JP6920739B2 - Defect inspection equipment - Google Patents

Description

The present invention relates to a defect inspection device for a continuous sheet-like product having a fibrous structure and a narrow width such as a woven fabric, a string, or a tape.

A defect inspection device for inspecting a continuous sheet-like material having a fiber structure is known. Such a defect inspection device is disclosed in, for example, Patent Document 1. This Patent Document 1 describes an inspection device that reduces the formation noise of a product by devising the arrangement of lighting.

Japanese Unexamined Patent Publication No. 8-50105

Here, as in Patent Document 1, the formation noise of a continuous sheet-like object made of a fiber structure is reduced, but when the surface unevenness of the product is large, the angle between the camera and the illumination is not specular. Even if this is done, the camera and the illumination will have a specular reflection relationship at any position on the uneven portion, and the formation noise of the product will increase. In addition, the formation noise becomes large even in the portion where the surface uneven portion is a shadow. Therefore, fine foreign matter and thin dirt defects are buried in the formation noise, which makes it difficult to detect the defects.

The present invention has been made in view of such circumstances, and an object of the present invention is to form a continuous sheet having a fibrous structure and a texture of an object to be inspected having a large surface unevenness in a narrow product such as a woven fabric, a string, or a tape. The purpose of the present invention is to provide a defect inspection device for reducing noise.

In order to solve the above problems, the defect inspection device of the present invention is a defect inspection device that is a continuous sheet-like object made of a fibrous structure and has a narrow width to be inspected, and the longitudinal direction is the above. Reflective illumination that is arranged on both sides in the width direction parallel to the transport direction of the object to be inspected and illuminates the object to be inspected from one main surface side of the object to be inspected, and the longitudinal direction is the width direction of the object to be inspected. The transmitted illumination that is arranged parallel to the object to be inspected and illuminates the object to be inspected from the other main surface side of the object to be inspected, and the reflected illumination and the object to be inspected that is illuminated by the transmitted illumination are on the main surface side of the one. A line sensor that captures images from the line sensor and a defect detection unit that detects defects in the object to be inspected based on image data obtained from the line sensor, and the reflected illumination corresponds to a position imaged by the line sensor. At the position in the longitudinal direction of the reflected illumination, a light-shielding plate is added to the surface facing the object to be inspected, or a plurality of illumination units arranged in the traveling direction of the object to be inspected are formed. in the longitudinal direction of the position of the reflected illumination corresponding to the position where the line sensor takes an image, and wherein the Rukoto spaced positions of the plurality of illumination portions.

Further, the defect inspection device of the present invention is the above-mentioned defect inspection device, and the object to be inspected is any of a woven fabric, a string, and a tape, and the unevenness on the surface of the object to be inspected is the reflection illumination. The line sensor captures the reflected light and the transmitted light generated by the object to be inspected when illuminated by the transmitted illumination.

Further, the defect inspection device of the present invention is the above-mentioned defect inspection device, and two reflection lights are arranged so as to sandwich the object to be inspected, and the interval between the two reflection lights arranged is the subject. The feature is that the width is equal to or larger than the width of the inspection object.

Further, the defect inspection device of the present invention is the above-mentioned defect inspection device, and the transmitted illumination is characterized in that the relationship with the line sensor is arranged so as to be positive transmission or scattering transmission.

Further, the defect inspection device of the present invention is the above-mentioned defect inspection device, and the reflected illumination and the transmitted illumination perform automatic dimming based on the amount of light entering the line sensor, and the line sensor is subjected to automatic dimming. It is characterized in that the ratio of the amount of incoming light is set within a predetermined range.

According to the present invention, it is possible to reduce the formation noise of an inspected object having a large surface unevenness in a product having a narrow width such as a woven fabric, a string, a tape, etc., which is a continuous sheet-like material having a fibrous structure, and is generated on the inspected object. It is possible to provide a defect inspection apparatus capable of detecting fine foreign matter and thin dirt defects and reducing the formation noise of an inspected object having a large surface unevenness.

It is a schematic block diagram which shows the structure of the defect inspection apparatus 1. It is a block diagram of the imaging unit 11a using the reflection illumination. It is a line sensor waveform which imaged the woven fabric by the image pickup unit 11a using the reflection illumination. It is a block diagram of the imaging unit 11b which devised the arrangement of the reflection illumination. It is a line sensor waveform which imaged the woven fabric by the image pickup unit 11b which devised the arrangement of the reflection illumination. It is a figure which shows the relationship of the specular reflection of the imaging unit 11a using the reflection illumination. It is a figure which shows the relationship of the specular reflection of the imaging unit 11b which devised the arrangement of the reflection illumination. It is a figure which shows the relationship of the specular reflection of the imaging unit 11b which devised the arrangement of the reflection illumination. It is a block diagram of the imaging unit 11c using transmission illumination. It is a line sensor waveform which imaged the woven fabric by the image pickup unit 11c using transmission illumination. It is a block diagram of the imaging unit 11d using the reflection illumination and the transmission illumination. It is a line sensor waveform which imaged the woven fabric by the image pickup unit 11d using the reflection illumination and the transmission illumination. It is a block diagram of the imaging unit 11 of this invention. It is a line sensor waveform which imaged the woven fabric by the image pickup unit 11 of this invention. It is a figure which showed the waveform of the defect part image | positioned with the image pickup part 11d of the prior art and the image pickup part 11 of this invention. It is a relationship diagram of the lighting arrangement of this invention and a product. It is a figure which showed the light amount ratio of the reflected illumination and the transmitted illumination of this invention. It is a figure which shows the fluctuation of the formation noise when the object 50 to be inspected is rotated. It is a relationship diagram of the lighting arrangement of Example 2 and a product. It is a figure which shows the fluctuation of the formation noise when the object 50 to be inspected in Example 2 is rotated.

Hereinafter, the defect inspection apparatus according to the embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a schematic configuration diagram showing the configuration of the defect inspection device 1.
The defect inspection device 1 has an imaging unit 11, a defect detection unit 12, a rotary encoder 13, and an inspection unit 14, and inspects the object to be inspected 50.
In the description of the present embodiment, as a comparative example with respect to the imaging unit 11, the configurations of the four imaging units (imaging unit 11a to 11d) will be described as comparative examples 1 to 4, and subsequently, the imaging unit 11 will be the imaging unit. It will be described that the formation noise can be reduced as compared with 11a to 11d of the imaging unit.

The object to be inspected 50 is in the form of a continuous sheet, and is a product having a narrow width such as a woven fabric, a string, or a tape. In the following embodiment, the case where the object to be inspected 50 is a woven fabric will be described as an example. The object to be inspected 50 has a long shape in the transport direction described later. The minor axis direction of the object to be inspected 50 corresponds to the width direction (X-axis direction), and the major axis (long) direction (Y-axis direction) corresponds to the transport direction (also referred to as traveling direction).
That is, the defect inspection device 1 is a continuous sheet-like material having a fibrous structure, and has a narrow width to be inspected, which is a predetermined width (limited to a predetermined width) according to the product specifications. do. Here, the predetermined width is, for example, a width of about 5 cm.

FIG. 13 is a block diagram of the imaging unit 11 of the present invention.
The reflective illumination 42 is arranged on both sides in the width direction in the longitudinal direction parallel to the transport direction of the object 50 to be inspected, and illuminates the object 50 to be inspected from one main surface side of the object 50 to be inspected.
As described above, the reflection illumination 42 is arranged on both sides in the width direction so that the amount of light when illuminating the object 50 having the above-mentioned predetermined width is uniform in the width direction.
The transmission illumination 82 is arranged in the longitudinal direction parallel to the width direction of the inspected object 50, and illuminates the inspected object 50 from the other main surface side of the inspected object 50.
The line sensor 21 takes an image of the object to be inspected 50 illuminated by the reflection illumination 42 and the transmission illumination 82 from one main surface side.
That is, the imaging unit 11 includes an illuminating device (reflected illumination 42 and transmitted illumination 82) that illuminates the object 50 to be inspected, and a line sensor 21 that receives the reflected light or transmitted light. The longitudinal direction of the imaging range of the imaging unit 11 is arranged so as to be orthogonal to the traveling direction of the object to be inspected 50.

Further, the object to be inspected 50 is any of a woven fabric, a string, and a tape, and when the unevenness on the surface of the object to be inspected 50 is illuminated by the reflection illumination 42 and the transmission illumination 82, the object to be inspected 50 is generated. The line sensor 21 captures the reflected light and the transmitted light.

As these line-shaped illumination devices (reflection illumination 42 and transmission illumination 82), for example, fluorescent lamps, quartz rod illuminations, LED illuminations, and the like are used.
Further, as the line sensor 21, for example, one having 2048 to 8192 elements is used. As for the number of elements, an appropriate number of elements and speed (for example, data rate, scan rate, etc.) are used for a predetermined number of elements according to the width, traveling speed, resolution, installation space, etc. of the object to be inspected 50. .. Further, the line sensor 21 receives the reflected light or transmitted light reflected by the object 50 to be inspected by the light emitted from the lighting device. The line sensor 21 receives the reflected light and the transmitted light in a state where the object to be inspected 50 is conveyed in the conveying direction.
The line sensor 21 outputs an electric signal (defect data) according to the shade of the color tone of the surface of the object to be inspected 50 to the defect detection unit 12. In other words, the line sensor 21 outputs an electric signal according to the light intensity distribution on the surface of the object to be inspected 50 in units of lines (hereinafter, inspection lines) in a direction orthogonal to the traveling direction of the object to be inspected 50. .. Examples of the line sensor 21 include a CMOS (complementary MOS) camera and a CCD (Charge-Coupled Device) camera.

Returning to FIG. 1, the defect detection unit 12 is composed of an image processing computer and an image board connected to the line sensor 21 of the imaging unit 11, and is a defect of the object 50 to be inspected based on the image data obtained from the line sensor 21. Is detected. The defect detection unit 12 includes, for example, a binarization unit, a run-length coding unit, and a connectivity processing unit. The defect detection unit 12 binarizes the defect data input from the line sensor 21 of the imaging unit 11 based on a predetermined threshold value, compresses the defect data, performs a connectivity process, and performs a defect feature amount. (Information representing the characteristics of the shape of the defect. For example, the area, width, length, aspect ratio, and area ratio of the defect) are measured. As the defect detection unit 12, an image processing device LSC-6000 manufactured by MEC Co., Ltd. can be used.

The rotary encoder 13 mainly brings the wheels of its own measuring unit into contact with the transfer roll, and measures the inspection length based on the rotation speed of the transfer roll. This transport roll transports the object to be inspected 50 in the transport direction. Here, the inspection length represents the distance from the inspection start position in the Y-axis direction of the object to be inspected 50.

The inspection unit 14 calculates the distance that the inspected object 50 is conveyed in the conveying direction based on the encoder value obtained from the rotary encoder 13, and obtains the coordinate value of the y coordinate. The inspection unit 14 determines the defect detection result based on the defect inspection result including the defect detection result obtained by the defect detection unit 12 and the coordinates (coordinate values of x-coordinates) representing the position of the defect detection result on the object to be inspected. Generates an inspection result map represented by a map representing the coordinates corresponding to the object 50 to be inspected. The inspection unit 14 outputs the obtained inspection result on the screen of a display device (not shown in FIG. 1). This display device may be provided by the inspection unit 14, or a display such as a liquid crystal display device may be connected to the outside. Further, this output may be not only display on the screen but also output of electronic data or printing.

(Comparative Example 1)
FIG. 2 is a configuration diagram of an imaging unit 11a using reflected illumination. The image pickup unit 11a includes a line sensor 21 and a reflection illumination 22.
When a woven fabric is the target (object 50 to be inspected), the light entering the line sensor 21 is not stable due to the unevenness of the surface, and the waveform greatly oscillates. Therefore, for the purpose of stabilizing the texture, the line-shaped illumination device is used as reflective illumination, and two reflective illuminations 22 are arranged. Similar to a general imaging unit, the line-shaped illumination device (reflection illumination 22) is arranged so as to be orthogonal to the traveling direction of the object 50 to be inspected. The imaging position of the line sensor 21 was set between the two reflected illuminations 22 arranged.

Table 1 shown below is a table for explaining examples of the equipment and the object to be inspected 50 used in the case of inspecting in the above-mentioned defect inspection device 1 and the image pickup unit 11a using the reflected illumination.

FIG. 3 is a line sensor waveform obtained by imaging a woven fabric with an imaging unit 11a using reflected illumination. The vertical axis of the figure is the output of the camera (256 gradations), and the horizontal axis is the camera pixels (256 pixels).
The woven fabric was imaged by the imaging unit 11a in FIG. 2 for the purpose of stabilizing the texture of the woven fabric, and the texture noise on the light side was 141% and the texture noise on the dark side was 42%. At this time, 100% is the average value of 128 scans of the acquired waveform.
From this, it was confirmed that the configuration of the imaging unit 11a in FIG. 2 is significantly affected by the surface unevenness of the woven fabric.

(Comparative Example 2)
FIG. 4 is a configuration diagram of the imaging unit 11b in which the arrangement of the reflected illumination is devised. The image pickup unit 11b includes a line sensor 21 and a reflection illumination 42.
In the image pickup unit 11a of FIG. 2, since the line-shaped illumination device (reflection illumination) is arranged so as to be orthogonal to the traveling direction, it was confirmed that the texture of the woven fabric is remarkably received.
Therefore, it was decided to arrange the lighting on the object 50 to be inspected without the lighting. In the imaging unit 11b shown in FIG. 4, two line-shaped illumination devices (reflection illumination 42) are arranged in parallel with the product (object 50 to be inspected). Taking advantage of the narrow product width, the lighting interval was the product width. That is, two reflective illuminations 42 are arranged so as to sandwich the object to be inspected 50.

Table 2 shown below is a table explaining examples of the equipment and the object to be inspected 50 used in the case of inspecting by the above-mentioned defect inspection device 1 and the imaging unit 11b in which the arrangement of the reflected illumination is devised.

FIG. 5 is a line sensor waveform in which the woven fabric is imaged by the image capturing unit 11b in which the arrangement of the reflected illumination is devised. The vertical axis of the figure is the output of the camera (256 gradations), and the horizontal axis is the camera pixels (256 pixels).
As a result of imaging the woven fabric by the imaging unit 11b of FIG. 4, the formation noise on the light side was 120% and the formation noise on the dark side was 56%. At this time, 100% is the average value of 128 scans of the acquired waveform.
From the above, it was confirmed that the image pickup unit 11b in FIG. 4 was able to reduce the noise on the bright side by 51% and the noise on the dark side by 24% as compared with the image pickup unit 11a in FIG.
However, since the formation noise on the dark side is still large, we investigated an imaging unit that can be further improved.

The degree of improvement of the bright side noise is calculated by assuming that the bright side noise is "bright side formation noise-100%" and the ratio of the bright side noise is "bright side noise of the imaging unit 11b" / "brightness of the imaging unit 11a". It can be confirmed by calculating with "side noise" and calculating the calculated ratio with "100% -ratio". Specifically, since the bright side noise of the imaging unit 11a = 141% -100% = 41% and the bright side noise of the imaging unit 11b = 120% -100% = 20%, the ratio of the bright side noise is 20%. / 41% = 48.8%, and the degree of improvement of bright noise is calculated as 100% -48.8% = 51.2% ≈ 51%.
On the other hand, the degree of improvement of the dark side noise is calculated by calculating the dark side noise as "100% -dark side ground noise" and setting the ratio of the dark side noise to "dark side noise of the imaging unit 11b" / "dark side of the imaging unit 11a". It can be confirmed by calculating with "noise" and calculating the calculated ratio with "100% -ratio". Specifically, since the dark side noise of the imaging unit 11a = 100% -42% = 58% and the dark side noise of the imaging unit 11b = 100% -56% = 44%, the ratio of the dark side noise is 44%. / 58% = 75.9%, and the degree of improvement of dark noise is calculated as 100% -75.9% = 24.1% ≈ 24%.

FIG. 6 is a diagram showing the relationship of specular reflection of the imaging unit 11a using reflected illumination. Here, for convenience of explanation, the line sensor 21 and the reflected illumination 22 of the imaging unit 11a shown in FIG. 2 will be described as the line sensor 61 and the reflected illumination 62 in FIG.
FIG. 6A is a diagram showing the relationship of specular reflection on a plane. When the angle of reflected illumination is α, the position of specular reflection is α × 2 for a smooth object to be inspected.
FIG. 6B is a diagram showing the relationship between specular reflection on an inclined surface. Since the object to be inspected 50 this time is a woven fabric made of fibers, the surface is uneven. When the inclination angle of the surface unevenness is 1 degree, the reflection angle is (α-1) × 2. Therefore, when the object to be inspected 50 is conveyed, the angle of specular reflection changes according to the unevenness of the product. In the case of FIG. 6B, when the uneven inclination angle of the object to be inspected is α / 2, the line sensor 61 and the reflected illumination 62 have a specular reflection relationship.
FIG. 6C is a three-dimensional view of the imaging unit 11a and the object to be inspected 50. The inclination image 64 here depicts a part of the uneven inclination of the object to be inspected 50.
From the above, in the image pickup unit 11a of FIG. 2, even if the line sensor 21 and the reflected illumination 22 are not related to specular reflection, if the object to be inspected has irregularities, a position where specular reflection occurs occurs. Therefore, it can be said that the formation noise becomes large.

7 and 8 are diagrams showing the relationship of specular reflection of the imaging unit 11b in which the arrangement of the reflected illumination is devised. Here, for convenience of explanation, the line sensor 21 and the reflected illumination 42 of the imaging unit 11b shown in FIG. 4 will be described as the line sensor 61 and the reflected illumination 73 in FIGS. 7 and 8.
FIG. 7A is a diagram showing the relationship of specular reflection on a plane. When the angle β of the reflected illumination is set when viewed from the transport direction of the object 50 to be inspected, the position where the specular reflection occurs is an angle β × 2 in the width direction of the object 50 to be inspected.
FIG. 7B is a diagram showing the relationship of specular reflection in a plane viewed from the width direction of the object to be inspected 50. The light emitted from the point A of the reflection illumination 73 is depicted as entering the line sensor 61 in the plan view, but as shown in FIG. 7A, the position of specular reflection is the width of the object to be inspected. Since it is a direction, the line sensor 61 and the reflected illumination 73 do not have a specular relationship.
FIG. 7C is a diagram of Example 1 showing the relationship of specular reflection on the inclined surface of the object to be inspected 50 when viewed from the width direction. Similar to FIG. 6, it is assumed that the uneven inclination angle of the object to be inspected 50 is β / 2. The light emitted from the point A of the reflection illumination 73 is reflected in the transport direction and the width direction (arbitrary angle δ × 2). Therefore, the line sensor 61 and the reflected illumination 73 do not have a specular reflection relationship.
FIG. 8A is a diagram of Example 2 showing the relationship of specular reflection on the inclined surface of the object to be inspected 50 when viewed from the width direction. Unlike FIG. 7C, it is a diagram assuming irradiation from point B of the reflective illumination 73. Even if the object to be inspected angle γ and the irradiation angle Θ are the positions where light enters the line sensor 61 on a flat surface, the position of specular reflection also faces the width direction, so that the line sensor 61 and the reflection illumination 73 still have specular reflection. It does not become a relationship.
FIG. 8B is a three-dimensional view of the imaging unit 11b and the object to be inspected 50. It can be seen that the line sensor 61 and the reflected illumination 73 do not have a specular reflection relationship even when the object to be inspected 50 has irregularities as described in FIGS. 7 (c) and 8 (a). The inclination image 64 here depicts a part of the uneven inclination of the object to be inspected 50.

(Comparative Example 3)
FIG. 9 is a configuration diagram of an imaging unit 11c using transmitted illumination. The image pickup unit 11c includes a line sensor 21 and a transmission illumination 82.
It was thought that the reflected light using the reflected illumination alone would create shadows in the gaps between the fibers, and the formation noise on the dark side could not be improved. Therefore, the object to be inspected 50 was imaged by the imaging unit 11c using transmitted illumination.
In FIG. 9, the line-shaped illuminating device (transmitted illuminating 82) is arranged so as to be orthogonal to the traveling direction of the object 50 to be inspected, similarly to the general imaging unit.

Table 3 shown below is a table for explaining examples of the equipment and the object to be inspected 50 used in the case of inspecting in the above-mentioned defect inspection device 1 and the imaging unit 11c using the transmitted illumination.

FIG. 10 is a line sensor waveform obtained by imaging a woven fabric with an imaging unit 11c using transmitted illumination. The vertical axis of the figure is the output of the camera (256 gradations), and the horizontal axis is the camera pixels (256 pixels).
As a result of imaging the woven fabric by the imaging unit 11c of FIG. 9, the formation noise on the light side was 203% and the formation noise on the dark side was 64%. At this time, 100% is the average value of 128 scans of the acquired waveform.
From the above, it was confirmed that the light side noise was increased by 151% and the dark side noise was reduced by 38% in the image pickup unit 11c of FIG. 9 as compared with the image pickup unit 11a of FIG.
The degree of improvement of the bright side noise is calculated by assuming that the bright side noise is "bright side formation noise-100%" and the ratio of the bright side noise is "bright side noise of the imaging unit 11c" / "bright side noise of the imaging unit 11a". It can be confirmed by calculating with and calculating the calculated ratio with "100% -ratio". Specifically, since the bright side noise of the imaging unit 11a = 141% -100% = 41% and the bright side noise of the imaging unit 11c = 203% -100% = 103%, the ratio of the bright side noise is 103%. / 41% = 251.2%, and the degree of improvement in bright noise is calculated to be 100% -251.2% = -151.22% ≈-151%.

On the other hand, the degree of improvement of the dark side noise is calculated by calculating the dark side noise as "100% -dark side ground noise" and setting the ratio of the dark side noise to "dark side noise of the imaging unit 11c" / "dark side of the imaging unit 11a". It can be confirmed by calculating with "noise" and calculating the calculated ratio with "100% -ratio". Specifically, since the dark side noise of the imaging unit 11a = 100% -42% = 58% and the dark side noise of the imaging unit 11c = 100% -64% = 36%, the ratio of the dark side noise is 36%. / 58% = 62.1%, and the degree of improvement in dark noise is calculated to be 100% -62.1% = 37.9% ≈ 38%.
At this time, the formation noise portion on the bright side was considered to be the gap portion between the fibers, which becomes noise on the dark side in the reflected light.

(Comparative Example 4)
FIG. 11 is a configuration diagram of an imaging unit 11d using reflected illumination and transmitted illumination. The configuration of the imaging unit 11d is the same as the configuration of the imaging unit described in the prior art document. The image pickup unit 11d includes a line sensor 21, a reflection illumination 22, and a transmission illumination 82.
It was considered that the portion of the imaging unit 11c using the transmitted illumination of FIG. 9 that had a formation noise on the bright side may be a shadow in the gap portion of the fiber in the reflected illumination.
Therefore, in order to confirm the possibility that the formation noise on the dark side can be further reduced by using the reflected illumination and the transmitted illumination together, the reflecting illumination 22 of FIG. 2 and the transmitted illumination 82 of FIG. 9 are used in the imaging unit 11d of FIG. The fabric was imaged with the combined configuration.

Table 4 shown below is a table explaining examples of the equipment and the object to be inspected 50 used in the case of inspecting in the above-mentioned defect inspection device 1 and the image pickup unit 11d using the reflected illumination and the transmitted illumination.

FIG. 12 is a line sensor waveform in which a woven fabric is imaged by an imaging unit 11d using reflected illumination and transmitted illumination. The vertical axis of the figure is the output of the camera (256 gradations), and the horizontal axis is the camera pixels (256 pixels). Moreover, the light receiving amount ratio of the line sensor of the reflected illumination and the transmitted illumination was set to 3: 1.
As a result of imaging the woven fabric by the imaging unit 11d of FIG. 11, the texture noise on the light side was 143% and the texture noise on the dark side was 70% in the waveform of the weave 1 in FIG. 12 (a). Further, in the waveform of the weaving method 2 of FIG. 12B, the formation noise on the light side was 151%, and the formation noise on the dark side was 76%. At this time, 100% is the average value of 128 scans of the acquired waveform.
From the results of the above weaving method 1, it was confirmed that the image pickup unit 11d in FIG. 11 was able to increase the light side noise by 5% and the dark side noise by 48% in the weaving method 1 as compared with the image pickup unit 11a in FIG. did it.
The degree of improvement of the bright side noise is calculated by assuming that the bright side noise is "bright side formation noise-100%" and the ratio of the bright side noise is "bright side noise of the imaging unit 11d" / "bright side noise of the imaging unit 11a". It can be confirmed by calculating with and calculating the calculated ratio with "100% -ratio". Specifically, since the bright side noise of the imaging unit 11a = 141% -100% = 41% and the bright side noise of the imaging unit 11d = 143% -100% = 43%, the ratio of the bright side noise is 43%. / 41% = 104.9%, and the degree of improvement of the light side noise is calculated as 100% -104.9% = -4.9% ≈ -5%.

On the other hand, the degree of improvement of the dark side noise is calculated by calculating the dark side noise as "100% -dark side ground noise" and setting the ratio of the dark side noise to "dark side noise of the imaging unit 11d" / "dark side of the imaging unit 11a". It can be confirmed by calculating with "noise" and calculating the calculated ratio with "100% -ratio". Specifically, since the dark side noise of the imaging unit 11a = 100% -42% = 58% and the dark side noise of the imaging unit 11d = 100% -70% = 30%, the ratio of the dark side noise is 30%. / 58% = 51.7%, and the degree of improvement of dark noise is calculated as 100% -51.7% = 48.3% ≈48%.

Further, from the result of the above weaving method 2, the light side noise is increased by 24% and the dark side noise is reduced by 59% in the weaving method 2 in the image capturing unit 11d of FIG. 11 as compared with the imaging unit 11a of FIG. Was confirmed.
The degree of improvement of the bright side noise is calculated by assuming that the bright side noise is "bright side formation noise-100%" and the ratio of the bright side noise is "bright side noise of the imaging unit 11d" / "bright side noise of the imaging unit 11a". It can be confirmed by calculating with and calculating the calculated ratio with "100% -ratio". Specifically, since the bright side noise of the imaging unit 11a = 141% -100% = 41% and the bright side noise of the imaging unit 11d = 151% -100% = 51%, the ratio of the bright side noise is 51%. / 41% = 124.4%, and the degree of improvement of the light side noise is calculated as 100% -124.4% = -24.4% ≈ -24%.

On the other hand, the degree of improvement of the dark side noise is calculated by calculating the dark side noise as "100% -dark side ground noise" and setting the ratio of the dark side noise to "dark side noise of the imaging unit 11d" / "dark side of the imaging unit 11a". It can be confirmed by calculating with "noise" and calculating the calculated ratio with "100% -ratio". Specifically, since the dark side noise of the imaging unit 11a = 100% -42% = 58% and the dark side noise of the imaging unit 11d = 100% -76% = 24%, the ratio of the dark side noise is 24%. / 58% = 41.4%, and the degree of improvement of dark noise is calculated as 100% -41.4% = 58.6% ≈ 59%.
It was confirmed that the formation noise on the dark side can be reduced by using the transmitted illumination together. It can be judged that the part where the formation noise on the light side is as expected is the formation noise on the dark side in the reflected illumination.

(First Embodiment)
Since the description of Comparative Examples 1 to 4 has been completed in the above description, the first embodiment of inspecting the object to be inspected 50 in the defect inspection device 1 described above and the imaging unit 11 using the reflection illumination and the transmission illumination is subsequently performed. The form will be described.
FIG. 13 is a block diagram of the imaging unit 11 of the present invention. As described above, the image pickup unit 11 includes a line sensor 21, a reflection illumination 42, and a transmission illumination 82.
The woven fabric was imaged in a configuration in which the reflection illumination 42 of FIG. 4 and the transmission illumination 82 of FIG. 9 were used in combination, which could reduce the formation noise as compared with the reflection illumination 22 of FIG.

Table 5 shown below is a table explaining examples of the equipment and the object to be inspected 50 used in the case of inspecting in the above-mentioned defect inspection device 1 and the image pickup unit 11 using the reflected illumination and the transmitted illumination.

FIG. 14 is a line sensor waveform obtained by imaging a woven fabric by the imaging unit 11 of the present invention. The vertical axis of the figure is the output of the camera (256 gradations), and the horizontal axis is the camera pixels (256 pixels). Moreover, the light receiving amount ratio of the line sensor of the reflected illumination and the transmitted illumination was set to 3: 1.
As a result of imaging the woven fabric by the imaging unit 11 of FIG. 13, the texture noise on the light side was 119% and the texture noise on the dark side was 78% in the waveform of the weave 1 in FIG. 14 (a).
Further, in the waveform of weaving method 2 in FIG. 14B, the formation noise on the light side was 123%, and the formation noise on the dark side was 78%. At this time, 100% is the average value of 128 scans of the acquired waveform.

From the above, it was confirmed that the image pickup unit 11 of the present invention was able to reduce the light side noise by 54% and the dark side noise by 62% in the weaving method 1 as compared with the image pickup unit 11a in FIG.
The degree of improvement of the bright side noise is calculated by assuming that the bright side noise is "bright side formation noise-100%" and the ratio of the bright side noise is "bright side noise of the imaging unit 11" / "bright side noise of the imaging unit 11a". It can be confirmed by calculating with and calculating the calculated ratio with "100% -ratio". Specifically, since the bright side noise of the imaging unit 11a = 141% -100% = 41% and the bright side noise of the imaging unit 11 = 119% -100% = 19%, the ratio of the bright side noise is 19%. / 41% = 46.3%, and the degree of improvement of bright noise is calculated as 100% -46.3% = 53.7% ≈ 54%.
On the other hand, the degree of improvement of the dark side noise is calculated by calculating the dark side noise as "100% -dark side ground noise" and setting the ratio of the dark side noise to "dark side noise of the imaging unit 11" / "dark side of the imaging unit 11a". It can be confirmed by calculating with "noise" and calculating the calculated ratio with "100% -ratio". Specifically, since the dark side noise of the imaging unit 11a = 100% -42% = 58% and the dark side noise of the imaging unit 11 = 100% -78% = 22%, the ratio of the dark side noise is 22%. / 58% = 37.9%, and the degree of improvement in dark noise is calculated to be 100% -37.9% = 62.1% ≈ 62%.

Further, from the above, it was confirmed that the image pickup unit 11 of the present invention was able to reduce the light side noise by 44% and the dark side noise by 62% in the weaving method 2 as compared with the image pickup unit 11a in FIG.
The degree of improvement of the bright side noise is calculated by assuming that the bright side noise is "bright side formation noise-100%" and the ratio of the bright side noise is "bright side noise of the imaging unit 11" / "bright side noise of the imaging unit 11a". It can be confirmed by calculating with and calculating the calculated ratio with "100% -ratio". Specifically, since the bright side noise of the imaging unit 11a = 141% -100% = 41% and the bright side noise of the imaging unit 11 = 123% -100% = 23%, the ratio of the bright side noise is 23%. / 41% = 56.1%, and the degree of improvement in light noise is calculated to be 100% -56.1% = 43.9% ≈ 44%.
On the other hand, the degree of improvement of the dark side noise is calculated by calculating the dark side noise as "100% -dark side ground noise" and setting the ratio of the dark side noise to "dark side noise of the imaging unit 11" / "dark side of the imaging unit 11a". It can be confirmed by calculating with "noise" and calculating the calculated ratio with "100% -ratio". Specifically, since the dark side noise of the imaging unit 11a = 100% -42% = 58% and the dark side noise of the imaging unit 11 = 100% -78% = 22%, the ratio of the dark side noise is 22%. / 58% = 37.9%, and the degree of improvement in dark noise is calculated to be 100% -37.9% = 62.1% ≈ 62%.

Further, as compared with the image pickup unit 11d (imaging unit having the configuration of the prior art) of FIG. 11, the light side noise can be reduced by 56% and the dark side noise can be reduced by 27% in the weaving method 1.
The degree of improvement of the bright side noise is calculated by assuming that the bright side noise is "bright side formation noise-100%" and the ratio of the bright side noise is "bright side noise of the imaging unit 11" / "bright side noise of the imaging unit 11d". It can be confirmed by calculating with and calculating the calculated ratio with "100% -ratio". Specifically, since the bright side noise of the imaging unit 11d = 143% -100% = 43% and the bright side noise of the imaging unit 11 = 119% -100% = 19%, the ratio of the bright side noise is 19%. / 43% = 44.2%, and the degree of improvement in light noise is calculated to be 100% -44.2% = 55.8% ≈56%.
On the other hand, the degree of improvement of the dark side noise is calculated by calculating the dark side noise as "100% -dark side ground noise" and setting the ratio of the dark side noise to "dark side noise of the imaging unit 11" / "dark side of the imaging unit 11d". It can be confirmed by calculating with "noise" and calculating the calculated ratio with "100% -ratio". Specifically, since the dark side noise of the imaging unit 11d = 100% -70% = 30% and the dark side noise of the imaging unit 11 = 100% -78% = 22%, the ratio of the dark side noise is 22%. / 30% = 73.3%, and the degree of improvement of dark noise is calculated as 100% -73.3% = 26.7% ≈ 27%.

Further, it was confirmed that the light side noise was reduced by 55% and the dark side noise was reduced by 8% in the weaving method 2 as compared with the imaging unit 11d (imaging unit having the configuration of the prior art) in FIG.
The degree of improvement of the bright side noise is calculated by assuming that the bright side noise is "bright side formation noise-100%" and the ratio of the bright side noise is "bright side noise of the imaging unit 11" / "bright side noise of the imaging unit 11d". It can be confirmed by calculating with and calculating the calculated ratio with "100% -ratio". Specifically, since the bright side noise of the imaging unit 11d = 151% -100% = 51% and the bright side noise of the imaging unit 11 = 123% -100% = 23%, the ratio of the bright side noise is 23%. / 51% = 45.1%, and the degree of improvement of bright noise is calculated as 100% -45.1% = 54.9% ≈ 55%.
On the other hand, the degree of improvement of the dark side noise is calculated by calculating the dark side noise as "100% -dark side ground noise" and setting the ratio of the dark side noise to "dark side noise of the imaging unit 11" / "dark side of the imaging unit 11d". It can be confirmed by calculating with "noise" and calculating the calculated ratio with "100% -ratio". Specifically, since the dark side noise of the imaging unit 11d = 100% -76% = 24% and the dark side noise of the imaging unit 11 = 100% -78% = 22%, the ratio of the dark side noise is 22%. / 24% = 91.7%, and the degree of improvement of dark noise is calculated as 100% -91.7% = 8.3% ≈ 8%.

FIG. 15 is a diagram showing waveforms of a defect portion imaged by the image pickup unit 11d of the prior art and the image pickup unit 11 of the present invention. FIG. 15A is a defect waveform acquired by the imaging unit 11d of FIG. FIG. 15B is a defect waveform acquired by the imaging unit 11 of FIG.
In FIG. 15A, the defect output can be confirmed for both dark defects and bright defects, but stable detection is difficult due to the large geological noise.
On the other hand, in FIG. 15B, since the formation noise is reduced, it can be confirmed that both dark defects and bright defects are clearly output.
As a result, the imaging unit 11 of the present invention cannot detect the conventional imaging units (imaging units 11a to 11d shown in Comparative Examples 1 to 4), and the fine foreign matter and thinness generated on the object to be inspected 50 are thin. It can be said that dirt defects can be detected.

FIG. 16 is a diagram showing the relationship between the lighting arrangement of the present invention and the product.
FIG. 16A is a relationship diagram viewed from the transport direction.
The distance between the reflected illuminations 42 in the examples was adjusted to the width X 1:50 mm of the object to be inspected 50.
In the present invention, since the object to be inspected is arranged so that there is no illumination, the distance between the reflected illuminations 42 needs to be equal to or greater than the width X1 of the object to be inspected 50.
Further, in the embodiment, the height Y1 of the reflective illumination 42 (height from the object to be inspected 50) was set to 100 mm. When the height Y1 of the reflection illumination 42 is 100 mm, the angle of view in which the illumination enters the field of view is 3.59 degrees because the height of the line sensor is Y2: 498 mm.
When the angle of view is 3.59 degrees, the field of view width X2 of the line sensor is 62 mm, which is 12 mm larger than the width X1: 50 mm of the object to be inspected 50. Therefore, even if the object to be inspected meanders and is conveyed, the amount of meandering is within 6 mm. If so, inspection is possible.

FIG. 16B is a relationship diagram seen from the width direction of the object to be inspected 50. In the embodiment, the center of the reflected illumination 42 is set as the visual field position of the line sensor. Further, the irradiation position of the transmission illumination 82 is also set to the center of the reflection illumination 42.
In the embodiment, the transmission illumination 82 has a relationship with the line sensor as the position of scattering transmission, but it may be a position of normal transmission.

FIG. 17 is a diagram showing the light intensity ratio of the reflected illumination and the transmitted illumination of the present invention. The vertical axis of the figure is the formation noise, and the horizontal axis is the light intensity ratio of the reflected illumination and the transmitted illumination.
It was confirmed that it is necessary to use both reflected illumination and transmitted illumination in order to reduce the formation noise of the object 50 to be inspected. Therefore, we confirmed the range in which the formation noise can be reduced by changing the light intensity ratio between the reflected illumination and the transmitted illumination.
The amount of light was adjusted by the amount of light received by the line sensor 21, and the ratio range was from 10: 0 to 0:10.
In FIG. 17, it was confirmed that the formation noise on the bright side was larger than that of the imaging unit 11d in FIG. 11 from the time when the ratio was 1: 1.
From the above, it was confirmed that the light intensity ratio between the reflected illumination and the transmitted illumination of the present invention needs to be dimmed (automatic dimming) in the range of 1.4: 1 to 11: 1 (predetermined range).

The dimming method of the reflected illumination and the transmitted illumination of the present invention needs to be performed alternately. With the object to be inspected placed, only the reflected illumination is turned on and the reflected illumination is dimmed. After dimming the reflected illumination, the transmitted illumination is turned on and the transmitted illumination is dimmed. The inspection is started after dimming the transmitted illumination.
For example, when the amount of light received from the reflected illumination is A and the amount of light received from the transmitted illumination is B, A + B = C is the amount of light to be inspected.
When the light intensity ratio between the reflected illumination and the transmitted illumination is 3: 1 and the target value of C is 120 Lv of the camera output (256 gradations), only the reflected illumination is turned on so that the received light amount of the camera is 90 Lv. Automatically dimming reflected lighting. After that, the transmitted light is turned on and the transmitted light is automatically adjusted so that the light receiving amount of the camera is 120 Lv, so that the light received amount of the camera is set to 120 Lv (A + B = C) with a light amount ratio of 3: 1. can. The order of dimming may be transmitted illumination first.

FIG. 18 is a diagram showing fluctuations in formation noise when the object to be inspected 50 is rotated. The vertical axis of the figure is the formation noise, and the horizontal axis is the rotation angle of the object to be inspected.
The surface of the object to be inspected 50 of the embodiment mainly has a warp structure (a structure along the transport direction). Therefore, the object to be inspected was rotated, and it was confirmed how many fiber structure angles could be reduced by the imaging unit of the present invention with respect to the transport direction.
In FIG. 18, it was confirmed that the formation noise fluctuated from 60 degrees. Therefore, it was confirmed that the fiber structure angle of the imaging unit 11 of the present invention, which can reduce the formation noise with respect to the transport direction, is 50 degrees or less.

It was confirmed that the imaging unit 11 of the present invention can reduce the formation noise and detect fine foreign matter and thin dirt defects. However, when the object to be inspected 50 is rotated by 60 degrees or more, an uneven angle is generated in which the line sensor 21 and the reflection illumination 42 have a specular reflection relationship, and the effect of the present invention is lost. Therefore, if the object to be inspected has a concavo-convex angle in which the line sensor and the reflected illumination have a specular reflection relationship in the width direction, the formation noise cannot be reduced.
In Example 2, the imaging unit 11 was devised on the assumption that there was an inspected object 50 having large irregularities in the width direction.

(Second embodiment)
FIG. 19 is a diagram showing the relationship between the lighting arrangement of the second embodiment and the product.
FIG. 19A is a relationship diagram viewed from the transport direction, and FIG. 19B is a relationship diagram viewed from the width direction. The change from FIG. 16 is that a shading plate 183 (non-illuminated portion) is added to the illumination at the center of the reflected illumination 42 (the position where the line sensor 21 takes an image).
By adding the light-shielding plate 183 to the center of the reflection illumination 42 (the position where the line sensor 21 takes an image), the light from the width direction is also blocked. Therefore, it is considered that the specular reflection component is not generated even if there is a concavo-convex angle in which the line sensor and the reflected illumination have a specular reflection relationship.

FIG. 20 is a diagram showing fluctuations in formation noise when the object to be inspected 50 in the second embodiment is rotated. FIG. 20A is a diagram showing fluctuations in noise when the light-shielding plate is 12.5 mm. FIG. 20B is a diagram showing fluctuations in noise when the light-shielding plate is 25 mm.
When a light-shielding plate (width X3: 12.5 mm) is attached, the formation noise increases from 70 degrees.
It was confirmed that when the light-shielding plate width (width X3: 25 mm) was attached, the formation noise could be maintained even when rotated up to 90 degrees.
From the above, it can be confirmed that by inserting a light-shielding plate into the reflected illumination 42, the effect of reducing the formation noise can be maintained even when the line sensor 21 and the reflected illumination 42 have a concavo-convex angle in which the line sensor 21 and the reflected illumination 42 have a specular reflection relationship. rice field.
In addition, as a configuration of the reflection illumination 42, a plurality of illumination units may be arranged in the traveling direction of the object to be inspected 50, and an interval position (non-illumination portion) of the plurality of illumination units may be provided at a position where the line sensor 21 takes an image. good.

The defect inspection apparatus according to the above-described embodiment may be realized by a computer. In that case, the program for realizing this function may be recorded on a computer-readable recording medium, and the program recorded on the recording medium may be read by the computer system and executed. The term "computer system" as used herein includes hardware such as an OS and peripheral devices. Further, the "computer-readable recording medium" refers to a portable medium such as a flexible disk, a magneto-optical disk, a ROM, or a CD-ROM, or a storage device such as a hard disk built in a computer system. Further, a "computer-readable recording medium" is a communication line for transmitting a program via a network such as the Internet or a communication line such as a telephone line, and dynamically holds the program for a short period of time. It may also include a program that holds a program for a certain period of time, such as a volatile memory inside a computer system that serves as a server or a client in that case. Further, the above program may be for realizing a part of the above-mentioned functions, and may be further realized for realizing the above-mentioned functions in combination with a program already recorded in the computer system. It may be realized by using a programmable logic device such as FPGA (Field Programmable Gate Array).

Although the embodiments of the present invention have been described in detail with reference to the drawings, the specific configuration is not limited to this embodiment, and includes designs and the like within a range that does not deviate from the gist of the present invention.

1 ... Defect inspection device, 11, 11a, 11b, 11c, 11d ... Imaging unit, 12 ... Defect detection unit, 13 ... Rotary encoder, 14 ... Inspection unit, 21,61 ... Line sensor, 22, 42, 62, 73 ... Reflective lighting, 50 ... Inspected object, 82 ... Transmitted lighting

Claims (5)

A defect inspection device that is a continuous sheet-like material made of a fibrous structure and inspects a narrow object to be inspected.
Reflective illumination whose longitudinal direction is parallel to the transport direction of the object to be inspected and is arranged on both sides in the width direction to illuminate the object to be inspected from one main surface side of the object to be inspected.
Transmitted illumination, which is arranged in the longitudinal direction parallel to the width direction of the object to be inspected and illuminates the object to be inspected from the other main surface side of the object to be inspected.
A line sensor that images the object to be inspected illuminated by the reflected illumination and the transmitted illumination from the one main surface side.
A defect detection unit that detects defects in the inspected object based on image data obtained from the line sensor is provided .
The reflected illumination has a light-shielding plate added to the surface facing the object to be inspected at a position in the longitudinal direction of the reflected illumination corresponding to the position imaged by the line sensor, or the object to be inspected. It is composed of a plurality of illumination portions arranged in the direction of travel, in the longitudinal direction of the position of the reflected illumination corresponding to the position where the line sensor takes an image, and wherein the Rukoto spaced positions of the plurality of illumination portions Defect inspection equipment. The object to be inspected is any of a woven fabric, a string, and a tape, and when the unevenness on the surface of the object to be inspected is illuminated by the reflected illumination and the transmitted illumination, the reflected light generated by the object to be inspected is generated. The defect inspection apparatus according to claim 1, wherein the line sensor captures the transmitted light. Two of the reflected lights are arranged so as to sandwich the object to be inspected.
The defect inspection apparatus according to claim 1 or 2, wherein the distance between the two reflected illuminations is equal to or greater than the width of the object to be inspected. The defect inspection apparatus according to any one of claims 1 to 3, wherein the transmitted illumination is arranged so that the relationship with the line sensor is positive transmission or scattered transmission. The reflected illumination and the transmitted illumination are characterized in that automatic dimming is performed based on the amount of light entering the line sensor, and the ratio of the amount of light entering the line sensor is set within a predetermined range. 4. The defect inspection apparatus according to any one of claims 4.

Images