurality of read image data obtained by reading the same object with the at least one line sensor;
The line sensor includes:
A light source extending in the main scanning direction and irradiating the object with light;
an image reading device having a slit member extending in the main scanning direction and disposed between the object and the light source; At least one line sensor capable of reading an image in a line shape along a main scanning direction;
and an image processing device according to any one of claims 1 to 4, which performs image processing on a plurality of read image data obtained by reading the same object with the at least one line sensor;
The line sensor is an image reading device having a line laser light source that irradiates light onto the object. At least one line sensor capable of reading an image in a line shape along a main scanning direction;
The image processing device according to any one of claims 1 to 4, which performs image processing on a plurality of read image data obtained by reading the same object with the at least one line sensor;
a plurality of conveying devices that convey the object along a conveying path that extends in a direction intersecting the main scanning direction;
The line sensor includes:
A light source that irradiates the object with light,
The plurality of conveying devices are disposed along the conveying path with gaps between them,
An image reading device, wherein the light source irradiates light through the gap onto the object being transported along the transport path, or causes the light irradiated onto the object being transported along the transport path to enter the gap. At least one line sensor capable of reading an image in a line shape along a main scanning direction;
The image processing device according to any one of claims 1 to 4, which performs image processing on a plurality of read image data obtained by reading the same object with the at least one line sensor;
a conveying device that conveys the object along a conveying path that extends in a direction intersecting the main scanning direction,
The line sensor includes:
A light source that irradiates the object with light,
The conveying device includes a conveying surface that forms the conveying path, and the conveying surface is optically transparent, so that light from the light source passes through the conveying surface and is irradiated onto the object. An image processing method for performing image processing on a plurality of scanned image data obtained by scanning a same object in lines along main scanning directions each extending in a different direction, the method comprising:
a spectrum generating step of generating a plurality of spectrum data by performing a Fourier transform process on each of the plurality of read image data;
a comparison step of comparing components corresponding to the same frequency of the plurality of spectrum data with each other;
a synthetic spectrum generating step of generating synthetic spectrum data by selecting one component from among the components corresponding to the same frequency of the plurality of spectral data based on a comparison result of the comparing step;
and an image generating step of generating output image data by performing an inverse Fourier transform process on the synthetic spectral data. An image processing program for performing image processing on a plurality of scanned image data obtained by scanning an identical object in lines along main scanning directions each extending in a different direction, the program comprising:
a spectrum generating step of generating a plurality of spectrum data by performing a Fourier transform process on each of the plurality of read image data;
a comparison step of comparing components corresponding to the same frequency of the plurality of spectrum data with each other;
a synthetic spectrum generating step of generating synthetic spectrum data by selecting one component from among the components corresponding to the same frequency of the plurality of spectral data based on a comparison result of the comparing step;
and an image generating step of generating output image data by performing an inverse Fourier transform process on the synthetic spectral data.

Images