JP2003004647A - Defect inspection apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a defect inspection apparatus which can highly accurately detect defects of an inspection object having a colored pattern. SOLUTION: The inspection object 100 having the colored pattern is illuminated with a light having a wavelength of not smaller than 780 nm by a light source 1. Only a light reflected at an inspection object face of the inspection object 100 and having a wavelength of not smaller than 780 nm is received by a camera 2, whereby an image of the inspection object 100 is formed. Defects of the inspection object 100 are detected by a CPU 7 or the like with the use of light and dark image data formed from the formed image.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a defect inspection device for inspecting defects on an inspection object having a color pattern.

[0002]

2. Description of the Related Art As a conventional defect inspection method, an image of an inspection object irradiated with visible light is taken by a camera,
By comparing the grayscale image data created using the captured image with a predetermined threshold value that serves as a reference value, scratches on the inspection object and foreign matter adhering to the surface, weaving of the fabric, bubbles in the inner layer of the transparent body, etc. There is known a method of detecting a local change in a physical quantity of a defect as a defect.

According to the above-mentioned defect inspection method, when inspecting a defect of an inspection object whose surface is made of a woven fabric, even if the inspection object is plain, the level of the gray image data of the normal portion of the weave structure and the weave image data are woven. Since there is no difference between the level of the grayscale image data of the defective portion where the texture is changed, the defect of the fabric cannot be detected.

On the other hand, the patent No. 3013 by the present inventors
In Japanese Patent No. 789, the tissue cycle of the fabric is calculated by using the grayscale image data created from the image of the fabric irradiated with visible light, and is set in the grayscale image data based on the calculated tissue cycle. A detection method is disclosed in which the statistics of a pair of comparison areas are compared to extract a defect. In this inspection method, the presence or absence of defects is determined by extracting the discontinuity of the weave structure, so even if there is no difference between the level of the grayscale image data of the defective part and the level of the grayscale image data of the normal part, the fabric The defect can be detected with high accuracy.

[0005]

However, the above-mentioned inspection method can be applied only to plain fabrics,
It is not possible to detect a defect of the inspection object having a color pattern. For example, when inspecting a fabric having a blue and red fabric pattern in the comparison area, the blue and red color information is reflected from the level fluctuation of the grayscale image data that reflects the shade information due to the unevenness of the surface of the fabric. The level fluctuation of the grayscale image data becomes much larger. Therefore, the true unevenness on the surface of the fabric cannot be extracted, and the color pattern is erroneously detected as a defect of the fabric.

Also in the above defect inspection method, when there is a color pattern in the vicinity of the defect part of the inspection object or in the defect part itself, the level fluctuation of the grayscale image data due to the color pattern is greater than the level fluctuation of the grayscale image data due to the defect. It becomes large, and the color pattern is erroneously detected as a defect.

An object of the present invention is to provide a defect inspection apparatus capable of detecting a defect of an inspection object having a color pattern with high accuracy.

[0008]

A defect inspection apparatus according to the present invention is a defect inspection apparatus for inspecting a defect of an inspection object having a color pattern, and is 780 nm reflected or transmitted on an inspection object surface of the inspection object. An image pickup means for receiving only light having the above wavelength to pick up an image of the inspection target surface, and a detection means for detecting a defect of the inspection target by using the image picked up by the image pickup means are provided.

In the defect inspection apparatus according to the present invention, 78 reflected or transmitted on the inspection object surface of the inspection object.
An image of the inspection target surface is captured by receiving only light having a wavelength of 0 nm or more, and a defect of the inspection target is detected using the captured image. At this time, it is possible to remove the light having the wavelength in the visible light region to remove the color pattern information of the inspection object, and to capture the image of the inspection object in which only the shadow information due to the physical unevenness is reflected. The defect of the inspection object having the color pattern can be detected with high accuracy without being affected by the color pattern.

The image pickup means preferably receives only near-infrared rays reflected or transmitted by the inspection target surface of the inspection target object to capture an image of the inspection target surface.

In this case, since an image of the surface to be inspected is picked up by receiving only the near infrared ray reflected or transmitted on the surface to be inspected of the object to be inspected, an inexpensive light source such as a light emitting diode for irradiating only the near infrared ray. Alternatively, a defect inspection device can be configured by using an inexpensive optical filter or the like that transmits only near infrared rays without transmitting visible light, and detects defects of an inspection object having a color pattern with high accuracy and at low cost. can do.

The image pickup means includes a light projecting means for irradiating the surface to be inspected of the object to be inspected with infrared rays, and the wavelength of near infrared rays emitted by the light projecting means is 800 nm or more and 1100 nm.
The following is preferable.

In this case, the light to be inspected can be irradiated with the light in the Herschel region, so that the absorption of the light derived from the overtone and bonding sound of the functional group of the inspection object becomes extremely weak, and the inspection object Since the difference in light absorption due to the material and the color that constitutes the color pattern of the inspection object does not occur due to the pigments and dyes, a grayscale image in which the color information of the color pattern is completely removed from the image of the inspection surface is obtained. Therefore, it is possible to detect the defect of the inspection object having the color pattern with higher accuracy.

[0014]

DESCRIPTION OF THE PREFERRED EMBODIMENTS A defect inspection apparatus according to an embodiment of the present invention will be described below with reference to the drawings.
FIG. 1 is a block diagram showing the configuration of a defect inspection apparatus according to an embodiment of the present invention.

The defect inspection apparatus shown in FIG. 1 comprises a light source 1, a camera 2, an A / D (analog / digital) converter 3, an image memory 4, an image processing circuit 5, an image display 6, a CPU (central processing unit). ) 7, ROM (Read Only Memory) 8,
A RAM (random access memory) 9, an image bus 10 and a CPU bus 11 are provided.

The camera 2 is connected to the A / D converter 3,
The D / D converter 3, the image memory 4, the image processing circuit 5, and the image display 6 are an image bus 1 for transferring image data and the like.
Connected to each other by 0. Further, the A / D converter 3, the image memory 4, the image processing circuit 5, the image display 6, the CPU
7, ROM 8 and RAM 9 are mutually connected by a CPU bus 11 for transferring various data and the like.

The light source 1 is composed of, for example, a light emitting diode or a semiconductor laser, and generates only light having a wavelength of 780 nm or more. For example, the light generated on the inspection target surface of the sheet-shaped inspection target 100 is used. Irradiate. In addition,
The form of the inspection object 100 is not particularly limited to the above-mentioned sheet shape, and may be other forms such as a long shape and a three-dimensional shape.

The camera 2 includes a camera lens and an area type C.
It is composed of an imaging device such as a CD (charge coupled device). The camera 2 collects the light reflected by the inspection target surface of the inspection target 100 of the light emitted by the light source 1 by the camera lens and forms an image on the light receiving surface of the area type CCD, and the inspection target 100 The image of the surface to be inspected is captured.

At this time, the camera 2 uses the inspection object 100.
Since only the light having a wavelength of 780 nm or more reflected on the inspection target surface is received to capture the image of the inspection target surface, the light in the visible light region is excluded to remove the color pattern information of the inspection target. However, it is possible to capture an image of the inspection target object in which only the shadow information due to the physical unevenness is reflected.

The light received by the camera 2 is preferably near infrared rays (for example, electromagnetic waves having a wavelength range of 780 nm to 2000 nm). In this case, since the image is captured using near-infrared light that does not include light in the visible light region, it is possible to configure the defect inspection device using a light source such as an inexpensive light-emitting diode that emits only near-infrared light. It is possible to detect a defect of an inspection object having a color pattern with high accuracy and at low cost.

Further, the wavelength of near infrared rays emitted from the light source 1 is more preferably 800 nm or more and 1100 nm or less. In this case, since it is possible to irradiate the inspection object 100 with light in the Herschel region,
The absorption of light derived from the overtones and the combined sounds of the functional groups of the inspection object 100 becomes extremely weak, and the material of the inspection object 100,
Since there is no difference in absorption of light due to pigments, dyes, etc. that determine the color that constitutes the color pattern of the inspection object 100, the shade that completely excludes the color information of the color pattern from the image of the inspection object surface of the inspection object 100. Images can be obtained.

The light source 1 is not particularly limited to the above example, and for example, a light source for irradiating light having a wavelength in the visible light region such as a halogen lamp is used, and as will be described later, an optical element for transmitting only near infrared rays is transmitted. Of the light emitted from the light source, only near infrared rays may be transmitted and emitted using a filter or the like. Further, the form of illumination by the light source 1, the directivity, the irradiation angle, and the like can be appropriately determined according to the characteristics of the inspection object and the defect to be extracted, and various changes can be made.

The CCD used in the camera 2 is not particularly limited as long as it has sensitivity to light in the near infrared region and can store sufficient electric charge within the update period of the CCD. Various CCDs can be used. Also, the form of the CCD is not particularly limited to the area type described above, and a line type CCD or the like may be used.

The A / D converter 3 converts the grayscale image signal of an analog signal corresponding to the image of the inspection target surface of the inspection target 100 output from the camera 2 into the grayscale image data of an 8-bit digital signal, for example. To do. The image memory 4 is A
The grayscale image data and the like output from the / D converter 3 are stored.

The image processing circuit 5 performs predetermined preprocessing such as differential emphasis processing on the grayscale image data stored in the image memory 4 in order to effectively extract defects in the inspection object 100. The image processing circuit 5 also calculates various statistical values from the processed grayscale image data. The RAM 9 stores various statistical values calculated by the image processing circuit 5.
Here, as the statistic, the maximum value, the average value, the minimum value, the variance value, the covariance value, the coefficient of variation, the skewness, the sharpness, the correlation coefficient, the standard deviation value, and the maximum value of the density obtained from the grayscale image data are used. Frequent density value etc.

The CPU 7 executes a predetermined processing program stored in the ROM 8 in advance, compares the reference values of various statistical values stored in the ROM 8 or the like with the statistical values stored in the RAM 9, and inspects them. The presence or absence of a defect in the object 100 is determined. Specifically, the CPU 7 compares difference values obtained by comparing the various statistical values in the vicinity area in the image of the inspection target surface, correlation of the image pattern, or comparison of the statistical value with the reference value. The defect of the inspection object 100 is extracted by performing. The image display 6 is composed of a display or the like, and displays an image or the like including a defect extracted by the CPU 7.

In this way, the presence or absence of a defect is determined by comparing the statistical value obtained from the grayscale image data created from the image taken by the camera 2 by the CPU 7 etc. with the preset reference value. You can Also,
When the image captured by the camera 2 includes periodic shading information, the CPU 7 or the like sets a pair of comparison regions based on the periodic information in the image and compares the statistical values extracted from the respective comparison regions. The presence or absence of a defect may be detected by comparing the value obtained in this way with a preset reference value.

The defect detection method performed after obtaining the image in which the color information is excluded is not particularly limited to the above-mentioned example, and various methods can be adopted. 2778531, Japanese Patent No. 28007
26, Japanese Patent No. 3013789, Japanese Patent No. 30
The defect inspection methods disclosed in Japanese Patent No. 13927, Japanese Patent No. 3013928, Japanese Patent No. 3063719, etc. may be used.

In this embodiment, the light source 1 and the camera 2 correspond to the image pickup means, and the A / D converter 3, the image memory 4, the image processing circuit 5, the image display 6, the CPU 7, the ROM.
8, the RAM 9, the image bus 10 and the CPU bus 11 correspond to the detecting means, and the light source 1 corresponds to the light projecting means.

Next, the light source 1 and the like shown in FIG. 1 will be described in more detail. FIG. 2 is a perspective view showing a specific example of a light source used as the light source 1 shown in FIG. Figure 2 (a)
The light source 1 shown in FIG. 1 is a module in which a plurality of light emitting diodes 21 are arranged in a line and is modularized, and can cover the imaging visual field of the camera lens of the camera 2 and can secure uniform illuminance within the imaging visual field. it can.

The light source 1 shown in FIG. 2B is a module in which a plurality of light emitting diodes 21 are arranged in a staggered pattern to cover the imaging field of view of the camera lens of the camera 2 and to capture an image. It is possible to secure a more uniform illuminance within the field of view.

The arrangement pattern of the light emitting diodes is as follows.
The present invention is not particularly limited to the above examples, and various changes can be made according to the imaged area of the inspection object 100 and the like. However, in order to reduce the illumination unevenness in the imaged region (illuminated region) of the inspection object 100, it is preferable to make the pitch between the light emitting diodes 21 as narrow as possible.

FIG. 3 is a perspective view showing another specific example of the light source 1 and the camera 2 shown in FIG. As shown in FIG. 3, the camera lens 41 of the camera 2 is arranged so as to face the optical filter 32 provided in the opening of the housing 31.
Is fixed in the housing 31, and the light source 1 is arranged outside the housing 31 so as to surround the optical filter 32.

Here, when the light source 1 irradiates the inspection target surface of the inspection target object 100 with light, the optical filter 32 transmits only near-infrared rays, and out of the light reflected by the inspection target surface of the inspection target object 100. Only near infrared rays are guided to the CCD in the camera 2 via the camera lens 41. Therefore, even when the light source 1 emits light in the visible light region other than near infrared rays, the camera 2 can receive only near infrared rays. Further, when the defect inspection apparatus is used in a room or the like, the inspection object is irradiated with light in the visible light region by an ordinary illumination device provided in the room, but the optical filter 32 causes the visible light by the illumination other than the light source 1 to be inspected. The light in the area can also be excluded, and in this case as well, the camera 2 can receive only near infrared rays.

As described above, in the present embodiment, the 780n reflected from the inspection target surface of the inspection target 100.
An image of the inspection target surface is captured by receiving only light having a wavelength of m or more, and the inspection target 100 is captured using the captured image.
The defect of is detected. Therefore, the light in the visible light region is eliminated to completely remove the color pattern information of the inspection object 100,
It is possible to capture an image of the inspection object 100 in which only the shadow information due to the physical unevenness is reflected. As a result, the defect of the inspection target 100 having the color pattern can be detected with high accuracy without being affected by the color pattern.

That is, it is possible to detect a local change in the physical quantity of the inspection object having a color pattern as a defect without being affected by the color pattern. For example, a scratch or a surface on the surface of an object having a color pattern can be detected. It is possible to highly accurately detect foreign matter adhering to the substrate, the weaving of a woven fabric having a color pattern, inner-layer air bubbles of a transparent body having a color pattern, etc. as defects of the inspection object.

Next, a specific defect detection result will be described. First, the result of inspecting defects such as scratches on the surface of the print sheet and adhesion of foreign matter using the light source 1 and the camera 2 shown in FIG. 3 will be described. Here, the light source 1 shown in FIG.
Is used as the light emitting diode having the relative radiant intensity characteristic shown in FIG. 4, and the optical filter 32 is the optical filter having the transmittance characteristic shown in FIG.
As the CCD of the camera 2, a CCD having the relative sensitivity characteristic shown in FIG. 6 was used.

The light emitting diode used as the light source 1 is shown in FIG.
Has a relative radiant intensity (%) shown in the center wavelength of 950n
m, the wavelength range is 900 to 1000 nm, and the full width at half maximum is 40 nm. The opening angle of this light emitting diode was 30 °, and the light emitted from the light emitting diode was applied to the inspection target surface of the print sheet at an irradiation angle of 45 °.

The optical filter 32 has the transmittance (%) shown in FIG. 5 and cuts off light having a wavelength shorter than 900 nm. Therefore, even when the light reflected by the printing sheet due to external illumination or the like has a wavelength shorter than 900 nm, the wavelength of the light incident on the CCD of the camera 2 via the camera lens 41 is 900 nm or more.

The CCD of the camera 2 has the relative sensitivity (%) shown in FIG. 6, and is sensitive to wavelengths other than the wavelength range of the light emitted from the light emitting diode, for example, wavelengths in the visible light region. However, since the wavelength of light incident on the CCD is limited to 900 nm or more by the optical filter 32 as described above, the CCD does not react to light having a wavelength shorter than 900 nm, and the CCD emits light. It has a sufficient sensitivity only to the wavelength range of light (900 to 1000 nm).

FIG. 7 is a diagram schematically showing the state of the surface of the print sheet which is the inspection object. The printing sheet 100a shown in FIG. 7 has a red stripe-shaped color pattern R and a blue stripe-shaped color pattern B, and has scratches D formed on the surface thereof and foreign matter F attached thereto. And

FIG. 8 shows a printing sheet 100 shown in FIG. 7 using a light emitting diode having the characteristics shown in FIGS.
FIG. 9 is a diagram showing the waveform of the density value of the grayscale image data created from the image of a, and FIG. 9 is a diagram showing an image of the print sheet shown in FIG. 7 photographed using visible light as in the conventional defect detection apparatus. It is a figure which shows the waveform of the density value of the produced grayscale image data. The horizontal axis of FIGS. 8 and 9 indicates the position of the print sheet 100a on the broken line A shown in FIG.

When a light emitting diode or the like having the characteristics shown in FIGS. 4 to 6 is used, in this embodiment, as shown in FIG. 8, a red striped color pattern R and a blue striped color are used. Printing sheet 10 without being affected by pattern B
When the density value of 0a is obtained and a portion having the reference value RV or less is detected as a defect of the print sheet 100a, the defect portion P1 corresponding to the scratch D can be detected and the foreign material F can be detected.
It was possible to detect the defective portion P2 corresponding to.

On the other hand, when visible light is used as in the conventional defect detecting apparatus, as shown in FIG. 9, printing is performed under the influence of the red striped color pattern R and the blue striped color pattern B. Red striped color pattern R on the sheet 100a
The density value of the portion P3 corresponding to and the portion P5 corresponding to the blue stripe-shaped color pattern B decreased. In this state, when detecting a portion of the reference value RV or less as a defect of the printing sheet 100a, a defect portion corresponding to the scratch D that is not affected by the red stripe color pattern R and the blue stripe color pattern B Although P4 could be detected, the defective portion P6 corresponding to the foreign substance F is influenced by the portion P5 corresponding to the adjacent blue stripe-shaped color pattern B, and is erroneously detected as a defective portion together with the portion P5. The portion P3 corresponding to the red striped color pattern R was also erroneously detected as a defective portion. As described above, when visible light is used, the decrease in the density value due to the color pattern is larger than the decrease in the density value due to the defect, and the defective portion is erroneously detected.

As described above, in the present embodiment, the red stripe-shaped color pattern R and the blue stripe-shaped color pattern B are used.
The density value does not decrease due to
It was possible to accurately detect the scratch D and the foreign matter F of a as defects.

Next, the result of detection of the collapse of the weave of the colored fabric (the state where the weave structure is locally collapsed) and the like will be described. FIG. 10 is a schematic diagram showing an example of the arrangement of the light source 1 and the camera 2 in the case of detecting the collapse of the fabric.

As the light source 1, the optical filter 32 and the CCD of the camera 2 shown in FIG. 10, a light emitting diode having the relative radiation intensity characteristic shown in FIG. 4, an optical filter having the transmittance characteristic shown in FIG. A CCD having the relative sensitivity characteristic shown in FIG. 6 was used. Further, the irradiation angle α of the light source 1 was set to 30 ° with respect to the horizontal plane of the woven fabric 100b in order to emphasize the warp yarn 101 and the weft yarn 102 which are the constituent elements of the woven fabric 100b.

When the fabric 100b illuminated by the light source 1 under the above illumination conditions is photographed by the camera 2, for example, the near-infrared ray 201 incident on the warp 101 to be inspected is included in the illumination light emitted from the light source 1. , Warp 10
1 is reflected by the near infrared ray 202 and is passed through the optical filter 32 and the camera lens 41 to the CCD of the camera 2.
Incident on.

On the other hand, of the illumination light emitted from the light source 1, for example, the near-infrared ray 203 that has entered the weft thread 102 that is not the inspection target is reflected by the weft thread 102 like the near-infrared ray 204 and does not enter the camera 2. As a result, the information on the color pattern of the woven fabric 100b can be completely eliminated, and the warp threads 101 and the weft yarns 1 constituting the woven fabric 100b can be eliminated.
It was possible to capture an image in which only the yarn component arranged on the entanglement point 02 was brightly emphasized.

The CPU 7 uses the grayscale image data obtained from the image photographed by the camera 2 as described above.
The tissue cycle of the warp 101 to be inspected was determined by. Specifically, the grayscale image data is subjected to density addition processing in the direction of the warp 101 to be inspected to convert it into one-dimensional density data, and this density data is subjected to FFT (Fast Fourier Trans).
The tissue cycle of the warp yarn 101 was obtained by obtaining the peak value of the spectrum obtained by performing the form treatment. The method for calculating the tissue cycle of the inspection target is not particularly limited to the above example, and various other methods can be used, for example,
The tissue cycle may be obtained by extracting the characteristics of the density waveform in the direction orthogonal to the yarn to be inspected.

Next, the warp 10 obtained by the CPU 7
Based on the tissue cycle of 1, a predetermined rectangular area was set in the captured image, and the statistical value of this rectangular area was calculated. Also,
The CPU 7 uses the calculated statistical values to compare the texture patterns of the fabric 100b in each rectangular area to extract defects of the fabric, and an image including the extracted defects is displayed on the image display 6.

FIG. 11 is a diagram showing an example of a rectangular area set in an image obtained by photographing a normal tissue portion of the fabric 100b. The warp 101R shown in FIG. 11 is a red yarn, the warp 101W is a white yarn, the warp 101B is a blue yarn,
A red color pattern is formed by the warp 101R, and the warp 101R
A blue color pattern is formed by B. This also applies to FIGS. 12 and 13 below.

When the normal tissue portion of the fabric 100b is photographed under the above illumination conditions, the warp yarns 101R, 10 are formed in the rectangular regions R1, R2 set in the photographed image.
The yarn component 103 (black square shown in FIG. 11) to be inspected, which is the yarn component of the warp 101R, 101W, 101B to be inspected on the entanglement point between 1W, 101B and the weft 102, is
It was detected as a highlighted and brightly photographed area.

In this case, each thread component 103 to be inspected is accurately detected without being affected by the red color pattern of the warp thread 101R and the blue color pattern of the warp thread 101B, and the thread to be inspected in each of the rectangular regions R1 and R2 is detected. When the patterns of the component 103 are compared with each other, the patterns match each other, and it is possible to detect that there is no defect in the woven fabric 100b by calculating the consistency of the pattern of the yarn component 103 to be inspected in each of the rectangular regions R1 and R2. I was able to.

FIG. 12 is a diagram showing an example of a rectangular area set in an image obtained by photographing a portion of the woven fabric 100b including a flow-in defect. When the portion including the flow-in defect of the woven fabric 100b is photographed as the weaving of the woven fabric 100b under the above illumination conditions, the yarn component 103 to be inspected (black square in FIG. 12) shown in FIG. It was

In this case, each thread component 103 to be inspected is accurately detected without being affected by the red color pattern by the warp thread 101R and the blue color pattern by the warp thread 101B. Since the pattern of the inspection target thread component 103 in each of the rectangular regions R3 and R4 does not match, the fabric 100b is calculated by calculating the consistency of the pattern of the inspection target yarn component 103 in each of the rectangular regions R3 and R4. It was possible to detect that there was a pouring defect.

FIG. 13 is a diagram showing an example of a rectangular area set in an image obtained by photographing a normal tissue portion of the fabric 100b using visible light as in the conventional defect detecting apparatus.
When the normal tissue portion of the woven fabric 100b was photographed using visible light, the yarn component 103 to be inspected (black square in FIG. 13) shown in FIG. 13 was detected in the rectangular regions R1 and R2.

In this case, although it was possible to detect the yarn component 103 to be inspected of the white warp 101W, the warp 101R
By the red color pattern by and the blue color pattern by the warp thread 101B, the inspection target thread component on the confounding point that should be brightly photographed by illumination is imaged darkly, and the inspection target thread components of the red warp thread 101R and the blue warp thread 101B are detected. It could not be detected.

As a result, when the patterns of the thread component 103 to be inspected in the respective rectangular areas R1 and R2 are compared, the patterns of both do not match, and the consistency of the pattern of the thread component 103 to be inspected in the respective rectangular areas R1 and R2 is confirmed. As a result of the calculation, the fabric 100
Although there is no defect in b, the inspected yarn component of the red warp 101R and the blue warp 101B, which could not be detected, was erroneously detected as a defect.

As described above, in this embodiment, the warp 1
The yarn component 103 to be inspected in each of the rectangular regions R1 and R2 in the normal tissue portion of the fabric 100b can be accurately extracted without being affected by the red color pattern of 01R and the blue color pattern of the warp yarn 101B. In addition to being able to accurately detect the absence of defects, the fabric 100b
Rectangular regions R3, R in the portion including the inflow defect of
It was possible to accurately extract the yarn component 103 to be inspected of No. 4 and to accurately detect the flow-in defect of the woven fabric 100b.

Although the light reflected on the surface of the inspection object is used in the above description, the present invention can be similarly applied to the case where the light transmitted through the inspection object is used. The effect of can be obtained.

[0062]

According to the present invention, a defect of an inspection object is detected using an image picked up by receiving only light having a wavelength of 780 nm or more reflected or transmitted on the inspection object surface of the inspection object. Therefore, it is possible to eliminate the light in the visible light range to remove the color pattern information of the inspection object, and to take an image of the inspection object that reflects only the shadow information due to physical unevenness. It is possible to detect a defect of an inspection object having a color pattern with high accuracy without being affected by the pattern.

[Brief description of drawings]

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

FIG. 2 is a perspective view showing a specific example of the light source shown in FIG.

FIG. 3 is a perspective view showing another specific example of the light source and the camera shown in FIG.

4 is a diagram showing a relative radiation intensity characteristic of a light emitting diode used as the light source shown in FIG.

5 is a diagram showing a transmittance characteristic of the optical filter shown in FIG.

6 is a diagram showing a relative sensitivity characteristic of a CCD used in the camera shown in FIG.

FIG. 7 is a diagram schematically showing a state of a surface of a print sheet which is an inspection object.

8 is a diagram showing waveforms of density values of grayscale image data created from an image obtained by photographing the print sheet shown in FIG. 7 using the light emitting diode having the characteristics shown in FIGS. 4 to 6;

9 is a diagram showing a waveform of a density value of grayscale image data created from an image obtained by photographing the print sheet shown in FIG. 7 using visible light as in the conventional defect detection apparatus.

FIG. 10 is a schematic diagram showing an example of the arrangement of a light source and a camera when detecting the weaving of the fabric.

FIG. 11 is a diagram showing an example of a rectangular region set in an image obtained by capturing an image of a normal tissue part of a fabric.

FIG. 12 is a diagram showing an example of a rectangular area set in an image obtained by capturing an image of a portion of a fabric including a flow-in defect.

FIG. 13 is a diagram showing an example of a rectangular area set in an image obtained by photographing a normal tissue portion of a fabric using visible light as in the conventional defect detection device.

[Explanation of symbols]

1 light source 2 camera 3 A / D converter 4 image memory 5 Image processing circuit 6 Image display 7 CPU 8 ROM 9 RAM 10 image bus 11 CPU bus 21 light emitting diode 32 Optical filter 41 camera lens

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁷ Identification symbol FI theme code (reference) G06T 1/00 400 G06T 1/00 400J 420 420D (72) Inventor Toru Sasaki 2-chome Katata, Otsu City, Shiga Prefecture No. 1 Toyobo Co., Ltd. Research Institute (72) Inventor Kenjiro Ueda 2-1-1 Katata, Otsu City, Shiga Prefecture Toyobo Co., Ltd. Research Institute F-term (reference) 2F065 AA49 BB13 CC00 CC02 DD03 FF04 GG06 GG14 GG22 HH12 JJ03 JJ09 JJ26 LL22 2G051 AA34 AA40 AB01 AB06 AB07 AB11 BA06 CA03 CB01 CC07 DA01 EA11 EA12 EB01 EC02 EC04 EC06 3B154 AB20 BA53 BB18 BC42 BF21 CA16 CA23 DA30 5B047 AA11 AB02 BB18 DB05 BC02 BC02 BC07 BC11 BC02

Claims (3)

[Claims] 1. A defect inspection apparatus for inspecting defects of an inspection object having a color pattern, which receives only light having a wavelength of 780 nm or more reflected or transmitted on the inspection object surface of the inspection object. A defect inspection apparatus comprising: an image pickup unit that picks up an image of an inspection target surface; and a detection unit that detects a defect of the inspection target by using the image picked up by the image pickup unit. 2. The defect according to claim 1, wherein the imaging unit receives only near-infrared rays reflected or transmitted by the inspection target surface of the inspection target object to capture an image of the inspection target surface. Inspection device. 3. The image pickup unit includes a light projecting unit that irradiates the inspection target surface of the inspection target object with the infrared light, and the wavelength of the near infrared light irradiated by the light projecting unit is 8
3. The defect inspection device according to claim 1, wherein the defect inspection device has a length of 00 nm or more and 1100 nm or less.