JP2006266934A - Method and apparatus for detecting defect in film - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method and an apparatus for detecting defects in a film capable of detecting the defects in the film with inexpensive and simple configulation. <P>SOLUTION: A parallel light forming means 4 converts a light 2a emitted from an illumination means 2 into an almost parallel luminous flux flux whose diffusion is suppressed in the Z direction orthogonal to the direction of the emitted light 2a. Since the light as an epi-illumination light 2b is radiated to a surface of the film 1 by a half mirror 5, the light is almost parallel in the Y direction of the film 1, diffused in the direction orthogonal to the Y direction, and radiated to the surface of the film 1. An intensity distribution of an original image as an imaged reflection light 2c is calculated. The defects are detected on the surface of the film 1 on the basis of the calculated intensity distribution. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a defect detection method and a defect detection apparatus for detecting a film defect, in particular, a concave defect (crater-like surface defect) present on the surface of the film.

  Conventionally, for example, as a technique for detecting defects generated on the surface of a manufactured film, light is irradiated from above the film surface, and reflected light reflected on the film surface is imaged by an imaging device such as a CCD camera sensor. A method for capturing an image of a film and detecting defects on the film surface based on the image is known.

  Here, as such a defect detection method for a film, a method using only one detection optical system is known, but in a method using only one detection optical system, a defect on the film surface is determined based on one image. As a result, it may be difficult to detect defects that occur on the film surface, or it may be difficult to distinguish the defects from foreign substances such as particles existing inside the film, and sufficient detection accuracy is obtained. There was a problem that it was not possible. In order to improve the detection accuracy of defects, an image having a large change in brightness (image brightness) of reflected light of a defective portion on the film surface and reflected light of a portion without a defect around the film surface, that is, Although it is necessary to obtain an image with high contrast, it is generally considered that it is difficult to obtain an image with high contrast in a method for detecting defects on the film surface based on one image.

Therefore, a method of accurately detecting a film defect using two detection optical systems is disclosed. More specifically, the surface of the film was irradiated with light from an angle within a predetermined range, the reflected light was imaged, and the light was irradiated from above the film surface, and the reflected light was imaged from above. A method for detecting a film surface defect based on an image has been proposed. By using this method, it becomes possible to detect defects on the film surface based on a plurality of images for the same part of the film surface. Therefore, the method for detecting defects on the film surface based on one image is used. In comparison, it is described that the detection accuracy can be improved (for example, see Patent Document 1). In addition, there has been proposed a defect detection method for a film surface that includes two detection optical systems of oblique illumination and epi-illumination and that can select an irradiation angle and wavelength of irradiation light. By using this method, it is possible to detect the density difference between the defect and the foreign matter such as particles existing in the film from the imaged image, so that it is possible to distinguish between the defect and the foreign matter, and the detection accuracy is improved. It is described that it can be increased (for example, see Patent Document 2).
JP 2001-201445 A JP 2003-215051 A

  Here, in the conventional film defect detection method, the film defect can be detected. However, as described above, since the structure using two detection optical systems is used, the detection system becomes complicated. There is a problem that the cost of the defect detection apparatus becomes high.

  Therefore, the present invention has been made in view of the above-described problems, and an object thereof is to provide a film defect detection method and a defect detection apparatus that can detect a film defect with an inexpensive and simple configuration.

  In order to achieve the above object, the invention described in claim 1 is a film defect detection method, wherein the film surface is substantially parallel to the film surface in one direction and orthogonal to the film direction. In this direction, diffusing light is irradiated, and defects on the surface of the film are detected based on a luminance distribution in one direction of the film of the original image in which the reflected light is imaged. In addition, the "perpendicular direction" here means a direction orthogonal in the plane of the film irradiated with light.

  According to the configuration of the first aspect, in the luminance distribution in one direction of the film, the difference between the luminance of the defective portion and the luminance of the peripheral portion having no defect becomes large. In addition, the difference between the luminance of the defective portion and the luminance of foreign matters such as particles existing inside the film also increases. Therefore, even when only one detection optical system is used, an image with a high contrast can be obtained when the reflected light is imaged. The generated defect can be detected and the defect can be distinguished from the foreign matter.

  The invention according to claim 2 is the film defect detection method according to claim 1, wherein the film is detected by detecting that the luminance distribution of the original image in one direction of the film has two minimum values. It is characterized by detecting defects on the surface.

  According to the configuration of the second aspect, in particular, when there is a concave defect on the surface of the film, the concave defect can be detected based on the luminance distribution of the original image having two dark portions specific to the concave defect. It becomes possible. Therefore, the detection accuracy of the concave defect is improved, and the discrimination accuracy between the concave defect and the foreign matter is improved.

  The invention according to claim 3 is a defect detection method for a film, wherein the light is diffused in a direction substantially parallel to one direction of the film and perpendicular to one direction of the film with respect to the surface of the film. A defect on the surface of the film is detected on the basis of a processed image obtained by performing image processing on an original image that is irradiated and the reflected light is captured.

  According to the configuration of the third aspect, in the captured original image, the difference between the luminance of the defective portion in one direction of the film and the luminance of the peripheral portion having no defect increases. In addition, the difference between the luminance of the defective portion and the luminance of foreign matters such as particles existing inside the film also increases. Therefore, even when only one detection optical system is used, an image with a high contrast can be obtained when the reflected light is imaged. The generated defect can be detected and the defect can be distinguished from the foreign matter. In addition, by performing image processing on the captured original image, it becomes possible to obtain a processed image in which a defective portion is revealed compared to the original image. Compared with the case where a defect is detected based on this, the defect detection accuracy is improved and the discrimination accuracy between the defect and the foreign matter is improved.

  The invention according to claim 4 is the film defect detection method according to claim 3, wherein it is detected that the brightness distribution in one direction of the film of the processed image has two minimum values, and It is characterized by detecting defects on the surface.

  According to the configuration of claim 4, in particular, when there is a concave defect on the surface of the film, the concave defect can be detected based on the luminance distribution of the processed image having two dark portions specific to the concave defect. It becomes possible. Therefore, the detection accuracy of the concave defect is improved, and the discrimination accuracy between the concave defect and the foreign matter is improved.

  Note that the image processing of the present invention may be any processing as long as the defect portion in the captured original image can be revealed, and as described in claim 5, claims 3 or 4 The film defect detection method according to claim 1, wherein a defect on the surface of the film is detected based on a luminance distribution in one direction of the film of the expansion / contraction image obtained by performing expansion / contraction processing on the original image. It is good as a method.

  According to the configuration described in claim 5, even when only one detection optical system is used, an image with high contrast can be obtained when the reflected light is imaged. In addition, with a simple configuration, it is possible to detect a defect generated on the film surface and to discriminate between a defect and a foreign substance. In addition, by performing expansion / contraction processing on the captured original image, it is possible to obtain an expansion / contraction image in which a defective portion is revealed compared to the original image. Compared with the case where a defect is detected based on the original image, the defect detection accuracy is improved and the discrimination accuracy between the defect and the foreign matter is improved. In particular, when there is a concave defect on the surface of the film, it becomes possible to detect the concave defect based on the luminance distribution of the expansion / contraction image having two dark portions peculiar to the concave defect. Therefore, the detection accuracy of the concave defect is improved, and the discrimination accuracy between the concave defect and the foreign matter is improved.

  Further, as a method for detecting defects on the surface of the film based on a processed image obtained by performing image processing, the method for detecting defects in a film according to claim 3 may be used. Then, the original image or the original image is subjected to expansion / contraction processing, and the obtained expansion / contraction image is binarized to obtain defects on the film surface based on the image pattern of the binarized image. A method of detecting can be used.

  According to the configuration of the sixth aspect, in the captured original image, the difference between the luminance of the defective portion in one direction of the film and the luminance of the peripheral portion having no defect increases. In addition, the difference between the luminance of the defective portion and the luminance of foreign matters such as particles existing inside the film also increases. Therefore, even when only one detection optical system is used, an image with a high contrast can be obtained when the reflected light is imaged. The generated defect can be detected and the defect can be distinguished from the foreign matter. In addition, by performing image processing on the captured original image, it is possible to obtain a binarized image in which a defective portion is revealed compared to the original image. Therefore, the original image before image processing is performed. Compared to the case where a defect is detected based on an image, the defect detection accuracy is improved and the discrimination accuracy between the defect and the foreign matter is improved.

  In addition, when detecting defects on the surface of the film while transporting the film, as described in claim 7, the defect detection method for film according to claim 1, wherein the defect is detected on the surface of the film to be transported. On the other hand, it is possible to adopt a method of irradiating diffused light in a direction substantially parallel to the film transport direction and orthogonal to the transport direction.

  According to the configuration of the seventh aspect, in the luminance distribution in the film transport direction, the difference between the luminance of the defective portion and the luminance of the peripheral portion having no defect becomes large. In addition, the difference between the luminance of the defective portion and the luminance of foreign matters such as particles existing inside the film also increases. Therefore, even when only one detection optical system is used, an image with a high contrast can be obtained when the reflected light is imaged. The generated defect can be detected and the defect can be distinguished from the foreign matter.

  The invention according to claim 8 is an apparatus for detecting a defect in a film, wherein illumination means for emitting light irradiated on the surface of the film and light irradiated on the surface of the film are substantially omitted in one direction of the film. Parallel light forming means that is parallel and that diffuses light in a direction perpendicular to one direction of the film, imaging means that captures reflected light reflected on the surface of the film, and an original image captured by the imaging means Computational processing means for computing the luminance distribution in one direction of the film and defect detection means for detecting defects on the surface of the film based on the luminance distribution.

  According to the structure of Claim 8, it becomes possible to irradiate the surface of a film with the substantially parallel light substantially parallel in one direction of a film. Therefore, in the luminance distribution in one direction of the film, the difference between the luminance of the defective portion and the luminance of the peripheral portion having no defect becomes large. In addition, the difference between the luminance of the defective portion and the luminance of foreign matters such as particles existing inside the film also increases. Therefore, even when only one detection optical system is used, an image with a high contrast can be obtained when the reflected light is imaged. The generated defect can be detected and the defect can be distinguished from the foreign matter.

  The invention according to claim 9 is a film defect detection device, wherein the illumination means for emitting light applied to the surface of the film and the light applied to the surface of the film are substantially omitted in one direction of the film. Parallel light forming means that is parallel and that diffuses light in a direction perpendicular to one direction of the film, imaging means that captures reflected light reflected on the surface of the film, and an original image captured by the imaging means Image processing means for performing image processing and generating a processed image; arithmetic processing means for calculating a luminance distribution in one direction of the processed image; and defect detection for detecting defects on the surface of the film based on the luminance distribution And means.

  According to the structure of Claim 9, it becomes possible to irradiate the surface of a film with the substantially parallel light substantially parallel in one direction of a film. Therefore, in the luminance distribution in one direction of the film, the difference between the luminance of the defective portion and the luminance of the peripheral portion having no defect becomes large. In addition, the difference between the luminance of the defective portion and the luminance of foreign matters such as particles existing inside the film also increases. Therefore, even when only one detection optical system is used, an image with a high contrast can be obtained when the reflected light is imaged. The generated defect can be detected and the defect can be distinguished from the foreign matter. In addition, by performing image processing on the captured original image, it becomes possible to obtain a processed image in which a defective portion is revealed compared to the original image. Compared with the case where a defect is detected based on this, the defect detection accuracy is improved and the discrimination accuracy between the defect and the foreign matter is improved.

  The invention according to claim 10 is a film defect detection device, wherein the illumination means for emitting light irradiated on the surface of the film and the light irradiated on the surface of the film are substantially omitted in one direction of the film. Parallel light forming means that is parallel and that diffuses light in a direction perpendicular to one direction of the film, imaging means that captures reflected light reflected on the surface of the film, and an original image captured by the imaging means And image processing means for generating a binarized image and defect detection means for detecting defects on the surface of the film based on the image pattern of the binarized image.

  According to the structure of Claim 10, it becomes possible to irradiate the surface of a film with the substantially parallel light substantially parallel in one direction of a film. Therefore, in the luminance distribution in one direction of the film, the difference between the luminance of the defective portion and the luminance of the peripheral portion having no defect becomes large. In addition, the difference between the luminance of the defective portion and the luminance of foreign matters such as particles existing inside the film also increases. Therefore, even when only one detection optical system is used, an image with a high contrast can be obtained when the reflected light is imaged. The generated defect can be detected and the defect can be distinguished from the foreign matter. In addition, by performing image processing on the captured original image, it is possible to obtain a binarized image in which a defective portion is revealed compared to the original image. Therefore, the original image before image processing is performed. Compared to the case where a defect is detected based on an image, the defect detection accuracy is improved and the discrimination accuracy between the defect and the foreign matter is improved.

  The parallel light forming means of the present invention makes the light irradiated on the surface of the film into light that is substantially parallel in one direction of the film and diffuses in a direction perpendicular to one direction of the film. Any material can be used, and a cylindrical lens or a linear Fresnel lens can be used as described in claim 11.

  According to the configuration of the eleventh aspect, a general-purpose optical member that is inexpensive and easily available can be used as the parallel light forming means.

  According to the present invention, it is possible to detect a defect generated on the film surface with a low-cost and simple configuration as compared with the prior art, and to distinguish between a defect and a foreign substance.

  Hereinafter, specific embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a schematic diagram showing the overall configuration of a defect detection apparatus for carrying out the film defect detection method of the present invention. This defect inspection apparatus detects a defect on the surface of the film 1 and includes one detection optical system. The detection optical system is configured to irradiate the surface of the film 1 with light, capture the reflected light 2c reflected on the surface of the film 1 with a CCD camera sensor 6 as an imaging means, and capture an image of the film 1. It has become.

  The light 2a emitted by the illuminating means 2 is guided toward the half mirror 5 by the light guide 3 arranged on the optical path of the light 2a, and then the coaxial epi-illumination (hereinafter referred to as “epi-illumination”). The surface of the film 1 is irradiated as 2b. Then, the reflected light 2 c reflected on the surface of the film 1 is imaged by the CCD camera sensor 6 having a light receiving surface in a direction orthogonal to the surface of the film 1.

  In this embodiment, a general adhesive film, for example, a film 1 produced by applying and drying an adhesive on a base material (PET film or the like) is an inspection target, and the film is formed on the base material 40. 1 is formed as an inspection object 41. Then, while continuously inspecting the inspection object 41 from the unwinding roll R1 to the take-up roll R2, or while stopping the inspection object 41 and stopping the film 1, the surface of the film 1 In this configuration, the surface of the film 1 is detected by picking up an image of the reflected light 2 c reflected by the CCD camera sensor 6.

  Here, in this embodiment, one direction of the film 1 with respect to the surface of the film 1 (the length direction of the film 1, and when the film is transported, the transport direction of the film 1. The arrow Y shown in FIG. In the direction perpendicular to one direction Y (the width direction of the film 1 and the direction perpendicular to the transport direction of the film 1. The direction of the arrow X shown in FIG. 5 described later). Is characterized by irradiating diffuse light. In addition, the orthogonal direction X means the direction orthogonal in the surface of the film 1 to which light is irradiated.

  More specifically, as shown in FIG. 1, there is provided parallel light forming means 4 for suppressing the diffusion (spreading) of the light 2a emitted by the illumination means 2 and forming the light 2a into substantially parallel light. ing. In the parallel light forming means 4, for example, a cylindrical lens having a certain curvature and a plurality of grooves are formed in a direction orthogonal to the emission direction of the light 2a (the direction of the arrow Z in the figure), and a constant refraction is achieved. A general-purpose optical member that is inexpensive and easily available, such as a linear Fresnel lens having power, can be used.

  When the light 2a emitted from the illuminating means 2 and guided by the light guide 3 is incident on the parallel light forming means 4, the parallel light forming means 4 collects the diffusing light 2a, and the emission direction of the light 2a. In order to suppress the diffusion of the light 2a in the orthogonal direction Z, the light 2a is changed into a substantially parallel light beam in the Z direction. By irradiating the surface of the film 1 with this light as the epi-illumination 2b by the half mirror 5, the light is diffused in the direction X which is substantially parallel in the one direction Y of the film 1 and orthogonal to the one direction Y. The surface of the film 1 is irradiated with light (hereinafter referred to as “substantially parallel light”). That is, in this embodiment, the light irradiated on the surface of the film 1 by the parallel light forming unit 4 is configured to be the above-described substantially parallel light.

  Next, the reflection mechanism on the surface of the film 1 will be described with reference to the drawings. FIG. 2 is a diagram for explaining a reflection mechanism when light that is not substantially parallel in one direction of the film is used. FIG. 3 is a reflection mechanism when light that is substantially parallel is applied in one direction of the film. It is a figure for demonstrating. FIG. 4 is a diagram showing an original image obtained by capturing reflected light when light that is not substantially parallel in one direction of the film is irradiated. FIG. 5 is a diagram in which light substantially parallel in one direction of the film is irradiated. It is a figure which shows the original image which imaged the reflected light in the case.

  For example, when detecting the concave defect 50 generated on the surface of the film 1, when the surface of the film 1 is irradiated with the light 2 a diffusing in one direction Y of the film 1 as the epi-illumination 2 b, it is not substantially parallel in the one direction Y. The light 2d is irradiated on the surface of the film 1. Then, as shown in FIG. 2, the light 2 d irradiated on the surface of the film 1 is reflected at the concave defect 50 portion on the surface of the film 1 in the same manner as the defect-free portion 51 around the concave defect 50. Therefore, the difference between the luminance of the reflected light 2c at the concave defect 50 portion and the luminance of the reflected light 2c at the peripheral portion 51 having no defect becomes small. Therefore, when the reflected light 2c is imaged, as shown in FIG. 4, only a low contrast image in which the portion 54 corresponding to the concave defect 50 cannot be clearly recognized can be obtained, and the concave generated on the surface of the film 1 can be obtained. It becomes difficult to detect the defect 50 sufficiently.

  On the other hand, when the surface of the film 1 is irradiated with substantially parallel light 2e generated by the parallel light forming means 4 and substantially parallel in one direction Y of the film 1, as shown in FIG. Since the intensity of the incident reflected light 2c at the concave defect 50 portion is smaller than the intensity of the reflected light 2c at the defect-free portion 51, the brightness of the reflected light 2c at the concave defect 50 portion and the surrounding defects The difference in the brightness of the reflected light 2c of the portion 51 having no gap increases. Therefore, when the reflected light 2c is imaged, a high-contrast image capable of recognizing the concave defect 50 as the dark portion 52 can be obtained as shown in FIG. 50 can be detected.

  In addition, since the bottom 50a of the concave defect 50 generally has a flat shape or a shape close to flat, as shown in FIG. 3, the incident light 2b has a defect-free portion 51 in the bottom 50a. Similarly reflected. Therefore, when the reflected light 2c is imaged, the bottom 50a becomes a bright portion in the same manner as the non-defect portion 51. Therefore, as shown in FIG. 5, an image unique to the concave defect 50, that is, two dark portions. An image having 52 is obtained.

  The original image captured by the CCD camera sensor 6 is converted from an analog signal to a digital signal by the A / D conversion unit 7 and stored in the image memory unit 8. The image data stored in the image memory unit 8 is input to the arithmetic processing unit 9, and the arithmetic processing unit 9 calculates the luminance distribution of the captured original image based on the input image data. Further, the defect detection unit 10 detects a defect on the surface of the film 1 based on the calculation result regarding the luminance of the calculation processing unit 9.

  Next, a specific defect detection method for the film 1 will be described with reference to the drawings. FIG. 6 is a diagram showing the luminance distribution of the image in one direction of the film and showing the luminance distribution in the aa cross section of the image shown in FIG. FIG. 7 is a diagram showing the luminance distribution of the image in one direction of the film and showing the luminance distribution in the AA cross section of the image shown in FIG. FIG. 8 is a luminance distribution in a direction perpendicular to one direction of the film, and is a diagram showing the luminance distribution in the BB cross section of the image shown in FIG.

  As described above, when the surface 2 of the film 1 is irradiated with the light 2d that is not substantially parallel in the one direction Y of the film 1, as shown in FIG. 6, the luminance 20 of the concave defect 50 portion and the surrounding defect-free portion Since the difference of the luminance 21 of 51 becomes small, even if it is a case where the determination threshold value 22 is preset, it becomes difficult to detect the concave defect 50. Further, since the difference between the luminance 20 of the concave defect 50 and the luminance 23 of the foreign matter such as particles existing in the film 1 is also small, it is difficult to distinguish the concave defect 50 from the foreign matter.

  On the other hand, when the surface of the film 1 is irradiated with substantially parallel light 2e substantially parallel in one direction Y of the film 1, as shown in FIG. 7, the luminance 24 of the concave defect 50 portion and the surrounding defect-free portion 51 are obtained. The difference of the luminance 21 of is increased. If it does so, it will become possible to detect the concave defect 50 by the determination by the predetermined determination threshold value 22. Further, since the difference between the luminance 24 of the concave defect 50 and the luminance 23 of the foreign matter such as particles existing in the film 1 is also increased, it is possible to distinguish the concave defect 50 from the foreign matter.

  That is, even when only one detection optical system is used, an image with a high contrast can be obtained when the reflected light 2c is imaged. It is possible to detect the recessed defect 50 generated on the surface of the surface and to distinguish the recessed defect 50 from the foreign matter.

  In the present embodiment, the concave defect on the surface of the film 1 can be detected by using only the luminance distribution in the one direction Y of the film 1 of the captured original image. However, the original image in the one direction Y can be detected. And the luminance distribution of the original image in the direction X orthogonal to the one direction Y shown in FIG. In this case, the determination threshold 25 is also set in advance in the luminance distribution of the original image in the X direction, and the luminance 26 of the concave defect 50 in the X direction and the determination threshold 25 are compared. That is, the concave defect 50 is detected by determination based on both the determination threshold values 22 and 25. In this way, the concave defect 50 can be easily detected.

  Further, as described above, when the substantially parallel light 2e substantially parallel in one direction Y is irradiated to the concave defect 50 formed on the surface of the film 1, the two dark portions 52 unique to the concave defect 50 shown in FIG. Since the original image is obtained, the portion corresponding to the dark portion 52 appears as two minimum values 27 in the luminance distribution in one direction Y as shown in FIG. Therefore, when the original image having the two dark portions 52 is obtained, but the determination threshold 22 does not determine that it is the concave defect 50, it is detected that the luminance distribution in one direction Y has the two minimum values 27. Thus, the concave defect 50 can be detected. As a result, the detection accuracy of the concave defect 50 is improved, and the discrimination accuracy between the concave defect 50 and the foreign matter is improved.

  Further, in order to make the original image picked up by the CCD camera sensor 6 appear, the image processing means 11 performs image processing such as expansion / contraction processing and binarization processing on the original image, and performs the image processing. It can also be set as the structure which detects the defect of the surface of the film 1 based on the process image obtained by this. In this case, a defect on the surface of the film 1 is detected based on the above-described processed image in which the defective portion in the original image is made apparent. Hereinafter, an example of the expansion / contraction process 12 will be described.

  In general, expansion processing refers to processing for setting a pixel to 1 if there is even one (for example, white) in the vicinity (for example, 4 or 8) of a pixel. If there is even one 0 (for example, black) in the vicinity of a certain pixel, this is processing for setting that pixel to zero. In the present embodiment, an expansion image is generated by removing foreign substances such as particles in the original image by performing expansion processing on the original image captured by the CCD camera sensor 6, and then the expansion image By performing the contraction process on the image, an expansion / contraction image in which the portion of the remaining concave defect 50 is clear is generated. Then, the image processing signal 13 after the expansion / contraction processing 12 is input to the arithmetic processing means 9, and the arithmetic processing means 9 is based on the input image processing signal 13 and the film 1 of the expansion / contraction image which is a processed image. The luminance distribution in one direction Y is calculated. And the defect detection means 10 determines the presence or absence of the defect of the surface of the film 1 based on the said luminance distribution.

  FIG. 9 shows an image obtained by expanding and contracting the original image shown in FIG. By the expansion / contraction process, two dark parts 52 peculiar to the concave defect 50 are revealed in the processed image as compared with the original image shown in FIG.

  That is, as described above, even when only one detection optical system is used by irradiating the surface of the film 1 with the substantially parallel light 2e that is substantially parallel in one direction Y of the film 1, the reflected light 2c is generated. When picked up, an original image with high contrast can be obtained. Then, by performing image processing on the original image, it is possible to obtain a processed image in which a portion of the concave defect 50 is revealed as compared with the original image, and a defect on the surface of the film 1 is detected based on the processed image. It is configured. Therefore, even when only one detection optical system is used, it is possible to detect the concave defect 50 generated on the surface of the film 1 with a cheaper and simpler configuration as compared with the conventional technique, and the concave In addition to the effect that the defect 50 can be distinguished from the foreign matter, the detection accuracy of the concave defect 50 is improved as compared with the case where the concave defect 50 is detected based on the original image before image processing is performed. The discrimination accuracy between the concave defect 50 and the foreign matter is improved.

  Further, as described above, since an original image having two dark portions 52 peculiar to the concave defect 50 is obtained, image processing such as expansion / contraction processing 12 is performed on the original image, and a processed image in one direction Y of the film 1 is obtained. It is also possible to adopt a configuration in which the concave defect 50 is detected by detecting that the luminance distribution has two minimum values 27. Also in this case, the detection accuracy of the concave defect 50 is improved, and the discrimination accuracy between the concave defect 50 and the foreign matter is improved.

  FIG. 10 is a flowchart showing the procedure of the film defect detection method of the present invention. First, the surface of the film 1 is irradiated with substantially parallel light 2e substantially parallel in one direction Y of the film 1 (step S1), and the reflected light 2c is imaged by the CCD camera sensor 6 (step S2). Then, the original image captured by the CCD camera sensor 6 is stored in the image memory unit 8 (step S3). Next, the image data stored in the image memory unit 8 is input to the arithmetic processing means 9 directly or via image processing such as expansion / contraction processing 12 by the image processing means 11 (step S4), and arithmetic processing is performed. The means 9 calculates the luminance distribution based on the input image data (step S5). Next, the calculation result by the calculation processing means 9 is input to the defect detection means 10 (step S6). The defect detection means 10 determines whether or not there is a portion having a luminance smaller than a predetermined determination threshold 22 in the luminance distribution of the original image or processed image in one direction Y of the film 1 (step S7). ), If there is a portion having a luminance smaller than the determination threshold 22, it is determined that the concave defect 50 exists on the surface of the film 1 (step S8). On the other hand, in the luminance distribution of the original image or the processed image in one direction Y of the film 1, when there is no portion having a luminance smaller than the determination threshold 22, the defect detection means 10 has two luminance distributions in the one direction Y. It is determined whether or not the minimum value 27 is present (step S9). If two minimum values 27 are detected in the luminance distribution of the original image or processed image in one direction Y, it is determined that the concave defect 50 exists on the surface of the film 1 (step S10). On the other hand, when the two minimum values 27 are not detected in the luminance distribution of the original image or the processed image in one direction Y, it is determined that the concave defect 50 does not exist on the surface of the film 1 (step S11).

  As described above, in the present invention, the light 2 a emitted from the illumination unit 2 is substantially parallel in the one direction Y of the film 1 by the parallel light forming unit 4 and is in the one direction Y of the film 1. In the orthogonal direction X, the light is diffused, and the surface of the film 1 is irradiated with substantially parallel light 2e substantially parallel in one direction Y of the film 1. Then, the reflected light 2c reflected on the surface of the film 1 is imaged by the CCD camera sensor 6, and the luminance distribution in the one direction Y of the film 1 is calculated by the arithmetic processing means 9 to detect the defect. The means 10 detects a defect on the surface of the film 1 based on the calculated luminance distribution. Accordingly, in the luminance distribution in one direction Y of the film 1, the difference between the luminance 24 of the concave defect 50 portion and the luminance 21 of the surrounding defect-free portion 51 is increased, and the luminance 24 of the concave defect 50 portion is Also, the difference from the brightness 23 of foreign matter such as particles existing in the film 1 is increased. As a result, even when only one detection optical system is used, an image with high contrast can be obtained when the reflected light is imaged. Therefore, the film 1 has a lower cost and simpler configuration than the prior art. It is possible to detect the recessed defect 50 generated on the surface of the surface and to distinguish the recessed defect 50 from the foreign matter.

  Further, the processing unit 9 calculates the luminance distribution in one direction Y of the film 1 of the processed image obtained by performing image processing on the original image captured by the CCD camera sensor 6 by the image processing unit 11. The defect detection means 10 detects defects on the surface of the film 1 based on the calculated luminance distribution. Therefore, in order to detect a defect on the surface of the film 1 based on the processed image in which the defective portion in the captured original image is made visible, the concave defect 50 is detected based on the original image before image processing is performed. Compared to the above, the detection accuracy of the concave defect 50 is improved, and the discrimination accuracy between the concave defect 50 and the foreign matter is improved.

  Further, as described above, the light 2 a emitted from the illumination unit 2 is substantially parallel in the one direction Y of the film 1 by the parallel light forming unit 4 and in the direction X orthogonal to the one direction Y of the film 1. Since the light is diffused and the surface of the film 1 is irradiated with substantially parallel light 2e that is substantially parallel in one direction Y of the film 1, the concave defect 50 is present particularly when the concave defect 50 exists on the surface of the film 1. An original image or a processed image having two dark portions 52 peculiar to 50 can be obtained. Therefore, it becomes possible to detect the concave defect 50 on the surface of the film 1 by detecting that the luminance distribution of the original image or the processed image in the one direction Y of the film 1 has two minimum values 27, and as a result. The detection accuracy of the concave defect 50 is improved, and the discrimination accuracy between the concave defect 50 and the foreign matter is improved.

In addition, this invention is not limited to the above-mentioned embodiment, A various deformation | transformation is possible.
For example, in the above-described embodiment, the reflected light 2c reflected on the surface of the film 1 is imaged by the CCD camera sensor 6 using the epi-illumination 2b, but is substantially parallel in one direction Y of the film 1. Any configuration may be used as long as the surface of the film 1 can be irradiated with substantially parallel light 2e. For example, as shown in FIG. 11, the substantially parallel light 2 e emitted from the illumination unit 2 and generated by the parallel light formation unit 4 is converted into oblique illumination 2 f at a preset angle with respect to the surface of the film 1. It is good also as a structure which irradiates the surface of the film 1 and images the reflected light reflected on the surface of the film 1 with the CCD camera sensor 6.

  Further, the surface of the film 1 is irradiated with substantially parallel light 2e that is substantially parallel in one direction Y of the film 1, and the reflected light 2c is imaged, or an expansion / contraction process 12 is performed on the original image. The defect on the surface of the film 1 can also be detected based on the image pattern of the binarized image obtained by binarizing the obtained expansion / contraction image. One example will be described below.

  FIG. 12 is a schematic diagram showing the overall configuration of another defect detection apparatus for carrying out the film defect detection method of the present invention. In this case, first, the image processing unit 11 performs the above-described expansion / contraction processing 12 on the captured original image or the original image, and the binarization processing 14 is performed on the obtained expansion / contraction image. Is called. Examples of the binarization processing 14 include a method of processing pixels having a brightness greater than a determination threshold set in advance with respect to the brightness of the image in white, and pixels having a small brightness in black. Next, an image processing signal 13 based on the binarization process 14 is input to the defect detection unit 10, and the defect detection unit 10 compares the binarized image with an image stored in advance in the storage unit 15. The presence or absence of defects on the surface of the film 1 is determined. That is, if the shape pattern of the portion recognized as a defect in the binarized image generated by the image processing unit 11 matches the pattern of the defect shape of the binarized image stored in the storage unit 15 in advance, the film 1 It is determined that there is a defect on the surface.

  FIG. 13 is an image obtained by inverting and outputting the result of the binarization process 14 after the above-described expansion / contraction process 12 is performed on the original image captured by the CCD camera sensor 6. In other words, as a result of comparison with the determination threshold value set in advance with respect to the brightness of the image, a pixel determined to be white is processed as a black pixel, and a pixel determined as black is processed as a white pixel. By the binarization process 14, the two dark parts 52 peculiar to the concave defect 50 are further manifested as white parts 55. In this case, the defect detection means 10 has the shape pattern of the white part 55 corresponding to the concave defect 50 in the binarized image shown in FIG. If it matches the shape pattern of the part, it is determined that the concave defect 50 exists on the surface of the film 1. Even in such a method, even when only one detection optical system is used, an image with high contrast can be obtained when reflected light is imaged. Thus, the concave defect 50 generated on the surface of the film 1 can be detected, and the concave defect 50 can be distinguished from the foreign matter. In addition, by performing image processing on the captured original image, it is possible to obtain a binarized image in which a defective portion is revealed compared to the original image. Therefore, the original image before image processing is performed. Compared with the case where a defect is detected based on an image, the detection accuracy of the concave defect 50 is improved and the discrimination accuracy between the concave defect 50 and the foreign matter is improved.

  Examples of the use of the present invention include a defect detection method and a defect detection device for detecting a film defect, particularly a concave defect present on the surface of the film.

It is the schematic which shows the whole structure of the defect detection apparatus which enforces the defect detection method of the film of this invention. It is a figure for demonstrating the reflection mechanism at the time of irradiating the light which is not substantially parallel in one direction of a film. It is a figure for demonstrating the reflection mechanism at the time of irradiating the substantially parallel light in one direction of a film. It is a figure which shows the original image which imaged the reflected light at the time of irradiating the light which is not substantially parallel in one direction of a film. It is a figure which shows the original image which imaged the reflected light at the time of irradiating substantially parallel light in one direction of a film. It is a luminance distribution of the image in one direction of a film, Comprising: It is a figure which shows the luminance distribution in the aa cross section of the image shown in FIG. It is a luminance distribution of the image in one direction of a film, Comprising: It is a figure which shows the luminance distribution in the AA cross section of the image shown in FIG. It is a luminance distribution in the direction orthogonal to one direction of a film, Comprising: It is a figure which shows the luminance distribution in the BB cross section of the image shown in FIG. It is a figure which shows the image which carried out the expansion / contraction process of the image shown in FIG. It is a flowchart which shows the procedure of the defect detection method of the film of this invention. It is the schematic which shows the whole structure of the other defect detection apparatus which enforces the defect detection method of the film of this invention. It is the schematic which shows the whole structure of the other defect detection apparatus which enforces the defect detection method of the film of this invention. It is a figure which shows the image which reversed and output the result of having performed the binarization process, after performing the expansion / contraction process with respect to the imaged original image.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Film, 2 ... Illuminating means, 2a ... Light, 2b ... Coaxial epi-illumination, 2c ... Reflected light, 2d ... Light which is not substantially parallel, 2e ... Substantially parallel light, 2f ... Oblique illumination, 3 ... Light guide, 4 DESCRIPTION OF SYMBOLS ... Parallel light formation means, 5 ... Half mirror, 6 ... CCD camera sensor, 7 ... A / D conversion part, 8 ... Image memory part, 9 ... Arithmetic processing means, 10 ... Defect detection means, 11 ... Image processing means, 12 ... Expansion / shrinkage processing, 13 ... Image processing signal, 14 ... Binarization processing, 15 ... Storage means, 22 ... Determination threshold, 25 ... Determination threshold, 27 ... Minimum value, 40 ... Substrate, 41 ... Inspection object, 50 ... concave defect, 51 ... non-defective part around the concave defect, R1 ... unwinding roll, R2 ... winding roll, Y ... one direction of film 1

Claims (11)

  The surface of the film is substantially parallel in one direction of the film and is irradiated with diffusing light in a direction perpendicular to the one direction of the film, and the reflected image of the original image of the film is captured. A defect detection method for a film, comprising: detecting a defect on the surface of the film based on a luminance distribution in one direction.   The film defect detection method according to claim 1, wherein a defect on the surface of the film is detected by detecting that the luminance distribution in one direction of the film of the original image has two minimum values. .   An image of an original image obtained by irradiating the surface of the film with light that is substantially parallel in one direction of the film and diffusing in a direction perpendicular to the one direction of the film A defect detection method for a film, comprising: detecting a defect on the surface of the film based on a processed image obtained by performing the process.   The film defect detection method according to claim 3, wherein a defect on the surface of the film is detected by detecting that a luminance distribution in one direction of the film of the processed image has two minimum values. .   The defect on the surface of the film is detected based on a luminance distribution in one direction of the film of the expansion / contraction image obtained by performing expansion / contraction processing on the original image. The film defect detection method according to claim 4.   Defects on the surface of the film based on the image pattern of the binarized image obtained by performing the expansion / contraction processing on the original image or the original image and binarizing the obtained expansion / contraction image The film defect detection method according to claim 3, wherein the defect is detected.   2. The film according to claim 1, wherein the film is irradiated with diffusing light in a direction that is substantially parallel to the surface of the film to be transported and is orthogonal to the transport direction. Defect detection method.   Illumination means for emitting light applied to the surface of the film and light applied to the surface of the film are substantially parallel in one direction of the film and orthogonal to the one direction of the film. Parallel light forming means for diffusing light, imaging means for imaging reflected light reflected on the surface of the film, and calculation for calculating the luminance distribution in one direction of the film of the original image captured by the imaging means A film defect detection apparatus comprising: processing means; and defect detection means for detecting defects on the surface of the film based on the luminance distribution.   Illumination means for emitting light applied to the surface of the film and light applied to the surface of the film are substantially parallel in one direction of the film and orthogonal to the one direction of the film. Parallel light forming means for diffusing light, imaging means for imaging reflected light reflected on the surface of the film, and image processing for generating a processed image by performing image processing of the original image captured by the imaging means Means, arithmetic processing means for calculating a luminance distribution in one direction of the film of the processed image, and defect detecting means for detecting a defect on the surface of the film based on the luminance distribution. Film defect detection device.   Illumination means for emitting light applied to the surface of the film and light applied to the surface of the film are substantially parallel in one direction of the film and orthogonal to the one direction of the film. Parallel light forming means for diffusing light, imaging means for imaging reflected light reflected on the surface of the film, and image processing of the original image captured by the imaging means to generate a binarized image A film defect detection apparatus comprising: an image processing unit; and a defect detection unit that detects a defect on the surface of the film based on an image pattern of the binarized image.   11. The film defect detection apparatus according to claim 8, wherein the parallel light forming means is a cylindrical lens or a linear Fresnel lens.

Images