JP6314557B2 - Sheet inspection device - Google Patents

Description

  The present invention relates to a technique for detecting an abnormal part of a sheet-like inspection object.

  In a production line for manufacturing or processing a sheet-like article, an abnormal part on the sheet (using an image obtained by irradiating the sheet with visible light or ultraviolet light and capturing the transmitted light or reflected light with a camera) Inspection apparatuses that detect foreign matter contamination, dirt, wrinkles, etc.) are used (see, for example, Patent Document 1).

  Although the conventional inspection apparatus can detect an abnormal portion on a sheet, it cannot accurately determine what type of abnormality is detected. For this reason, conventionally, a sheet in which an abnormal part is detected must be discarded, a rank-reduced product, or a detailed inspection by visual inspection. However, in practice, there are various abnormalities that can occur in the sheet, and depending on the type, application, material, etc. of the product, there are some that do not have to be defective (defects).

  For example, a microporous polyolefin film is generally used as a separator of a lithium ion secondary battery. However, since the separator itself is not touched by human eyes, even if there is some dirt, there is no problem with functionality. It is not necessary to make it non-defective. On the other hand, metal contamination or adhesion and pinholes (holes) can be said to be a kind of abnormality that should never be overlooked because there is a risk of short circuit. On the other hand, in the case of paper material, small pinholes are acceptable, but there are cases where it is desirable to detect dirt and wrinkles that affect the appearance as defective.

JP 2010-8174 A (Patent No. 4950951)

  The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a technique that enables detection of an abnormal part included in a sheet-like inspection object and determination of the type of abnormality. .

  A first aspect of the present invention is a sheet inspection apparatus for inspecting a sheet-like inspection object, the first light source for irradiating visible light to the first surface of the inspection object, and the inspection object A second light source that irradiates the first surface with invisible light, a third light source that irradiates visible light onto the second surface opposite to the first surface of the inspection object, and The first light source is arranged so that the inspection object can be imaged by visible light irradiated from the first light source and reflected by the inspection object, and visible light irradiated from the third light source and transmitted through the inspection object. A first imaging sensor, a second imaging sensor arranged so that the inspected object can be imaged by invisible light emitted from the second light source and reflected by the inspected object, and the first imaging sensor. Obtained by the obtained first image of the object to be inspected and the second imaging sensor. Based on the second image of the inspection object, a processing unit that detects an abnormal part included in the inspection object and discriminates the type of abnormality occurring at the detected abnormal part, and the processing And an output unit that outputs information on an abnormal part including at least information indicating the type of abnormality determined by the unit.

  Here, the “visible light” emitted from the first light source and the third light source may be light including at least a wavelength in the visible light region, and is not necessarily limited to light including a wavelength in the entire visible light region (white color). Light), and conversely, light having a wavelength outside the visible light region may be included. Further, the “invisible light” emitted from the second light source may be light including at least a wavelength outside the visible light region, and does not exclude including the wavelength in the visible light region.

  If there is any abnormality on the object to be inspected, characteristics such as light absorption, reflectance, and transmittance may change at the abnormal portion as compared to other portions (that is, a portion having no abnormality). In addition, the manner of change depends on the wavelength of light and the type of anomaly. Therefore, depending on what type of abnormality has occurred on the object to be inspected, the change in the pixel value of each of the first image obtained by the first image sensor and the second image obtained by the second image sensor is changed. Features appear in the way. In particular, in the above configuration, visible light is emitted from both sides (both sides) of the object to be inspected using the first light source and the third light source, and the reflected light or transmitted light is imaged by the same first image sensor. Therefore, in the case of only the reflected light of the light from the first light source, the pixel value of the first image is the reflected light of the light from the first light source and the transmitted light of the light from the third light source. In both cases, it may be different from the case of only the transmitted light from the third light source, the amount of received light may increase or decrease due to the occurrence of an abnormality, and the light from the third light source Considering that the color balance of transmitted light may change due to the difference in the absorption rate for each wavelength when transmitting through an abnormal location, there are many variations in the way the pixel value of the first image changes. There is. Therefore, according to the configuration of the light source and the image sensor of the present invention, various abnormality detection and detection are performed based on the first image obtained by the first image sensor and the second image obtained by the second image sensor. It is possible to perform type discrimination.

  The type of abnormality indicates how the pixel values of the first image and the second image increase or decrease with respect to the normal state where there is no abnormality when the abnormality occurs. Each can be categorized in advance. Therefore, for example, the processing unit determines which type the change of the pixel value of the first image and the pixel value of the second image corresponding to the same position on the inspection object corresponds to the normal state. By doing so, it is possible to determine the type of abnormality occurring at the position. Note that the association between the type of change (increase or decrease) of the pixel value of the first image and the pixel value of the second image and the type of abnormality may be defined by a lookup table, or in the program You may implement as judgment logic.

  The output unit includes the first image of the region including the abnormal part, the second image of the region including the abnormal part, a graph showing a change in the pixel value of the first image at the abnormal part, It is also preferable to output at least one of the information indicating the change in the pixel value of the second image at the abnormal portion. By outputting information on such anomalous locations, the user (inspector) can grasp the details of the anomaly that has occurred, determine whether the anomaly should be a defect (defect), It can be used for feedback to driving conditions.

The visible light emitted from the first light source and the visible light emitted from the third light source are lights having the same spectral distribution including a plurality of color components, and the first imaging sensor Visible light irradiated from one light source and reflected from the inspection object can be received, and visible light irradiated from the third light source and transmitted straight through the inspection object can be received. The pixels of the first image have pixel values for each of a plurality of color components, and the processing unit corresponds to the first image corresponding to an arbitrarily selected target position on the inspection object. The second image corresponding to the position of interest when the total value of the pixel values for each of the plurality of color components is larger than that in the normal state in which there is no abnormality in the inspection object. Is smaller than that in the normal state, and When the balance of the plurality of color components in the pixels of the first image corresponding to the eye position is the same as the balance in the normal state, the type of abnormality occurring at the target position is a pinhole defect It is good to distinguish. The pinhole defect can be accurately determined by the arrangement of the light source and the imaging sensor of the present invention and the determination conditions thereby. This function is particularly useful for products in which pinholes are one of the major defects, such as secondary battery separators.

  The visible light emitted from the first light source and the visible light emitted from the third light source are lights having the same spectral distribution including a plurality of color components, and the first imaging sensor Visible light irradiated from one light source and reflected from the inspection object can be received, and visible light irradiated from the third light source and transmitted straight through the inspection object can be received. The pixels of the first image have pixel values for each of a plurality of color components, and the processing unit corresponds to the first image corresponding to an arbitrarily selected target position on the inspection object. And when the pixel value of the second image corresponding to the target position is in the normal state when the total value of the pixel values of the plurality of color components in the pixel is larger than that in the normal state The first corresponding to the attention position is smaller than the first position. When the balance of the plurality of color components in the pixels of the image is not the same as the balance in the normal state, and when the pixel value of the second image corresponding to the target position is larger than in the normal state, It may be determined that the type of abnormality is different from the pinhole defect. Thereby, the kind of abnormality can be classified more finely.

  The visible light emitted from the first light source and the visible light emitted from the third light source are lights having the same spectral distribution including a plurality of color components, and the first imaging sensor Visible light irradiated from one light source and reflected from the inspection object can be received, and visible light irradiated from the third light source and transmitted straight through the inspection object can be received. The pixels of the first image have pixel values for each of a plurality of color components, and the processing unit corresponds to the first image corresponding to an arbitrarily selected target position on the inspection object. Each of the plurality of color components in the pixel of the first image corresponding to the target position in a case where the total value of the pixel values of the plurality of color components in the pixel is smaller than that in the normal state. Between the pixel value of the second image and the pixel value of the second image , Wherein when the ratio of reduction relative to the normal state is the same, or an abnormality of the kind occurring in the target position is determined to be in metal defects. A metal defect can be accurately discriminated by the arrangement of the light source and the image sensor of the present invention and the determination conditions thereby. This function is particularly useful for a product such as a separator of a secondary battery where metal contamination or adhesion is one of the serious defects.

  When evaluating a change in pixel value, a relative value based on the value in the normal state, in other words, the degree of decrease or increase in the pixel value with respect to the normal state (state in which there is no abnormality in the inspection object), in other words In addition, it is preferable to evaluate with a value (normalized pixel value) normalized by a value in a normal state. As a typical example of the normalized pixel value, a value (ratio) obtained by dividing the pixel value based on the output signal of the imaging sensor by the pixel value in the normal state can be used. By evaluating with such standardized pixel values, due to fluctuations in the light amount of the light source, differences in transmittance, reflectance, and absorption rate at each abnormal location, differences in transmittance, reflectance, and absorption rate of the inspection object Variations can be canceled. Therefore, it is possible to make it less susceptible to the influence of disturbances, abnormalities or variations in the inspected object, and it is possible to stabilize the accuracy of detecting abnormal parts and determining the type of abnormalities.

  The invisible light may be infrared light or ultraviolet light. If light of these wavelengths is used, the output signal of the sensor varies significantly depending on the type of abnormality, so that it is possible to discriminate abnormalities that cannot be identified only by visible light.

  It is preferable that the first imaging sensor not to receive the invisible light and the second imaging sensor not to receive the visible light are provided in one imaging device. As a result, the apparatus can be miniaturized and the degree of freedom of installation is improved.

  The imaging apparatus may include a spectroscopic element that divides one optical path and guides light to each of the first imaging sensor and the second imaging sensor. This eliminates the need for alignment for correcting the deviation between the measurement position of the first image sensor and the measurement position of the second image sensor, thereby simplifying the apparatus configuration and processing.

  Note that the present invention can also be understood as a sheet inspection apparatus having at least a part of the above-described configuration, and a control method, a sheet inspection method, or an abnormal type determination of a sheet having at least a part of the above-described processing. It can also be understood as a method. The present invention can also be understood as a program for causing a computer to execute such a method, or a computer-readable storage medium in which such a program is stored non-temporarily. Each of the above configurations and processes can be combined with each other as long as no technical contradiction occurs.

  According to the present invention, it is possible to detect an abnormal part included in a sheet-like inspection object and to determine the type of abnormality.

1 is a block diagram of a sheet inspection apparatus according to Embodiment 1. FIG. The figure which shows an example of the normalization pixel value obtained in each signal processing part. An example of the result output screen which an output part outputs. The figure which shows the mode of reflection of light in the case of "dirt", and the change of each output value. The figure which shows the mode of reflection of light in the case of "metal defect", and the change of each output value. The figure which shows the mode of reflection of light and transmission in the case of a "pinhole defect", and the change of each output value. The figure which shows the mode of reflection of light and transmission in the case of "abnormality A", and the change of each output value. The figure which shows the mode of reflection of light and transmission in the case of "abnormal B", and the change of each output value. The flowchart of abnormality detection and kind discrimination | determination by a sheet inspection apparatus. The detailed flowchart of abnormality kind discrimination | determination. FIG. 6 is a block diagram of a sheet inspection apparatus according to a second embodiment. FIG. 10 is a diagram illustrating an arrangement of sensors according to the second embodiment. FIG. 9 is a block diagram of a sheet inspection apparatus according to a third embodiment. FIG. 6 is a diagram illustrating an internal structure of an imaging apparatus according to a third embodiment.

  DESCRIPTION OF EMBODIMENTS Hereinafter, embodiments for carrying out the present invention will be exemplarily described in detail with reference to the drawings. However, the dimensions, materials, shapes, relative arrangements, and the like of the components described in this embodiment are not intended to limit the scope of the present invention only to those unless otherwise specified.

<Example 1>
FIG. 1 is a block diagram of a sheet inspection apparatus 1 according to the present embodiment. As an illumination system, the sheet inspection apparatus 1 includes a visible light source 31 (first light source) that irradiates the upper surface (first surface) of the inspection object 2 with visible light, and ultraviolet light or red light on the upper surface of the inspection object 2. An IR / UV light source 32 (second light source) that irradiates at least one of outside light, and a visible light source for transmission 33 (third light source) that irradiates visible light onto the lower surface (second surface) of the inspection object 2. And have. The sheet inspection apparatus 1 includes a visible light camera 41 (first image sensor) and an IR / UV light camera 42 (second image sensor) as a measurement system. The visible light source 31 and the visible light source for transmission 33 are installed so as to irradiate light at the same position of the inspection object 2 (however, the surface that is exposed to light is different), and the visible light camera 41 is irradiated from the visible light source 31. The inspection object 2 is arranged so that it can be imaged by the light reflected from the upper surface of the inspection object 2 and the light irradiated from the transmission visible light source 33 and transmitted straight through the inspection object 2. On the other hand, the IR / UV light camera 42 is arranged so that the inspection object 2 can be imaged by the light emitted from the IR / UV light source 32 and reflected from the upper surface of the inspection object 2. Further, the sheet inspection apparatus 1 detects the abnormal part included in the inspection object 2 and determines the type of abnormality based on the output signal of the visible light camera 41 and the output signal of the IR / UV light camera 42. have.

  The inspection object 2 is formed in a sheet shape, for example, and is conveyed in the direction of the arrow in FIG. Examples of the inspection object 2 include paper, film, resin, and cellulose. Further, the inspection object 2 may be a separator used for a secondary battery, an optical sheet used for liquid crystal, or the like. In this embodiment, the illumination system and the measurement system are fixed and the inspection object 2 is moved. Instead, the inspection object 2 is fixed and the illumination system and the measurement system are moved. Also good.

  The sheet inspection apparatus 1 detects an abnormal part included in the inspection object 2 based on the first image obtained by the visible light camera 41 and the second image obtained by the IR / UV light camera 42, and detects the detected abnormality. It has a function of discriminating the type and outputting the result. In this example, a porous film made of an olefin resin used for a separator of a secondary battery or the like is an inspection object 2, and “dirt”, “metal defect”, “pinhole defect”, “abnormal A”, Five types of abnormalities “abnormality B” are detected and determined. “Dirt” is dirt or foreign matter adhering to the film surface. “Metal defects” are, for example, defects in which metal powder or the like generated or peeled off from a manufacturing apparatus or a conveying apparatus is mixed or attached to a film, and “pinhole defects” are holes formed in the film. “Abnormal A” is unevenness (portion where the porous surface is rough) or the portion where oil adheres and penetrates during processing of the porous film, and the light transmittance is higher than that of the normal film. It is an abnormal condition that has increased. “Abnormal B” is unevenness (portion where the porous material is dense) generated during processing of the porous film, and an abnormal state in which the reflectance of light is increased compared to a normal film. It is. "Metal defects" and "pinhole defects" are serious defects that should not be overlooked in secondary battery separators, whereas "dirt", "abnormal A", and "abnormal B" If it is small, it is an abnormality that does not have to be a defect (defect).

As the illumination system, an LED or the like in which the wavelength region is limited, or an illumination system in which the wavelength region is limited using a wavelength filter can be used. The visible light source 31 and the visible light source for transmission 33 are preferably configured to be able to irradiate the inspection object 2 with light having the same spectral distribution including a plurality of color components. In this embodiment, the visible light source 31 and the transmissive visible light source 33 are produced using the same type of white light source.
For the measurement system, for example, an imaging apparatus including a CCD image sensor in which 4096 light receiving elements are arranged in series can be used. In each light receiving element, light is converted into electric charge according to the amount of received light. In the present embodiment, the visible light camera 41 includes three CCD image sensors for R, G, and B components. The IR / UV light camera 42 includes a CCD image sensor that detects at least one of infrared light and ultraviolet light. The charge output from each light receiving element is input to the processing device 5 as an output signal (imaging data).

  In the present embodiment, a plurality of cameras can be provided in the width direction of the inspection object 2 in accordance with the width of the inspection object 2 so that the entire width of the inspection object 2 can be captured by the camera. Further, the visible light camera 41 and the IR / UV light camera 42 are arranged so as to be shifted in the transport direction.

The processing device 5 processes the imaging data output from the visible light camera 41 for each of the RGB components. The R signal processing unit 51, the G signal processing unit 52, the B signal processing unit 53, and the IR / UV light camera 42. And an IR / UV signal processing unit 54 for processing imaging data output from. The R signal processing unit 51 performs white shading processing on the R component signal (R signal) for one line (4096 pixels) output from the visible light camera 41, and corrects variation in output level for each light receiving element. . Similarly, the G signal processing unit 52 performs white shading processing on the G component signal (G signal), and the B signal processing unit 53 performs white shading processing on the B component signal (B signal). Further, the IR / UV signal processing unit 54 performs white shading processing on one line of infrared light or ultraviolet light signal (IR / UV light signal) output from the IR / UV light camera 42. Thereafter, the R, G, B, and IR / UV values output from the respective signal processing units 51, 52, 53, and 54 are respectively converted into output pixel values I _R , I _G , I _B , and I _{IR /} UV. _{Indicated as UV} .

Further, each signal processing unit 51, 52, 53, 54 calculates a normalized pixel value for each of R, G, B, and IR / UV. Here, the normalized pixel value refers to the output pixel values I _R , I _G , I _B , and I _{IR / UV} after white shading, and the output pixel values (normal values) when there is no abnormality in the inspection object 2. It is a value divided by N _R , N _G , N _B , N _{IR / UV} . In this embodiment, the normalized pixel value has a range of 0 to 255, and is normalized so that the normal value is 128, which is the median value of the range. The “output pixel value (normal value) when there is no abnormality” may be an average value of output pixel values when imaging is performed a plurality of times. The normalized pixel value becomes smaller as the degree of decrease in the output pixel value (camera received light amount) becomes larger, and becomes larger as the increase in the output pixel value (camera received light amount) becomes larger, and the output pixel value fluctuated. Correlate with degree. The normalized pixel value is a value close to 128 in a portion where the inspection object 2 is not abnormal (also referred to as a texture). In the following description, when it is not necessary to distinguish the output pixel value and the normalized pixel value, they are simply referred to as pixel values.

  FIG. 2 is a diagram illustrating an example of normalized pixel values obtained by the signal processing units 51, 52, 53, and 54. The horizontal axis is a pixel (light receiving element), and the vertical axis is a normalized pixel value. The normalized pixel value is about 128 in a portion where there is no abnormality (ground), but the normalized pixel value increases or decreases in the pixel corresponding to the abnormal portion. The direction of change (increase or decrease) of the normalized pixel value and the degree of change thereof may differ for each wavelength and for each type of abnormality. Even in the formation portion, the normalized pixel value varies slightly from pixel to pixel due to the influence of irregularities on the surface of the object 2 to be inspected.

  The processing device 5 includes an alignment processing unit 55 that aligns an image obtained from the visible light camera 41 and an image obtained from the IR / UV light camera 42. Here, since the visible light camera 41 and the IR / UV light camera 42 are arranged so as to be shifted in the transport direction of the object 2 to be inspected, the portion imaged by the visible light camera 41 is the IR / UV light camera. It takes a certain time to reach the position imaged at 42. In order to compare the pixel values at the same location obtained from the visible light camera 41 and the IR / UV light camera 42, the alignment processing unit 55 includes one line of image data obtained from the visible light camera 41, IR / UV Position alignment (time alignment) with image data for one line obtained from the UV light camera 42 is performed.

  Here, since the conveyance speed of the inspection object 2 and the distance from the visible light camera 41 to the IR / UV light camera 42 are set in advance, the image is picked up by the visible light camera 41 based on these values. It is possible to calculate the time delay until the spot is imaged by the IR / UV light camera 42. That is, alignment can be performed by shifting the data by this time delay. Similarly, when different locations are imaged with the R signal, G signal, and B signal, or when different locations are imaged with ultraviolet light and infrared light, these positions are aligned. .

  Further, the processing device 5 includes an abnormality detection unit 56 that detects an abnormal part included in the inspection object 2 and a detection threshold value storage unit 56A that stores a threshold value used for abnormality determination. In this embodiment, as will be described later, when the degree of change in the pixel value of the image obtained from the visible light camera 41 is large to some extent, it is determined that there is an abnormality. Therefore, the threshold value of the degree of change of the pixel value to be determined as abnormal is held in the detection threshold value storage unit 56A. This threshold value is determined by the type of the inspection object 2 and the inspection standard set by the user.

  In addition, the processing device 5 includes a determination unit 57 that determines the type of abnormality when an abnormal part is detected, and a determination threshold value storage unit 57A that stores a plurality of threshold values used for processing to determine the type of abnormality. Yes. The determination unit 57 determines the correspondence between the type of change (increase or decrease) in pixel values of the image obtained from the visible light camera 41 and the image obtained from the IR / UV light camera 42 and the five types of abnormalities described above. The type of abnormality is determined by determining in which type the pixel value change method corresponds in advance. Detailed processing will be described later.

  The output unit 58 is a function for outputting information related to the abnormal part. The output destination of information is typically a display device, but outputs information to the printing device, outputs a message or alarm from a speaker, or sends a message to the user's terminal by e-mail or the like. Information may be transmitted to an external computer. FIG. 3 is an example of a result output screen that the output unit 58 outputs to the display device. In this screen, information 580 indicating the type of anomaly detected and determined (“pinhole defect” in the example of FIG. 3), an image 581 of the anomalous portion captured by the visible light camera 41, and the IR / UV light camera 42 Graphs 584 and 585 showing changes in the output signals (output pixel values or normalized pixel values) of the visible light camera 41 and the IR / UV light camera 42 in a line 583 passing through the abnormal portion, and the imaged abnormal portion 582, 586, 587, etc. are displayed. By outputting information on such anomalous locations, the user (inspector) can grasp the details of the anomaly that has occurred, determine whether the anomaly should be a defect (defect), It can be used for feedback to manufacturing conditions and operating conditions.

  FIGS. 4 to 8 show the correspondence between the type of change (increase or decrease) in the normalized pixel value obtained from each signal processing unit 51 to 54 and the type of abnormality. Such a correspondence can be obtained by carrying out an experiment for each type of abnormality that can actually occur. 4 to 8 show examples of a porous film used for a secondary battery separator or the like. Since how the standardized pixel values change, how to classify them, and the types of abnormalities that can occur vary depending on the material and physical properties of the object to be inspected, the correspondence relationships shown in FIGS. 4 to 8 are assumed. It is preferable to prepare in advance for each object to be inspected and mount it in a program of the processing device 5 as a lookup table or determination logic.

  4 to 8 show output signals (normalized pixels) of visible light and invisible light when “dirt”, “metal defect”, “pinhole defect”, “abnormal A”, and “abnormal B” occur, respectively. Value) is schematically shown. In each figure, (a) is a schematic diagram showing how light is reflected and transmitted at an abnormal location, (b) is a change in the normalized pixel value of the R signal, and (c) is a normalized pixel value of the G signal. (D) is a schematic diagram showing a change in the normalized pixel value of the B signal and (e) a change in the normalized pixel value of the IR signal.

  As shown in FIG. 4, when “dirt” 61 adheres to the film 60, visible light 62 is absorbed by the dirt 61, so that the R signal, G signal, and B signal (output signals of the visible light camera 41) (Hereinafter collectively referred to as RGB signals) is significantly smaller than the normal state. On the other hand, since the absorption of the invisible light 63 (in this example, infrared light) is small, the degree of reduction of the IR signal is significantly smaller than that of the RGB signal. The visible light 64 irradiated from the lower surface of the film 60 hardly passes through the film 60 and is not received by the visible light camera 41.

  In the case of a “metal defect”, as shown in FIG. 5, both the visible light 62 and the invisible light 63 are absorbed by the metal 65. Therefore, the RGB signal that is the output signal of the visible light camera 41 and the IR / UV light camera are used. The IR signal that is the output signal of 42 is significantly smaller than the normal state. In addition, the rate of decrease of the R signal, G signal, B signal, and IR signal is approximately the same. Also in this case, all visible light 64 is reflected and is not received by the visible light camera 41.

  In the case of a “pinhole defect”, as shown in FIG. 6, both visible light 62 and invisible light 63 from the upper surface are transmitted to the lower surface side through the pinhole 66 and are not received by the camera. However, the visible light 64 from the lower surface passes straight through the pinhole 66 to the upper surface side and enters the visible light camera 41. Therefore, the RGB signal is significantly larger than the normal state, but the IR signal is significantly smaller than the normal state. Since the visible light 64 is directly incident on the visible light camera 41, the color balance in the output signal of the visible light camera 41 is the same as that of the original visible light 64.

  “Abnormal A” is a state where unevenness (portion where the porous surface is rough) or oil adhered or penetrated during processing of the porous film. In this case, the portion where unevenness or oil adheres and penetrates becomes a transparent spot 67 with a slight tint. Then, as shown in FIG. 7, the visible light 62 and the invisible light 63 from the upper surface are almost transmitted through the spot 67 to the lower surface side. On the other hand, visible light 64 from the lower surface passes through the spot 67 to the upper surface side and is received by the visible light camera 41. However, unlike the pinhole defect of FIG. 6, when the light passes through the spot 67, the light of a part of the wavelength is absorbed and attenuated, so that the color balance of the transmitted light is lost. In the example of FIG. 7, the R signal, the G signal, and the B signal all increase significantly compared to the normal state, but the G signal and the B signal are significantly smaller than the R signal, and the color balance changes. You can see that As in the case of the pinhole defect, the IR signal decreases compared to the normal state.

  “Abnormal B” is unevenness (portion in which the porous material is dense) generated during the processing of the porous film, and is a state in which the reflectance of light is increased as compared with a film in a normal state. In this case, as shown in FIG. 8, since the visible light 62 and the invisible light 63 from the upper surface are reflected at the spot 68 where the unevenness occurs, all of the R signal, the G signal, the B signal, and the IR signal are reflected. Increases significantly compared to normal conditions. The visible light 64 irradiated from the lower surface of the film 60 hardly passes through the film 60 and is not received by the visible light camera 41.

  By clarifying the correspondence between the type of change of each signal and the type of abnormality as described above, the type of abnormality can be easily and accurately determined. In this embodiment, the degree of change of each signal is evaluated using the normalized pixel value obtained by normalizing the output value of each signal with the value in the normal state. Thereby, it is possible to cancel variations caused by fluctuations in the amount of light of the light source, differences in transmittance, reflectance, and absorption rate at each abnormal location, differences in transmittance, reflectance, and absorption rate of the inspection object. Therefore, it is possible to make it less susceptible to the influence of disturbances, abnormalities or variations in the inspected object, and it is possible to stabilize the accuracy of detecting abnormal parts and determining the type of abnormalities.

  Further, the type of defect may be determined using only one or two of the R component, the G component, and the B component for visible light. Further, component light having a wavelength different from that of R, G, and B (for example, cyan, magenta, yellow, etc.) may be used. Further, the type of defect may be determined using one or both of infrared light and ultraviolet light. Further, when selecting any one of the R component, the G component, and the B component and one of infrared light and ultraviolet light, a combination that increases the difference in wavelength may be selected. For example, combining with ultraviolet light having a short wavelength is an R component having a long wavelength in visible light, and combining with infrared light having a long wavelength is a B component having a short wavelength among visible light. Thereby, since the difference in the reflection ratio of defects appears more remarkably, the determination accuracy can be improved.

  Further, a Si (silicon) -based semiconductor may be used for the light receiving element of the camera. If a Si-based semiconductor light receiving element is used, any of ultraviolet light, visible light, and infrared light can be detected. Further, the number of pixels can be increased, wide-range or high-speed measurement is possible, and the cost can be kept low.

Next, the flow of processing of the sheet inspection apparatus 1 will be described with reference to FIG. FIG. 9 is a flowchart of processing executed by the processing device 5.
In step S101, the visible light source 41, the IR / UV light source 32, and the transmissive visible light source 33 are turned on, and the visible light camera 41 and the IR / UV light camera 42 shoot the inspection object 2, The output signal is taken into the processing device 5.

In step S102, an R signal processing unit 51, a G signal processing unit 52, an R signal, a G signal, and a B signal output from the visible light camera 41, and an IR / UV signal output from the IR / UV light camera 42, Each of the B signal processing unit 53 and the IR / UV signal processing unit 54 performs white shading processing to generate output pixel values I _R , I _G , I _B , and I _{IR / UV} . The signal processing units 51 to 54 generate standardized pixel values of R, G, B, and IR / UV from the output pixel values I _R , I _G , I _B , and I _{IR / UV} , respectively. The data of the output pixel value and the normalized pixel value is output to the alignment processing unit 55. Hereinafter, the normalized pixel values of the R signal, G signal, B signal, and IR / UV signal are also referred to as R value, G value, B value, and IR value.

  In step S103, the alignment processing unit 55 aligns the RGB image and the IR / UV image based on the conveyance speed of the inspection object 2 and the distance between the cameras 41 and 42.

ここで、Ｉ_Ｒ，Ｉ_Ｇ，Ｉ_ＢはＲ，Ｇ，Ｂの出力画素値であり、Ｎ_Ｒ，Ｎ_Ｇ，Ｎ_ＢはＲ，Ｇ，Ｂの通常状態の出力画素値（地合の出力画素値）である。また、Ｉ_ｇｒａｙは輝度値（グレー値）であり、ＲＧＢの色成分ごとの画素値（Ｉ_Ｒ，Ｉ_Ｇ，Ｉ_Ｂ）を総合する値である。Ｃｏｅｆｆ_Ｒ，Ｃｏｅｆｆ_Ｇ，Ｃｏｅｆｆ_Ｂは重み係数であり、Ｃｏｅｆｆ_ｄｉｖは除算係数であり、ｓは１又は−１の値をとる符号係数である。本実施例では、（Ｉ_Ｒ−Ｎ_Ｒ）、（Ｉ_Ｇ−Ｎ_Ｇ）、（Ｉ_Ｂ−Ｎ_Ｂ）のなかで最大値をとるものの符号を、符号係数ｓとして用いる。
そして、異常検出部５６は、グレー画像において、輝度値Ｉ_ｇｒａｙが１２８×０．９より小さい画素からなる領域（画素群）、又は、グレー値Ｉ_ｇｒａｙが１２８×１．１より大きい画素からなる領域（画素群）を検出し、当該領域の面積が所定値を超えていた場合に当該領域を「異常個所」と判定する。なお、本実施例では、異常か否かの判別閾値を通常値（１２８）の±１０％の値に設定したが、これは一例にすぎず、被検査物やカメラの特性などに応じて閾値は適宜設定すればよい。また、本実施例では上記式で計算した輝度値により異常判定を行ったが、単純に、Ｒ、Ｇ、Ｂのいずれか（例えばＧ）の画素値をそのまま輝度値として用いてもよいし、Ｒ、Ｇ、Ｂの画素値の平均値や最大値を輝度値として用いてもよい。 In step S104, the abnormality detection unit 56 detects an abnormality. For example, the abnormality detection unit 56 generates a gray image from an RGB image using the following color / gray conversion formula.

_Here, I _R, I _G, the _{I B} is the output pixel value of the _{R, G, B, N R} , N G, N B is R, G, the output pixel value of the normal state of B (formation output Pixel value). Further, I _gray is a luminance value (gray value), and is a value that _combines pixel values (I _R , I _G , I _B ) for each RGB color component. Coeff _R , Coeff _G , and Coeff _B are weighting coefficients, Coeff _div is a division coefficient, and s is a code coefficient that takes a value of 1 or -1. In this embodiment, the code having the maximum value among (I _R −N _R ), (I _G −N _G ), and (I _B −N _B ) is used as the code coefficient s.
Then, the abnormality detection unit 56 includes, in the gray image, a region (pixel group) including pixels having a luminance value I _gray smaller than 128 × 0.9 or a pixel having a gray value I _gray larger than 128 × 1.1. A region (pixel group) is detected, and when the area of the region exceeds a predetermined value, the region is determined as an “abnormal part”. In this embodiment, the threshold value for determining whether or not there is an abnormality is set to a value of ± 10% of the normal value (128). However, this is only an example, and the threshold value is set according to the characteristics of the object to be inspected or the camera. May be set as appropriate. Further, in this embodiment, the abnormality determination is performed based on the luminance value calculated by the above formula. However, a pixel value of any one of R, G, and B (for example, G) may be simply used as the luminance value as it is. An average value or a maximum value of R, G, and B pixel values may be used as the luminance value.

  In step S105, it is determined whether an abnormal part is detected in step S104. If an affirmative determination is made in step S105, the process proceeds to step S106. On the other hand, if a negative determination is made in step S105, it is determined that there is no abnormality, and this routine is terminated.

In step S106, the determination unit 57 determines the type of abnormality.
FIG. 10 shows a detailed flow of abnormality type discrimination. First, in step S201, the determination unit 57 compares the detected luminance value of the abnormal part with the luminance value of the normal state (the luminance value of the ground), and the abnormality is a dark defect (the luminance value is smaller than that of the normal state). It is determined whether it is (that is, I _gray <128)) or a light defect (the luminance value is larger than the normal state (that is, I _gray > 128)). In the case of a dark defect, the routine enters a discrimination routine for dirt (FIG. 4) / metal defect (FIG. 5). In the case of a bright defect, pinhole defect (FIG. 6) / abnormal A (FIG. 7) / abnormal B (FIG. 8). ) Is entered.

  In step S <b> 202, the determination unit 57 evaluates the difference in the rate of decrease in visible light and invisible light. Specifically, the absolute value of the difference between the R value and the IR value (denoted as | R-IR |), the absolute value of the difference between the G value and the IR value (denoted as | G-IR |), the B value, The absolute value of the difference between the IR values (denoted as | B-IR |) is obtained, and the minimum value among them, min | V-IR | (where V = R, G, B) is greater than the threshold value TH1. Check if it is big. If min | V-IR | ≧ TH1, that is, if the rate of decrease in visible light and invisible light is significantly different, it falls under the category of FIG. 4 and is determined as “dirt” (step S203). On the other hand, if min | V-IR | <TH1, that is, if the rate of decrease in visible light and invisible light is substantially the same, it falls under the category of FIG. S204).

  In step S205, the determination unit 57 checks whether the IR value is greater than the threshold value TH2. For example, TH2 may be set to the same level as 128, which is the value of the normal state (form). When IR> TH2, it corresponds to the type of FIG. 8 and is determined as “abnormal B” (step S206). On the other hand, if IR ≦ TH2, the process proceeds to step S207.

  In step S207, the determination unit 57 checks whether or not the difference maxV−minV between the maximum value maxV and the minimum value minV among the R value, the G value, and the B value is larger than the threshold value TH3. This corresponds to processing for evaluating the balance of pixel values for each color component in the pixels of the image obtained by the visible light camera 41. If maxV−minV <TH3, that is, if the color balance is substantially the same as the balance in the normal state, it corresponds to the type of FIG. 6 and is determined as a “pinhole defect” (step S208). On the other hand, if maxV−minV ≧ TH3, that is, if the color balance is significantly different from the balance in the normal state, it corresponds to the type of FIG. 7 and is determined as “abnormal A” (step S209). The method for evaluating the color balance may be another method. The threshold value TH3 may be set to about 10% of the maximum value maxV among the R value, G value, and B value, for example. Note that a color balance evaluation method other than the method described in this embodiment may be used. For example, when the saturation of the pixel is calculated from the pixel values of R, G, and B, and the saturation is smaller than the threshold (that is, close to an achromatic color), the color balance is substantially the same as the balance in the normal state. You may determine that there is.

  If the type of abnormality is specified by the above determination logic, the process proceeds to step S107 in FIG. In step S107, the output unit 58 outputs information on the abnormal part (see FIG. 3). At this time, information is output only in the case of serious defects such as “metal defects” and “pinhole defects”, and information is not output in the case of other abnormalities, or only when the area of the abnormal part is large to some extent. Information may be output as a defect (defect). In addition, in the case of “metal defect”, “pinhole defect”, or other abnormality having a large area, control may be performed to output a warning or alarm or stop the film manufacturing apparatus.

  According to the present embodiment described above, the abnormality of the sheet-like inspection object 2 can be detected and the type of the detected abnormality can be finely discriminated. This makes it possible to distinguish between abnormalities that may affect product quality and abnormalities that do not have to be defective (defects), so that too much inspection (overdetection) is suppressed and product yield is increased. be able to.

<Example 2>
FIG. 11 is a block diagram of the sheet inspection apparatus 1 according to the second embodiment. In this embodiment, only one imaging device 4 is provided. This single image pickup device 4 serves as both the visible light camera 41 and the IR / UV light camera 42 according to the first embodiment. That is, the imaging apparatus 4 according to the present embodiment includes a light receiving element that measures at least one component of R, G, and B and a light receiving element that measures at least one of infrared light and ultraviolet light. The visible light source 31, the IR / UV light source 32, and the transmissive visible light source 33 irradiate light at the same location (however, the transmissive visible light source 33 irradiates the lower surface). Since other devices are the same as those in the first embodiment, the description thereof is omitted.

  Here, FIG. 12 is a diagram showing the arrangement of the light receiving elements provided in the imaging device 4. R detects the R component in visible light, G detects the G component in visible light, B detects the B component in visible light, IR / UV is infrared light Or it is a line sensor which detects ultraviolet light. The R, G, B, and IR / UV sensors are arranged so as to be shifted in the transport direction. For this reason, as in the first embodiment, it is necessary to align the output signals of the sensors. By using such an imaging device 4, the size of the device can be reduced.

<Example 3>
FIG. 13 is a block diagram of the sheet inspection apparatus 1 according to the third embodiment. FIG. 14 is a diagram illustrating the internal structure of the imaging device 4. R detects the R component in visible light, G detects the G component in visible light, B detects the B component in visible light, IR / UV is infrared light Or it is a sensor which detects ultraviolet light. The present embodiment is the same as the second embodiment in that one imaging device 4 is provided, but the light incident on the imaging device 4 using the spectroscopic element 43 that divides the optical path is converted into B light, G light, R light, The second embodiment is different from the second embodiment in that the spectrum is split into IR / UV light and each is measured by a corresponding light receiving element.

That is, in this embodiment, since the light is separated by the spectroscopic element 43 and received by each sensor, visible light and IR / UV light incident from the same position of the inspection object 2 can be imaged by a single imaging device 4. it can. And since the data of the same position can be obtained simultaneously, alignment is not necessary. For this reason, the alignment processing unit 55 required in the first and second embodiments is not necessary. Since other devices are the same as those in the second embodiment, description thereof is omitted.
By using such an image pickup device 4, it is not necessary to align, so that the processing can be simplified. In addition, since it is not affected by the alignment accuracy, the accuracy of abnormality detection can be improved.

<Others>
The above-described embodiments are merely illustrative of the present invention, and the present invention is not limited to the above specific modes. The present invention can be variously modified within the scope of its technical idea. For example, in the above embodiment, three signals of R, G, and B are output as visible light, but any wavelength signal may be used as long as it is two or more signals. For example, light including B and G components may be received by a single sensor to obtain a cyan signal, and two signals of an R signal and a cyan signal may be used as visible light signals.
In the above embodiment (FIG. 9), the abnormal part detection process is first performed, and the abnormality type determination process is applied only to the detected abnormal part. However, the abnormality type determination process is applied to the entire image. You may apply. For example, the same effect as in the above-described embodiment can be obtained even when the abnormality location detection process (step S104) and the abnormality type determination process (step S106) are performed in parallel and the results of both processes are combined.

1: Sheet inspection device, 2: Inspection object, 4: Imaging device, 5: Processing device 31: Visible light source, 32: IR / UV light source, 33: Visible light source for transmission 41: Visible light camera, 42: IR / UV Optical camera, 43: Spectroscopic element 51: R signal processing unit, 52: G signal processing unit, 53: B signal processing unit, 54: IR / UV signal processing unit, 55: Positioning processing unit, 56: Abnormality detection unit, 56A: Detection threshold storage unit, 57: Determination unit, 57A: Determination threshold storage unit, 58: Output unit 60: Film, 61: Dirt, 62: Visible light, 63: Invisible light, 64: Visible light, 65: Metal, 66: Pinhole, 67: Spot, 68: Spot

Claims (7)

A sheet inspection apparatus for inspecting a sheet-like inspection object,
A first light source that emits visible light to the first surface of the object to be inspected;
A second light source that irradiates the first surface of the inspection object with invisible light;
A third light source for irradiating visible light to a second surface opposite to the first surface of the inspection object;
It arrange | positioned so that the said to-be-inspected object can be imaged with the visible light irradiated from the said 1st light source, and reflected by the said to-be-inspected object, and the visible light irradiated from the said 3rd light source, and permeate | transmitted the to-be-inspected object. A first imaging sensor;
A second imaging sensor arranged so that the inspected object can be imaged by invisible light irradiated from the second light source and reflected by the inspected object;
Pixel values for each of a plurality of color components included in the first image of the inspection object obtained by the first imaging sensor and the second image of the inspection object obtained by the second imaging sensor Using the pixel value,
A state in which there is no abnormality in the inspection object, which is a value that combines the pixel values for each of the plurality of color components in the pixels of the first image corresponding to the position of interest arbitrarily selected on the inspection object The degree of change to the state,
In between the pixel value of the first pixel value of each of the plurality of color components in a pixel of the image and the second image corresponding to the target position, whether the rate of reduction is the same for the normal state Or based on the balance of the plurality of color components in the pixels of the first image corresponding to the target position and the pixel values of the second image corresponding to the target position,
A processing unit for detecting an abnormal part included in the inspection object and discriminating the type of abnormality occurring at the detected abnormal part,
And an output unit that outputs information on an abnormal part including at least information indicating the type of abnormality determined by the processing unit. The output unit includes, as information about the abnormal part, the first image of the area including the abnormal part, the second image of the area including the abnormal part, and the pixel value of the first image at the abnormal part. A graph showing a change in the second image, a graph showing a change in the pixel value of the second image at the abnormal portion,
The sheet inspection apparatus according to claim 1, further outputting at least one of the information. Visible light emitted from the first light source and visible light emitted from the third light source are light having the same spectral distribution including a plurality of color components,
The first imaging sensor is capable of receiving visible light irradiated from the first light source and reflected by the inspection object, and is irradiated from the third light source and passes straight through the inspection object. It is placed at a position where it can receive visible light,
The pixels of the first image have pixel values for a plurality of color components,
The processor is
Usually, the value obtained by combining the pixel values for each of the plurality of color components in the pixels of the first image corresponding to the position of interest arbitrarily selected on the inspection object is in a state where there is no abnormality in the inspection object. When it is larger than the state,
The pixel value of the second image corresponding to the target position is smaller than that in the normal state, and the balance of the plurality of color components in the pixels of the first image corresponding to the target position is the normal value. The sheet inspection apparatus according to claim 1, wherein when the balance in the state is the same, the type of abnormality occurring at the target position is determined to be a pinhole defect. Visible light emitted from the first light source and visible light emitted from the third light source are light having the same spectral distribution including a plurality of color components,
The first imaging sensor is capable of receiving visible light irradiated from the first light source and reflected by the inspection object, and is irradiated from the third light source and passes straight through the inspection object. It is placed at a position where it can receive visible light,
The pixels of the first image have pixel values for a plurality of color components,
The processor is
In a case where the total value of the pixel values for each of the plurality of color components in the pixel of the first image corresponding to the arbitrarily selected position of interest on the inspection object is larger than that in the normal state ,
The pixel value of the second image corresponding to the target position is smaller than that in the normal state, and the balance of the plurality of color components in the pixels of the first image corresponding to the target position is the normal value. If the balance in the state is not the same,
A pixel value of the second image corresponding to the target position is larger than that in a normal state;
The sheet inspection apparatus according to claim 1, wherein the sheet inspection apparatus discriminates a type of abnormality different from a pinhole defect. Visible light emitted from the first light source and visible light emitted from the third light source are light having the same spectral distribution including a plurality of color components,
The first imaging sensor is capable of receiving visible light irradiated from the first light source and reflected by the inspection object, and is irradiated from the third light source and passes straight through the inspection object. It is placed at a position where it can receive visible light,
The pixels of the first image have pixel values for a plurality of color components,
The processor is
In a case where the total value of the pixel values for each of the plurality of color components in the pixel of the first image corresponding to the arbitrarily selected position of interest on the inspection object is smaller than that in the normal state ,
When the rate of decrease with respect to the normal state is the same between the pixel value of each of the plurality of color components in the pixel of the first image corresponding to the target position and the pixel value of the second image The sheet inspection apparatus according to claim 1, wherein the type of abnormality occurring at the target position is determined to be a metal defect. The first imaging sensor that is configured not to receive the invisible light and the second imaging sensor that is configured not to receive visible light are provided in one imaging device. Item 6. The sheet inspection apparatus according to any one of Items 1 to 5. The sheet inspection apparatus according to claim 6, wherein the imaging apparatus includes a spectroscopic element that divides one optical path and guides light to each of the first imaging sensor and the second imaging sensor. .

Images