JP6132126B2 - Translucent object inspection apparatus and translucent object inspection method - Google Patents

Description

  The present invention inspects uneven defects present between plate-like objects or between layers of a light-transmitting object formed by laminating thin plate-like objects or a plate-like light-transmitting object formed from a plurality of layers. The present invention relates to a translucent object inspection apparatus or a translucent object inspection method.

  The touch panel has a thin plate-like glass substrate in close contact with the display, a film-like transparent electrode layer on which an ITO electrode is formed, and a transparent conductive layer or surface layer. It has a structure in which several layers are formed. When a person touches the surface with a finger, the resistance or capacitance of the transparent electrode layer changes, so that an input can be detected and the image displayed on the display can be changed. As described above, the touch panel has a structure in which a plurality of transparent layers are formed on a plate-like glass substrate. However, the glass substrate and the transparent electrode layer are bonded with a transparent adhesive. Is poor, continuous minute irregularities (hereinafter referred to as orange peel defects) are formed in a wide region between the glass substrate and the transparent electrode layer. If there is this orange peel defect, there is a problem that the display on the display looks distorted. Therefore, after bonding a glass substrate and a transparent electrode layer, it is necessary to inspect whether an orange peel defect is formed at the bonded position. As a method for inspecting a translucent object formed from a plurality of layers, as shown in Patent Document 1, light is irradiated obliquely on the surface of the translucent object and the reflected light is visually observed. There is a method of observing or making an image by reflected light incident on a camera and comparing it with a normal one. With this method, the irradiated light is reflected not only at the surface of the translucent object, but also at the boundary between layers, so that if there are irregularities between the layers as well as the surface, it will look normal, Or, there is a difference in images, and abnormality can be detected.

JP 2001-116925 A

  However, the orange peel defect that occurs between the glass substrate of the touch panel and the transparent electrode layer has a very small degree of unevenness, and the reflected light is visually observed by irradiating light obliquely to the surface of the translucent object. Even if it is observed or an image by reflected light is created, there is almost no difference compared to a normal one. Therefore, it is difficult to detect an orange peel defect with the conventional method. For this reason, at present, defects are often found in the inspection after the touch panel is made up to the finished product, and there is a problem that the manufacturing cost increases.

  The present invention has been made to solve the above-described problems, and the object thereof is to provide a translucent object formed by laminating thin plate-like objects or a plate-like translucent object formed of a plurality of layers. To provide a translucent object inspection apparatus or a translucent object inspection method capable of accurately detecting an uneven defect even if the unevenness degree of the uneven defect existing between plate-like objects or between layers is small. It is in.

In order to achieve the above object, a feature of the present invention is that a translucent object formed by laminating thin plate-like objects or a plate-like translucent object formed of a plurality of layers is provided with a translucent object. Light receiving means that emits light so that reflected light is generated from the surface and inside of the light, and light receiving image data in which the reflected light from the translucent object is received by the light receiver, and the light receiving intensity at each light receiving point is lightened Inspecting the presence or absence of defects existing between thin plate-like objects or between layers of translucent objects by analyzing the received light image data created by the image data creation means In the translucent object inspection apparatus, the light emitted from the light irradiating means is made to interfere with the light emitted from the light irradiating means before the light irradiated from the light irradiating means is created by the image data creating means. According to received image data Interference means for creating interference fringes in the image when the image is created, and the received light image data at each light receiving location of the light receiver are differentiated using the received light image data at the light receiving locations around each light receiving location Differential data calculation means for calculating the differential value for each light receiving location, and differential data calculation for each light receiving location of a plurality of lines perpendicular to the interference fringe line defined at a predetermined interval in the interference fringe line direction In the change curve created from the differential value calculated by the means, a point that crosses the line of the preset value is detected as a boundary point, and the detected boundary point is detected for each group that is the same boundary in the line direction of the interference fringes. Conclusion, a boundary variation degree calculating means for calculating a deviation from a straight line of the boundary point as a boundary fluctuation degrees for each group are summarized, the boundary variation degree calculating means When the calculated boundary fluctuation degree for each group is compared with a preset allowable value, and the presence of a defect is determined by a first determination unit that determines the presence or absence of a defect, The variation calculation means for calculating the variation of the boundary fluctuation degree using the boundary fluctuation degree of each group calculated by the boundary fluctuation degree calculation means as a variable, and the variation of the boundary fluctuation degree calculated by the variation calculation means are preset. And a second determination means for comparing the set value and determining whether or not the defect is an uneven defect continuously present in a predetermined range or more .

According to this, the image created based on the received light image data created by the image data creating means by irradiating the light to the translucent object with the light irradiating means, receiving the reflected light from the translucent object, Since the light emitted from the light irradiation means is caused to interfere by the interference fringe creating means, an interference fringe is generated. When the unevenness degree of the uneven defect existing between thin plate-like objects or between layers of the light transmitting object is small, the image of the interference fringe does not show a large difference compared to the case where the uneven defect does not exist. However, the inventor has confirmed that even if the degree of unevenness is small, a large difference appears in the differential image of the interference fringe image. Specifically, it was confirmed that the boundary between the bright part and the dark part in the differential image composed of the stripes of the bright part and the dark part greatly deviated from the straight line. Therefore, the differential data calculation means calculates the differential value differentiated using the received light image data for each light receiving location, and the boundary fluctuation degree calculation means calculates in advance the differential value change curve in the direction perpendicular to the line of the interference fringes. The point where the set value line crosses is detected as a boundary point, the detected boundary point is grouped into groups that are the same boundary in the line direction of the interference fringes, and the deviation of the boundary point from the straight line is grouped for each group. If the degree of boundary fluctuation is calculated, it is possible to accurately detect uneven defects even if the degree of unevenness is small, depending on whether the calculated value exceeds a threshold value. In addition, according to this, uneven defects that exist continuously beyond a predetermined range between thin plate-like objects or layers between light-transmitting objects are distinguished from localized defects and foreign substances adhering to the surface. Can be detected above. Specifically, even when there is a defect localized between the thin plate-like objects or between the layers of the translucent object or between the layers of the translucent object, or when there is a foreign object attached to the surface of the translucent object, There is a possibility that the value of the boundary fluctuation degree calculated by the boundary fluctuation degree calculation means becomes large. However, since the adhesion of localized defects and foreign matters is not continuous, the variation in the boundary fluctuation degree with the boundary fluctuation degree for each group as a variable continuously exists within a predetermined range between plate-like objects or between layers. There is a difference with respect to the uneven defects. In other words, the variation in the boundary fluctuation degree is small in the case of uneven defects that are continuously present within a predetermined range between the plate-like objects or between the layers, and the variation in the boundary fluctuation degree is in the case of adhesion of localized defects or foreign matters. growing. Therefore, when it is determined that there is a defect from the boundary fluctuation degree calculated by the boundary fluctuation degree calculation means, if the variation of the boundary fluctuation degree is calculated by the fluctuation calculation means and compared with the set value, it will be more than a predetermined range. It is possible to detect the concavo-convex defects that exist continuously, while distinguishing them from localized defects and foreign matters adhering to the surface.

  Another feature of the present invention is that, in the above-described present invention, differential image data obtained by converting the differential value for each light receiving location calculated by the differential data calculation means into lightness is generated from the differential image data. It is to display the image.

  According to this, it is possible to detect the irregular defect by looking at the displayed image. It is also possible to detect irregularities that are continuously present in a predetermined range or more after distinguishing them from localized defects and foreign matters adhering to the surface.

  Further, the present invention is a touch panel in which a transparent electrode layer is formed on at least a transparent substrate, and a light-transmitting property for inspecting unevenness continuously existing in a predetermined range or more between the transparent substrate and the transparent electrode layer. It is particularly useful when applied to an object inspection apparatus.

  Furthermore, in carrying out the present invention, the present invention is not limited to the invention of the translucent object inspection apparatus, but can also be implemented as an invention of a translucent object inspection method.

It is a whole block diagram of the translucent object inspection apparatus of this invention. (A) is the figure which showed simply the image of the interference fringe created from the light reception image data of a camera, (B) is the figure which showed the change of the light reception image data in the orthogonal | vertical direction of the line direction of an interference fringe. (C) is the figure which showed simply the image produced from the differential image data which converted the differential value of the light reception image data of a camera into the brightness, (D) is the differential image in the orthogonal | vertical direction of the line direction of an interference fringe. It is the figure which showed the change of data. (A) is the figure which showed the boundary point when an orange peel defect does not exist visually, (B) is the figure which showed the boundary point when an orange peel defect exists visually. It is the figure which showed visually a mode that the inspection location (photographing location of a camera) of a test subject moves. It is the figure which showed visually the process which detects the boundary point in the orthogonal | vertical direction of the line direction of an interference fringe, and identifies each boundary point with the variables p and q. It is the figure which showed the reason why the line of an interference fringe does not become the same in the boundary point of an interference fringe line, even if the variable q is the same. It is a flowchart which shows the movement of the camera in the defect detection program performed by the controller of FIG. It is the first half part of the flowchart which shows the analysis process and determination process of light reception image data in the defect detection program performed by the controller of FIG. FIG. 6 is a second half of a flowchart showing analysis processing and determination processing of received light image data in a defect detection program executed by the controller of FIG. 1.

  Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. FIG. 1 is an overall configuration diagram of a translucent object inspection apparatus according to the present invention. This translucent object inspection apparatus is a plate-like light-transmitting device in which a film-like transparent electrode layer is bonded to a plate-like glass substrate on a touch panel, and several layers such as a conductive layer and a surface layer are formed thereon. It is an apparatus for inspecting a natural object (hereinafter referred to as a flat plate for touch panel) OB, and in particular an apparatus for inspecting an orange peel defect that may occur between a glass substrate and a transparent electrode layer. In the following description, the direction perpendicular to the paper surface in FIG. 1 is referred to as the X direction, and the horizontal direction in FIG. 1 is referred to as the Y direction.

  This translucent object inspection device irradiates light onto the stage 2 for setting the flat plate OB for touch panel on the flat base 3 and the flat plate for touch panel OB set on the stage 2, and receives the reflected light to receive light. A measurement unit 1 that acquires image data is provided. A controller that outputs commands to several circuits and camera C provided in the measurement unit 1 and inputs digital data from the several circuits and camera C provided in the measurement unit 1 to perform program processing. 50. The controller 50 is an electronic control device including a microcomputer including a CPU, ROM, RAM, and the like, a storage device such as a hard disk and a nonvolatile memory, an input / output interface, and the like. The storage device stores various programs including a defect detection program to be described later. Then, the controller 50 shows the input device 52 such as a keyboard and a mouse for inputting commands and data to the controller 50, the operating state of the translucent object inspection device, the inspection result, and the received image data from the camera C. A display device 54 composed of a liquid crystal panel, a CRT, or the like, which displays an image created based on the connection, is connected.

  The measuring unit 1 includes an irradiation optical system including a laser light source 10, a collimating lens 12, expander lenses 14 and 16, a laser diameter adjusting ring 18 and an interference fringe creating unit 20, a camera C as a light receiving optical system, and a camera. And a camera moving mechanism 30 that moves C in two directions orthogonal to each other. The laser light source 10 emits laser light having a set intensity when a drive signal is input from the laser drive circuit 40. The laser drive circuit 40 supplies the laser light source 10 with a voltage and current with the intensity set so that the laser light with the intensity set from the laser light source 10 is emitted when a laser irradiation start command is input from the controller 50. When a laser irradiation stop command is input from 50, supply of voltage and current to the laser light source 10 is stopped. The laser light emitted from the laser light source 10 is highly coherent, and interference fringes are generated at locations where the light is received after being reflected by the interference fringe creating unit 20 described later. The collimating lens 12 makes the laser light emitted from the laser light source 10 parallel light and enters the expander lens 14. The expander lenses 14 and 16 increase the cross-sectional diameter of the laser light and make it incident on the diameter adjusting ring 18. The diameter adjusting ring 18 allows only laser light within a predetermined radius from the circle center of the laser cross section to pass through and enter the interference fringe creating unit 20. The laser light intensity in the radial direction in the cross section of the laser light has a Gaussian distribution, so that the weak laser light near the periphery in the cross section of the laser light is excluded, and only the strong laser light has interference fringes. The light enters the creation unit 20.

  The interference fringe creation unit 20 is a unit in which the prism 22 having a right angle and the surface of the sheet glass 24 are combined through the projections 26 and 28. There are two protrusions 26 having the same height in the direction perpendicular to the paper surface, and the protrusions 28 are slightly different in height from the protrusions 26, and the position in the direction perpendicular to the paper surface is an intermediate position between the two protrusions 26. That is, when viewed from the vertical direction of the surface of the sheet glass 24, the protrusions 26 and 28 are located at the vertices of the isosceles triangle. For this reason, the surface of the prism 22 facing the 90-degree angle and the surface of the glass sheet 24 are slightly deviated from being parallel, and the laser beam is incident on one of the 90-degree angles of the prism 22 perpendicularly. However, when the light is reflected by the surface of the prism 22 facing the 90-degree angle and the surface of the plate glass 24, the reflected two light beams have different optical path differences depending on the location, so that interference fringes are generated. That is, an interference fringe is generated at a location where light emitted from the other surface of the prism 22 forming an angle of 90 degrees is received.

  The stage 2 includes a flat plate 2b attached to the base 3 and a flat plate 2a that forms an angle of 45 degrees with the flat plate 2b. When the touch panel flat plate OB is mounted on the flat plate 2a, the stage 2 supports the touch panel flat plate OB. Protrusions 2c are provided for this purpose. Several protrusions 2c are formed in one line in the X direction and several in one line in the Y direction. When the touch panel flat plate OB is placed in contact with the protrusion, the X direction line and the Y direction of the protrusion 2c are arranged. The corner of the touch panel flat plate OB comes to the point where the lines intersect. The laser light emitted from the laser light source 10 is adjusted to be perpendicular to the base 3, and the surface of the prism 22 facing the 90 ° angle is adjusted to be 45 ° with the base 3. Therefore, the light emitted from the interference fringe creation unit 20 is parallel to the base 3 and enters the touch panel flat plate OB at an angle of 45 degrees. Therefore, the light reflected by the touch panel flat plate OB becomes perpendicular to the base 3 and enters the camera C. Reflection of light on the touch panel flat plate OB occurs between the surface and the inner layers. Then, the stage 2 is arranged so that the interference fringe line formed by the light emitted from the interference fringe creation unit 20 is parallel to the horizontal direction of the flat plate 2a of the stage 2 (the vertical direction in FIG. 1, ie, the X direction). It has been adjusted. Accordingly, since the surface of the touch panel flat plate OB is flat and rectangular, when the touch panel flat plate OB is placed on the flat plate 2a of the stage 2, the line of interference fringes is generated in parallel to the surface lateral direction of the touch panel flat plate OB. It will be.

  In the camera C, an image is formed by an imaging element such as a CCD or CMOS by a condensing lens, and the intensity (corresponding to the brightness at each pixel) output from each pixel of the imaging element is converted into digital data and output. To do. The optical axis of the condenser lens of the camera C is adjusted to be perpendicular to the base 3, and the imaging element is perpendicular to the optical axis of the condenser lens. Further, the optical axis direction position of the condenser lens of the camera C is adjusted so that the center of the depth of field is near the intersection of the surface of the touch panel flat plate OB and the optical axis of the condenser lens. An image of the surface of the touch panel flat plate OB is formed by the light received at. As described above, since the interference fringe is generated at the place where the light emitted from the interference fringe creation unit 20 is received, the interference fringe is generated on the surface of the touch panel flat plate OB, and an image of the surface of the touch panel flat plate OB is formed. Interference fringes also occur in the C image sensor. As described above, the light from the interference fringe creating unit 20 is reflected between the surface of the touch panel flat plate OB and the inner layer, and therefore, the interference fringe generated in the image sensor of the camera C is the surface of the touch panel flat plate OB and the inner layer. This is due to the reflected light from. The vertical and horizontal alignment directions (corresponding to the vertical and horizontal directions of the captured image) of the pixels in the image sensor of the camera C are adjusted to be parallel to the X direction and the Y direction. The line of interference fringes generated in is substantially parallel to the horizontal alignment direction of the pixels of the image sensor. That is, interference fringes are observed in a substantially horizontal direction of the captured image. When a command is input from the controller 50 to be described later, the camera C starts outputting digital data having a value corresponding to the brightness of each pixel to the controller 50. The controller 50 creates image data based on the digital data and outputs it to the display device 54 to display the captured image.

  The camera moving mechanism unit 30 moves the camera C in the X direction and the Y direction. The camera C is fixed to the moving body 32, and the moving body 32 is screwed with a Y-direction screw rod 33 that is driven to rotate around the axis at one end by a Y-direction motor 34. The other end of the Y direction screw rod 33 is rotatably supported by a support base 31, and the support base 31 fixedly supports a Y direction motor 34. The moving body 32 is guided by a guide member (not shown) so as to be movable only in the axial direction of the Y direction screw rod 33, and the moving body 32 and the Y direction screw rod 33 constitute a screw feed mechanism. Therefore, the moving body 32 and the camera C move in the Y direction by the rotation of the Y direction motor 34 around the axis of the Y direction screw rod 33. The support base 31 is screwed with an X-direction screw rod (not shown) that is driven to rotate around the axis at one end by the X-direction motor 36. The other end of the X direction screw rod is rotatably supported by a support base 35, and the support base 35 fixedly supports an X direction motor 36. The support base 31 is guided by a guide member (not shown) so as to be movable only in the axial direction of the X-direction screw rod, and the support base 31 and the X-direction screw rod constitute a screw feed mechanism. Accordingly, the support base 31, the moving body 32, and the camera C move in the X direction by the rotation of the X direction motor 36 around the axis of the X direction screw rod.

  The X direction motor 36 rotates when a drive signal is supplied from the X direction motor drive circuit 44, and the Y direction motor 34 rotates when a drive signal is supplied from the Y direction motor drive circuit 48. The X-direction motor 36 and the Y-direction motor 34 include an encoder 36a and an encoder 34a that output pulse train signals ΦA and ΦB that alternately switch between a high level and a low level each time they rotate by a predetermined minute angle. The pulse train signals ΦA and ΦB are signals that are out of phase with each other by π / 2 in order to identify the rotation direction. The X-direction motor drive circuit 44 and the Y-direction motor drive circuit 48 receive the pulse train signal output from the encoder 36a and the encoder 34a when supplying drive signals to the X-direction motor 36 and the Y-direction motor 34. The drive signal is controlled so that the number of pulses per time becomes a set value. Thereby, the camera C moves at the set constant speed.

  When the X direction position is input from the controller 50, the X direction motor drive circuit 44 until the X direction movement position input from the X direction movement position detection circuit 42 described later becomes the movement position input from the controller 50. A drive signal is supplied to 36. The same applies to the Y-direction motor drive circuit 48. When the Y-direction position is input from the controller 50, the Y-direction movement position input from the Y-direction movement position detection circuit 46, which will be described later, becomes the movement position input from the controller 50. A drive signal is supplied to the Y direction motor 34. Thereby, the camera C moves to the movement position designated by the controller 50.

  The X-direction movement position detection circuit 42 counts up or down the number of pulses of the pulse train signals ΦA and ΦB input from the encoder 36a in the X-direction motor 36 according to the rotation direction of the X-direction motor 36, and from the count value The X direction movement position is calculated, and the digital data of the X direction movement position is output to the X direction motor drive circuit 44 and the controller 50. The origin position (position of movement position 0) of the X direction movement position detected by the X direction movement position detection circuit 42 is set by an instruction from the controller 50 when the power is turned on. That is, when the power is turned on, the controller 50 instructs the X direction motor drive circuit 44 to move the camera C (support base 31) to the origin position, and instructs the X direction movement position detection circuit 42 to set the origin. In response to this instruction, the X direction motor drive circuit 44 rotates the X direction motor 36 to move the camera C (support base 31) in the direction of the origin. In the present embodiment, it is assumed that the back surface direction in the direction perpendicular to the paper surface of FIG. This origin position is the drive limit position of the camera C (support base 31), and the X-direction movement position detection circuit 42 continues to input the pulse train signals ΦA and ΦB from the encoder 36a while the camera C (support base 31) is moving. ing. When the camera C (support 31) reaches the drive limit position (that is, the origin position) and the rotation of the X-direction motor 36 stops, the X-direction movement position detection circuit 42 inputs the pulse train signals ΦA and ΦB from the encoder 36a. Stop is detected and the count value is set to zero. That is, the movement position is set to 0. At the same time, a signal for stopping the output is output to the X direction motor drive circuit 44, whereby the X direction motor drive circuit 44 stops supplying the drive signal to the X direction motor 36. Thereafter, when the X-direction motor 36 is driven, the X-direction movement position detection circuit 42 counts up the number of pulses of the pulse train signals ΦA and ΦB from the encoder 36 a according to the rotation direction of the X-direction motor 36. Count down, calculate the X-direction movement position based on the count value, and output the digital data.

  The Y-direction movement position detection circuit 46 is operated in the same manner as the X-direction movement position detection circuit 42. That is, the number of pulses of the pulse train signals ΦA and ΦB input from the encoder 34a in the Y direction motor 34 is counted up or down according to the rotation direction of the Y direction motor 34, and the Y direction moving position is calculated from the count value. The digital data of the Y direction movement position is output to the Y direction motor drive circuit 48 and the controller 50. The setting of the origin position of the Y-direction movement position detected by the Y-direction movement position detection circuit 46 (the position of the movement position 0) is the same as that of the X-direction movement position detection circuit 42. That is, when the power is turned on, the controller 50 instructs the Y-direction motor drive circuit 48 to move the camera C (the moving body 32) to the origin position and instructs the Y-direction movement position detection circuit 46 to set the origin. In response to this instruction, the Y-direction motor drive circuit 48 rotates the Y-direction motor 34 to move the camera C (the moving body 32) in the origin direction. In the present embodiment, the left direction in FIG. 1 is the origin direction. This origin position is the drive limit position of the camera C (moving body 32), and the Y-direction moving position detection circuit 46 continues to input the pulse train signals ΦA and ΦB from the encoder 34a while the camera C (moving body 32) is moving. ing. When the camera C (the moving body 32) reaches the drive limit position (that is, the origin position) and the rotation of the Y-direction motor 34 stops, the Y-direction movement position detection circuit 46 inputs the pulse train signals ΦA and ΦB from the encoder 34a. Stop is detected and the count value is set to zero. That is, the movement position is set to 0. At the same time, a signal for stopping the output is output to the Y-direction motor drive circuit 48, whereby the Y-direction motor drive circuit 48 stops supplying the drive signal to the Y-direction motor 34. Thereafter, when the Y-direction motor 34 is driven, the Y-direction movement position detection circuit 46 counts up the number of pulses of the pulse train signals ΦA and ΦB from the encoder 34 a according to the rotation direction of the Y-direction motor 34. It counts down, calculates the Y-direction movement position based on the count value, and outputs the digital data.

  In the translucent object inspection apparatus configured as described above, when the controller 50 executes a defect detection program, which will be described later, by input from the input device 52, light from which interference fringes are formed at the light receiving location from the interference fringe creating unit 20 , The captured image data acquired by the camera C is processed, and a determination result as to whether or not the touch panel flat plate OB is defective is displayed on the display device 54. Before explaining the operation of the translucent object inspection apparatus in the inspection, how the captured image data of the camera C is processed and determined will be described. FIG. 2 visually shows a method of processing the captured image data of the camera C performed by the controller 50. When the interference fringes are generated on the touch panel flat plate OB as shown in FIG. 2A, the received light intensity on the line in the vertical direction of the interference fringe lines in the image pickup device of the camera C (that is, in this line of the image pickup device). When a brightness curve is created from image data obtained by converting the signal intensity output from a pixel into digital data, the result is as shown in FIG. FIG. 2C shows a curve obtained by differentiating the lightness curve and making the differential value an absolute value. When the absolute value of this differential value is calculated for all lines in the vertical direction of the interference fringe lines and an image converted to lightness is created, the result is as shown in FIG. Actually, it is not necessary to consider the vertical direction of the line of the interference fringes, and the interference fringes are generated almost in parallel with the arrangement of the pixels of the image sensor of the camera C as described above. Of the four pixels that are equidistant in the arrangement direction, the pixel having the greatest difference in brightness (image data) may be selected to calculate the differential value. In this way, the pixel in the vertical direction of the interference fringe line among the four pixels has the most difference in brightness, so that the image of FIG. 2D can be obtained.

  If the touch panel flat plate OB has no orange peel defect, the boundary line between the bright part and the dark part of the interference fringe is almost a straight line as shown in FIG. 2A, and the bright part in FIG. The boundary line of the dark part is almost a straight line. On the other hand, if there is an orange peel defect in the touch panel flat plate OB, the boundary line between the bright and dark portions of the interference fringes is slightly deviated from the straight line. It is difficult to do. However, the boundary line between the bright part and the dark part in the differential image is greatly deviated from the straight line. Therefore, the lightness curve of FIG. 2C is created for a plurality of lines in the vertical direction of the interference fringe line, and the lightness curve crosses a predetermined value in each lightness curve (corresponding to the boundary between the light part and the dark part). The coordinate point (pixel position of the image sensor) is detected. The coordinate points are grouped for each boundary line between the bright part and the dark part, and a straight line is drawn by the least square method for each group. If the orange peel defect does not exist in the touch panel flat plate OB, the deviation of the coordinate point from the straight line is small as shown in FIG. 3A, but if the orange peel defect exists, the deviation from the straight line as shown in FIG. Therefore, the orange peel defect can be detected from the degree of deviation of the coordinate point from the straight line. That is, if the degree of deviation of the coordinate point from the straight line is obtained as a numerical value, an orange peel defect can be automatically detected depending on whether or not the numerical value is within an allowable value. The numerical value of the degree of deviation of the coordinate point from the straight line is calculated by considering the deviation of each coordinate point from the straight line as a deviation. A calculated correlation coefficient or the like can be used.

  Even if the orange peel defect does not exist on the touch panel flat plate OB, if there is a point-like defect or foreign matter adhering to the surface, the deviation from the straight line of the coordinate point becomes large at that point, the calculated coordinates The numerical value of the degree of deviation of the point from the straight line increases. However, since point-like defects and adhesion of foreign matter to the surface do not exist continuously in a wide range like orange peel defects, coordinates for each group (per boundary line between bright and dark areas) There is a difference in the variation in the degree of deviation of the point from the straight line. That is, in the case of an orange peel defect, the variation is small because it is continuously present in a wide range, and in the case of a spot-like defect or adhesion of foreign matter, the variation is large because it exists locally. Therefore, the variation of the degree of deviation from the straight line of coordinate points for each group is calculated, and depending on whether this value is less than or equal to the set value, whether it is an orange peel defect, a dot-like defect or foreign matter adhering to the surface, Can be determined.

  Next, the operation of the apparatus for inspecting the touch panel flat plate OB in the translucent object inspection apparatus configured as described above will be described. First, the operator turns on a power source (not shown) of the translucent object inspection apparatus. At this time, the controller 50 outputs commands to the X-direction movement position detection circuit 42, the X-direction motor drive circuit 44, the Y-direction movement position detection circuit 46, and the Y-direction motor drive circuit 48 in FIG. The (moving body 32, support base 31) moves to the origin position (drive limit position). Next, the operator sets the touch panel flat plate OB on the flat plate 2 a of the stage 2. At this time, the side surface of the touch panel flat plate OB is brought into contact with the protrusion 2c on the X direction line of the flat plate 2a and the protrusion 2c on the Y direction line. As a result, regardless of the size of the touch panel flat plate OB, when the touch panel flat plate OB is viewed from above, the position of the lower left corner is always a constant position (the point where the X direction line and the Y direction line of the protrusion 2c intersect each other). )become. Then, the operator operates the input device 52 to cause the controller 50 to execute the defect detection program.

  This defect detection program is started in step S100 in FIG. First, the controller 50 sets the variables n and m to “1” in step S102. Variables n and m represent the photographing positions of the camera C. As shown in FIG. 4, n represents the photographing position in the X direction (perpendicular to the paper surface of FIG. 1), and m represents the Y direction (horizontal in FIG. 1). Direction). That is, the shooting position of the camera C is represented from (n, m). As described above, when the touch panel flat plate OB is viewed from above, the position of the lower left corner is always a fixed position, so the shooting position of the camera C represented by (n, m) is fixed. It is stored in the controller 50 in advance.

  In step S104, the controller 50 instructs the laser drive circuit 40 to start laser irradiation. In response to this command, the laser drive circuit 40 supplies the laser light source 10 with the current and voltage having the set intensity, and the laser light source. 10 is irradiated with laser light. The laser light interferes with the interference unit 20 as described above, is reflected by the touch panel flat plate OB, and enters the camera C. Next, the controller 50 commands the camera C to start photographing in step S106, and in response to this command, the camera C outputs digital data corresponding to the brightness value output by each pixel of the image sensor to the controller 50. Start that. Next, in step S108, the controller 50 moves the camera C to the position (1, 1) that is the initial photographing position. As shown in FIG. 4, this position is a position in which the vicinity of the lower left corner of the touch panel flat plate OB is the lower left of the shooting screen of the camera C. Next, the controller 50 captures image data of the camera C (digital data corresponding to the brightness value output by each pixel of the image sensor) in step S110, and creates an image created based on the image data captured in step S112. It is displayed on the display device 54.

  After step S112, the controller 50 analyzes the image data captured in step S112 in step S114, and performs various determinations. This process is performed by starting the program process of the flow shown in FIG. 8A in step S200. Hereinafter, description will be made along the program flow shown in FIG. 8A. First, the controller 50 sets the variable p to “1” in step S202. Assuming a photographed image of the camera C created from image data as shown in FIG. 5, the variable p is assigned to a plurality of lines parallel to the Y direction at equal intervals in the X direction from the side closer to the origin. Represents the assigned number. In other words, the variable p represents a number assigned to pixels parallel to the Y direction in the imaging element of the camera C, and the arrangement of the pixels with respect to the variable p is stored in the controller 50 in advance. Next, the controller 50 extracts the image data of the line of the variable p in step S204. That is, the image data output by the pixels in the line (pixel arrangement) corresponding to the variable p is extracted. In step S206, a brightness curve is created by arranging the image data in pixel order. This brightness curve is the curve shown in FIG.

  After step S206, the controller 50 determines in step S208 whether there is an interference fringe from the brightness curve created in step S206. Specifically, it is determined whether or not the lightness curve crosses a preset level more than a preset number of times. When the crossing is not performed more than the set number of times, when the touch panel flat plate OB is forgotten to be placed on the stage 2, or when the touch panel flat plate OB is not placed at a predetermined position of the stage 2. The case where the thing different from the flat plate OB for touch panels is mounted in the stage 2 etc. can be considered. In this case, it determines with "No" and progresses to step S210. Since the camera C is immediately after the start of photographing, n = 1 and m = 1. In both step S210 and step S212, “Yes” is determined. In step S214, “abnormal” is determined, and the process proceeds to step 284. Then, the image data analysis and various determination processing programs are terminated. On the other hand, if there is an interference fringe and the lightness curve crosses a preset level for a preset number of times, it is determined as “Yes” and the process proceeds to step S220.

  The controller 50 creates differential data of all the image data in step S220. Specifically, for each pixel, image data of pixels that are separated by an equal distance in the horizontal direction and the vertical direction (X direction and Y direction) is extracted, and the lightness value is selected by selecting the pixel having the most different lightness value. Is normalized (for example, a value of 0 to 255). Since the difference in brightness value should be the largest for the pixels in the vertical direction of the interference fringe line, the differential data is a value obtained by differentiating the brightness curve in FIG. When the lines are arranged in the vertical direction (Y direction), a curve shown in FIG. Next, in step S222, the controller 50 causes the display device 54 to display a differential image created based on the differential data created in step S220. The image displayed on the display device 54 is the image shown in FIG. 2D when the touch panel flat plate OB has no defect.

  After the process of step S222, the controller 50 performs the processes of steps S224 to S234 to extract all the bright and dark boundary points of the differential image for each line represented by the variable p, which are indicated by the points in FIG. Perform the process. Specifically, differential data of the line (pixel arrangement) represented by the variable p is extracted in step S224, and a differential curve is created by arranging the differential data in the pixel arrangement order in step S226. In step S228, the Y coordinate value at which the differential curve crosses a predetermined value is calculated and stored. The Y coordinate value is a number assigned to the pixel as 0, 1, 2,... In the pixel arrangement order from the side closer to the origin, and this Y coordinate value is represented by a variable p and a Y coordinate value as shown in FIG. Are stored in correspondence with the variable q attached in order from the smallest value. Then, in step S234, the variable p is incremented, and in step S232, until the variable p reaches the predetermined number B (the total number of lines represented by the variable p) and it is determined “Yes”, steps S224 to S228 are performed. By repeatedly executing the process, the Y coordinate value at which the differential curve crosses the predetermined value is stored in all the lines represented by the variable p. In step S230, it is determined whether or not the Y coordinate value is a very small number A (several). This is because the shooting position of the camera C moves and a touch panel flat plate OB (interference fringe) is formed on a part of the screen. It is assumed that there is no image). If the camera C is in the initial shooting position, it is always determined as “No”. This point will be described in detail later.

  If “Yes” is determined in step S232, the process proceeds to step S236, and in step S236, a variable pmax used in a subsequent step is set to a predetermined number B. This is done because the number of lines represented by the variable p does not become the predetermined number B when the shooting position of the camera C moves and the touch panel flat plate OB (interference fringe image) disappears in a part of the screen. It is. This point will also be described in detail later. In step S242, it is determined whether or not the number of Y coordinate values having the variable p of 1, that is, the number of the differential curve crossing the predetermined value in the leftmost line in FIG. This also assumes a case where the shooting position of the camera C is moved and the touch panel flat plate OB (interference fringe image) disappears in a part of the screen. Is determined. This point will also be described in detail later.

  If “No” is determined in step S242, the process proceeds to step S246, and by the processes in steps S246 to S266, the bright and dark boundary points (Y-coordinate values) are determined for each bright and dark boundary line of the differential image. Is calculated as the boundary variation degree. That is, as shown in FIG. 3, the degree to which the bright and dark boundary points (Y coordinate values) of the differential image fluctuate from the straight line is calculated. Specifically, the following processing is performed. In step S246, the variable q is set to 1. In step S248, the variable p is set to 1. In step S250, the Y coordinate value corresponding to the variable p and the variable q is extracted. Thereby, the Y coordinate value having the smallest value of the leftmost line in FIG. 5 is extracted. Next, in step S252, the variable p is incremented, and in step S254, among the Y coordinate values corresponding to the variable p, the Y coordinate value closest to the Y coordinate values corresponding to the variable p-1 and the variable q. To extract. This is a process of extracting the Y coordinate value closest to the first extracted Y coordinate value in the right one of the leftmost lines in FIG. If the difference between the extracted Y coordinate value and the previously extracted Y coordinate value is within the allowable value in step S256, it is determined as “Yes” and the process proceeds to step S258. Since the variable p is 2, it is determined as “No” in step S258, and the processes of steps S252 to S256 are performed again. This is a process of extracting the Y coordinate value closest to the Y coordinate value extracted in the second line from the left side in the third line from the left side in FIG. The process of extracting the Y coordinate value closest to the Y coordinate value extracted in the previous line in this way until the variable p reaches pmax and “Yes” is determined in step S258. Run repeatedly. As a result, all the Y coordinate values (boundary points) of the bright and dark boundary lines to which the Y coordinate values (boundary points) whose variables p and q are 1 belong are extracted. That is, the boundary point in the lowermost bright and dark boundary line in FIG. 5 is extracted.

  The reason for extracting the Y coordinate value (boundary point) belonging to one boundary line by repeatedly performing the processing of steps S252 to S256 without extracting the Y coordinate value (boundary point) having the same variable q is as follows. This is because the direction of the boundary line between light and dark is not always completely parallel to the X direction. That is, when the direction of the boundary line between light and dark is slightly deviated from the X direction, even if Y coordinate values (boundary points) corresponding to the same value of the variable q are extracted, they do not belong to one boundary line. Because there is a possibility. Visually, as shown in FIG. 6, if the direction of the interference fringe line is deviated from the X direction, the boundary point where the variable q is 1 is the same boundary in the line corresponding to the variables p and a + 1. This is because a boundary point belonging to a line but having a variable q of 1 in a line corresponding to a variable p of a + 2 belongs to another boundary line. In FIG. 6, the deviation of the bright and dark boundary lines from the X direction is exaggerated. However, since the interval and the width of the bright and dark boundary lines are smaller, the bright and dark boundary lines Such a problem occurs even if the deviation from the X direction is very small. Then, the reason for performing the determination in step S256 is that, as shown in FIG. 6, if the difference between the extracted Y coordinate value and the previous extracted Y coordinate value is greater than or equal to the allowable value, the extracted Y coordinate value Is in a case where another boundary line belongs, in this case, it is determined as “No”, and the repetition processing of steps S252 to S258 is interrupted. In this case, calculation of the degree of variation from a straight line, which will be described later, is not performed, and processing at the boundary line corresponding to the next variable q is performed.

  If “Yes” is determined in step S258, the process proceeds to step S260, and in step S260, {a, Y coordinate value (1, 1)}, {2a, Y coordinate value (1, 2)}, { 3a, Y coordinate value (1, 3)}... The regression equation when {pmax · a, Y coordinate value (1, pmax)} is the coordinate value is calculated by the least square method. a is the line-to-line spacing in the line corresponding to the variable p. If the deviation between the bright and dark boundary lines from the X direction is very small, all the Y coordinate values may be averaged. However, if the inspection accuracy is to be improved, it is desirable to calculate a regression equation. In step S262, {a, Y coordinate value (1, 1), {2a, Y coordinate value (1, 2)}, {3a, Y coordinate value (1, 3)}... {Pmax · The distance from each coordinate value of a, Y coordinate value (1, pmax)} to the straight line indicated by the regression equation is calculated, and the calculated distance is regarded as the deviation, and the standard deviation (q) is calculated. This value is the boundary fluctuation degree. In this case, q = 1 (lowermost boundary line), so the standard deviation (1).

  When the calculation in step S262 ends, the process proceeds to step S264, and it is determined in step S264 whether the variable q has reached the maximum value qmax. qmax is the number of Y coordinate values (boundary points) when the variable p is 1. Since the variable q is 1 at this time, it is determined “No”, the variable q is incremented in step S266, the process returns to step S248, and the processes of steps S248 to S262 are performed again. In this case, the standard deviation (b) is defined as the boundary fluctuation degree of the boundary point between the bright and dark boundary points when the variable q is 2, that is, the boundary fluctuation degree of the boundary point belonging to the second bright and dark boundary line from the lower side of FIG. 2) is calculated. When the calculation is completed, it is again determined as “No” in step S264, the variable q is incremented in step S266, and the process returns to step S248. Such a process is performed until the variable q increases by 1 until the variable q reaches the maximum value qmax and it is determined as “Yes” in step S264. As a result, standard deviation (1), standard deviation (2), standard deviation (3)... Standard deviation (qmax) are calculated as the degree of boundary fluctuation in all the light and dark boundary lines.

  If “Yes” is determined in step S 264, the process proceeds to step S 268, and the standard deviation (1), standard deviation (2), standard deviation (3), which are the boundary fluctuation degrees, calculated in the above-described processing. The average value of the standard deviation (qmax) is calculated, stored as the average boundary fluctuation degree (n, m), and displayed on the display device 54. The maximum value of standard deviation (1), standard deviation (2), standard deviation (3)... Standard deviation (qmax) is extracted and stored as the maximum boundary fluctuation degree (n, m) and displayed. It is displayed on the device 54. Then, in step S270 and step S272, it is determined whether the calculated average boundary fluctuation degree (n, m) and maximum boundary fluctuation degree (n, m) are less than the allowable value. “Yes” is determined, and the process proceeds to step S 274 where “normal” is displayed on the display device 54. On the other hand, if either the average boundary fluctuation degree (n, m) or the maximum boundary fluctuation degree (n, m) is larger than the allowable value, the process proceeds to step S276. In step S276, the standard deviation of the standard deviation (1), standard deviation (2), standard deviation (3)... Standard deviation (qmax), which are the boundary fluctuation degrees, is calculated, and the boundary fluctuation degree variation (n , M) and displayed on the display device 54. In step S278, it is determined whether the boundary variation degree variation (n, m) is greater than or equal to a set value. If it is less than the set value, it is determined as “No” and the process proceeds to step S280. In this case, since the standard deviation, which is the degree of boundary fluctuation, is the same value for all light and dark boundary lines, it is determined that this is a defect that continuously exists in a wide range. If the inspection object is the flat panel OB for the touch panel, since such a defect is an orange peel defect, “orange peel defect” is displayed on the display device 54 in step S280. On the other hand, if the boundary variation degree variation (n, m) is equal to or larger than the set value, it is determined as “Yes” and the process proceeds to step S282. In this case, since there are light and dark boundary lines with a large standard deviation, which is the degree of boundary fluctuation, and a small light and dark boundary line, this is determined to be a locally existing defect. The Therefore, “local defect” is displayed on the display device 54 in step S282. As described above, when any one of “normal”, “orange peel defect”, and “local defect” is displayed on the display device 54, the process proceeds to step S284, and the execution of the subprogram ends.

  When the execution of the subprogram ends, the program of the flow of FIG. 7 proceeds from step S114 to step S116, and it is determined whether or not an “abnormal” determination has been made in step S116. In the subprogram of the flow of FIG. 8A, the process proceeds to step S214 to determine whether it is determined as “abnormal”. If there is an “abnormal” determination, the determination is “Yes” and processing proceeds to step S136. In this case, as described above, since the interference fringes are not photographed even though the camera C is in the initial photographing position, “measurement impossible” is displayed in step S136, and step S138 and In step S140, the photographing stop of the camera C and the light irradiation from the measuring unit 1 are stopped, and the execution of the program is ended. Then, the operator investigates the cause of the measurement failure. On the other hand, if there is no “abnormal” determination, it is determined “No” and the process proceeds to step S118, where it is determined whether or not the shooting position of the camera C has reached the maximum position in the X direction. As will be described later, in the subprogram of the flow of FIG. 8A, there is a process for detecting that the photographing position of the camera C has reached the maximum position in the X direction, and it is determined whether or not the detection has been performed. . In this case, since the camera C is at the initial shooting position, the determination is “No” and the process proceeds to step S120, and the variable n representing the shooting position in the X direction is incremented and the process proceeds to step S122. In step S122, the camera C is moved to the shooting position represented by n = 2 and m = 1, and the process returns to step S110 to capture image data from the camera C, and the same processing as described above is executed.

  When the camera C is at the shooting position represented by n = 2 and m = 1, the processing of the subprogram in the flow of FIG. 8A is performed at the initial shooting position represented by n = 1 and m = 1. Same as there are. Accordingly, “No” is similarly determined in step S118, n = 3 in step S120, and the process proceeds to step S122. The camera C moves to the photographing position represented by n = 3 and m = 1. In this way, the photographing position of the camera C moves to the right in the X direction as shown in FIG. 4 every time the variable n increases by one. In FIG. 4, n = 4, but when the shooting position of the camera C moves to the right in the X direction, the touch panel flat plate OB (interference fringes) may not be shot on the right side of the shot image. Occur. In this case, if the variable p is incremented in step S234 in the subprogram of the flow of FIG. 8A, “Yes” may be determined in step S230. That is, when the line represented by the variable p shown in FIG. 6 moves to the right, the differential curve does not cross the predetermined value at a location where the touch panel flat plate OB (interference fringes) is not photographed. The coordinate value (boundary point) becomes 0, and “Yes” is determined in step S230. In this case, the process proceeds to step S238, and it is determined that the shooting position is the maximum position in the X direction. In step S240, a value obtained by subtracting 1 from the current value of p is set as pmax. This is a process of setting pmax at the maximum p among the lines where the differential curve crosses a predetermined value (the line represented by p and the touch panel flat plate OB (interference fringes)). The other processes in the subprogram of the flow of FIG. 8 are the same as when the camera C is at the initial photographing position represented by n = 1 and m = 1.

  As described above, when it is detected by the subprogram of the flow of FIG. 8A that the photographing position has reached the maximum position in the X direction, in the loop processing of steps S110 to S122 of the program of the flow of FIG. In step S124, it is determined whether the shooting position of the camera C has reached the maximum position in the Y direction. As will be described later, in the subprogram of the flow of FIG. 8A, there is a process for detecting that the photographing position of the camera C has reached the maximum position in the Y direction, and it is determined whether or not the detection has been performed. . In this case, since the camera C is at the maximum position in the X direction but not at the maximum position in the Y direction, the determination is “No” and the process proceeds to step S126, and the variable m indicating the shooting position in the Y direction is incremented to “2”. Then, the process proceeds to step S128, and the variable n representing the photographing position in the X direction is set to “1”. Then, by proceeding from step S128 to step S122, the process returns to the loop process of steps S110 to S122. Since n = 1 and m = 2 at this time, the shooting position of the camera C is one higher than the initial shooting position as shown in FIG. Then, by repeating the loop processing of steps S110 to S122, the shooting position of the camera C moves rightward from the shooting position one above the initial shooting position, and the above-described processing is performed at each shooting position. The process when the shooting position is the maximum position in the X direction is the same as the process when the shooting position is the maximum position in the X direction when m = 1.

  Each time it is detected that the shooting position is the maximum position in the X direction, m is incremented and n is set to “1”, so that the shooting position of the camera C becomes the maximum position in the X direction. The X direction returns to the same position as the initial shooting position, and the Y direction moves upward. In FIG. 4, m = 3, but when the shooting position of the camera C moves upward in the Y direction, the touch panel flat plate OB (interference fringes) may not be shot on the upper part of the shot image. Occur. In this case, in the subprogram of the flow of FIG. 8A, the number of Y coordinate values (boundary points) in the line (leftmost line) corresponding to the variable p = 1 in step S242 (the differential curve crosses the set value). Number) is smaller than a predetermined number C (the number of Y coordinate values when there are interference fringes in all of the shooting screens in the Y direction). Therefore, the determination is “Yes” and the process proceeds to step S244, where the shooting position is the maximum in the Y direction. It is determined that the position has been reached. Other processes are the same as those described above. However, even if it is detected that the shooting position has reached the maximum position in the Y direction, the program of the flow in FIG. 7 will be used unless the shooting position has become the maximum position in the X direction and it is determined “Yes” in step S118. The loop processing of S110 to S122 is not changed. When the shooting position becomes the maximum position in the X direction and the maximum position in the Y direction, it is detected in the subprogram of the flow in FIG. 8A that the shooting position is the maximum position in both the X direction and the Y direction. Thereby, after the subprogram of the flow of FIG. 8A and FIG. 8B ends, in the program of the flow of FIG. 7, “Yes” is determined in step S118 and “Yes” in step S124, and steps S110 to S122 are performed. Exit from the loop process and proceed to S130.

  As shown in FIG. 4, if the touch panel flat plate OB (interference fringes) is photographed in the photographing screen of the camera C when the photographing position of the camera C reaches the maximum position in the X direction or the Y direction. No problem. However, when the horizontal length and the vertical length of the flat plate OB for the touch panel are divisible by the horizontal length and the vertical length of the shooting area of the camera C, the maximum position in the X direction and the Y direction in the above-described processing. Cannot be detected. This is because even if the shooting position reaches the maximum position, the shooting screen does not change from the other shooting positions, and when the shooting position moves in the X direction or Y direction from that position, the touch screen flat plate OB ( This is because the (interference fringes) are not photographed. In order to detect the maximum position in the X direction and the Y direction in this case (exactly, detection of having passed the maximum position), the subprogram of the flow in FIG. 8A includes the processes of steps S208 to S218. That is, if no interference fringe is detected on the leftmost line (line corresponding to p = 1) on the shooting screen, “No” is determined in step S208, and the process proceeds to step S210, but the initial shooting position (n = 1). , When m = 1), when “No” is determined, as described above, there is some abnormality. However, when it is determined “No” at the shooting position that is not the initial shooting position, it is highly likely that the touch panel flat plate OB (interference fringes) has not been shot on the shooting screen because the shooting position has passed the maximum position. Accordingly, if “No” is determined when n = 1, the process proceeds from step S210 to step S218 regardless of the value of m, and the imaging position is determined to be the maximum position in the X direction. If it is determined “No” when n = 1 and m = 1, the process proceeds to step S210, step S212, and step S216. In steps S216 and S218, the shooting position is set in the Y direction, X It is determined that the direction is the maximum position. In this case, the actual shooting position is the shooting position next to the maximum position. However, if “No” is determined in step S208, the subprogram of the flow in FIG. 8A ends after the above-described determination. Since the process proceeds to step S116 in the flow of FIG. 7, the camera C can be immediately moved to the next shooting position. Alternatively, the loop process of steps S110 to S122 can be finished immediately and the process can proceed to step S130. Therefore, there is almost no time loss.

  At the stage of proceeding to step S130, the determination result of “normal”, “orange peel defect”, or “local defect” is displayed on the display device 54 at all the photographing positions of the camera C. In step S130, it is determined whether or not all photographing positions are normal. If it is normal, “Yes” is determined and “pass” is displayed on the display device 54 in step S130. On the other hand, if there is any photographing position determined as “orange peel defect” or “local defect”, “fail” is displayed on the display device 54. After this display, in step S138 and step S140, the photographing stop of the camera C and the light irradiation from the measurement unit 1 are stopped, and the execution of the program is ended. The operator looks at the displayed inspection result, moves the inspected touch panel flat plate OB to a corresponding place, sets the next touch panel for test touch panel OB on the stage 2, and again performs the controller 50 by the above-described operation. Causes the program described above to be executed. In addition, when it is determined as “fail” and the cause is “local defect” in any of the photographing positions, there is a possibility that foreign matter may adhere to the surface of the touch panel flat plate OB. Therefore, the operator displays the stored differential image, determines whether or not foreign matter is attached, and when it is determined that the possibility of foreign matter attachment is high, the surface is cleaned and then inspected again. You may do it.

  According to the embodiment of the translucent object inspection apparatus that operates as described above, the touch panel flat plate OB is irradiated with light from the measurement unit 1, and the reflected light from the touch panel flat plate OB is received by the camera C. In the image created by the controller 50 based on the image data output by C, interference fringes are generated because the interference fringe creation unit 20 interferes with the laser light emitted from the laser light source 10. Even if an orange peel defect exists between the glass substrate of the flat plate OB for the touch panel and the transparent electrode layer, the interference fringe image does not show a great difference compared to the case where the orange peel defect does not exist. However, when an orange peel defect exists, a large difference appears in the differential image of the interference fringe image. Specifically, the boundary between the bright part and the dark part in the differential image composed of the stripes of the bright part and the dark part greatly deviates from the straight line. Therefore, the differential data differentiated using the image data is calculated for each pixel by the program processing in the controller 50, and similarly, the change in the differential value in the direction perpendicular to the line of the interference fringes is calculated by the program processing in the controller 50. Detects a point that crosses a preset value line in the curve as a boundary point, summarizes the detected boundary points into groups that are the same boundary in the line direction of the interference fringes, and from the straight line of the boundary points for each group If the deviation is calculated as the degree of boundary fluctuation, an orange peel defect can be detected with high accuracy depending on whether or not the calculated value exceeds a threshold value.

  Further, according to the embodiment of the translucent object inspection device that operates as described above, the orange peel defect that may exist between the glass substrate of the touch panel flat plate OB and the transparent electrode layer is localized. It can be detected after being distinguished from defects and foreign matter adhering to the surface. That is, even when there is a defect localized between the surface of the touch panel flat plate OB, the glass substrate of the touch panel flat plate OB, and the transparent electrode layer, or when there is foreign matter attached to the surface of the touch panel flat plate OB, the controller 50. There is a possibility that the value of the boundary fluctuation degree calculated by the program processing in will increase. However, since the attachment of localized defects and foreign substances is not continuous, the variation in the degree of boundary fluctuation differs from the orange peel defect. That is, the variation in the boundary variation degree is small in the case of an orange peel defect, and the variation in boundary variation degree is large in the case of adhesion of a localized defect or a foreign substance. Therefore, if the variation of the boundary variation degree is calculated by the program processing in the controller 50, the orange peel defect is attached to the localized defect and the surface by determining using the boundary variation degree and the variation of the boundary variation degree. It can be detected after distinguishing it from foreign objects.

  In addition, according to the embodiment of the translucent object inspection device that operates as described above, an image created based on data obtained by differentiating image data is displayed. Defects can be detected. In addition, there is a big difference in the images created based on the differentiated image data between the orange peel defect and the localized defect and foreign matter adhering to the surface. It is also possible to detect the foreign matter adhering to the surface while distinguishing it.

  The embodiments of the present invention have been described above. However, the present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the object of the present invention.

  In the above embodiment, the imaging area of the camera C is moved, the image data captured from the camera C is processed in all areas of the touch panel flat plate OB, and the presence / absence of a defect is determined for each imaging area. However, since the orange peel defect occurs in a wide range in most cases, instead of this, only one central portion of the touch panel flat plate OB is photographed with the camera C, the image data is processed, and the presence or absence of the defect is detected. You may make it determine.

  Further, in the above-described embodiment, the degree of deviation of the boundary line from the straight line of the coordinate point group is obtained by calculating the standard deviation by regarding the deviation from the straight line of each coordinate point as a deviation. However, any value may be calculated in place of the standard deviation as long as it represents a deviation from the straight line. For example, the variance and the average deviation may be calculated, or the correlation coefficient may be calculated in accordance with the regression equation calculated by the least square method and used.

  Further, in the above embodiment, the degree of deviation from the straight line of the coordinate point group of the boundary line is calculated as the boundary fluctuation degree, and when the boundary fluctuation degree exceeds the allowable value, the variation of the boundary fluctuation degree is calculated, and the defect is detected. Whether it was an orange peel defect, a local defect or a foreign matter adhesion was determined. However, if the probability of occurrence of local defects or foreign matter adhesion is extremely low, only the boundary fluctuation degree may be calculated, and if this value exceeds the allowable value, it may be determined that an orange peel defect exists.

  Moreover, in the said embodiment, the interference fringe preparation part which shifted | deviated from the parallel right-angled surface and the surface of a sheet glass slightly from the parallel interferes with the location which received the light reflected by the flat plate for touchscreens. Stripes were generated. However, the interference fringe creation unit may have any configuration as long as interference fringes can be generated at locations where the light reflected by the touch panel flat plate is received. For example, a plurality of slits may be provided, or a diffraction grating may be used.

  Moreover, in the said embodiment, the translucent object test | inspected with a translucent object inspection apparatus was made into the flat plate for touch panels. However, for inspection of defects existing between or between layers of translucent objects formed by laminating thin plate objects or plate-shaped translucent objects formed from multiple layers, The sex object may be anything.

DESCRIPTION OF SYMBOLS 1 ... Measuring part, 2 ... Stage, 3 ... Base, C ... Camera, OB ... Touch panel flat plate, 10 ... Laser light source, 12 ... Collimating lens, 14, 16 ... Expander lens, 18 ... Laser diameter adjustment ring , 20 ... interference fringe creation section, 22 ... prism, 24 ... plate glass, 26 and 28 ... projection, 30 ... camera moving mechanism section, 31 ... support base, 32 ... moving body, 33 ... Y-direction screw rod, 34 ... Y direction motor, 35 ... support base, 36 ... X direction motor, 40 ... laser drive circuit, 42 ... X direction movement position detection circuit, 44 ... X direction motor drive circuit, 46 ... Y direction movement position detection circuit, 48 ... Y Direction motor drive circuit, 50 ... controller, 52 ... input device, 54 ... display device

Claims (4)

Light is applied to a translucent object formed by laminating thin plate-like objects or a plate-like translucent object formed of a plurality of layers so that reflected light is generated from the surface and inside of the translucent object. Light irradiation means for irradiating;
Receiving the reflected light from the translucent object with a photoreceiver, and comprising image data creating means for creating received light image data with the light receiving intensity of each light receiving location as brightness,
In a translucent object inspection apparatus that inspects for the presence or absence of defects existing between thin plate-like objects or between layers of the translucent object by analyzing the received light image data created by the image data creating means ,
Before the light irradiated from the light irradiating means is irradiated onto the translucent object, the light irradiated from the light irradiating means is caused to interfere, and an image is formed by the received light image data generated by the image data generating means. Interference means for creating interference fringes in the image when created;
Differential data calculation means for calculating a differential value obtained by differentiating the received light image data at each light receiving location of the light receiver using the received light image data at the light receiving locations around the respective light receiving locations for each of the light receiving locations. When,
In the change curve created from the differential value calculated by the differential data calculation means for each of the light receiving points of a plurality of lines perpendicular to the interference fringe line determined at a predetermined interval in the line direction of the interference fringe, Detecting a point that crosses a set value line as a boundary point, collecting the detected boundary point into groups that are the same boundary in the line direction of the interference fringes, and straight lines of the boundary points for each group Boundary variation degree calculation means for calculating a deviation from the boundary variation degree;
A first determination unit that compares the boundary variation degree for each group calculated by the boundary variation degree calculation unit with a preset allowable value and determines the presence or absence of the defect;
When the first determination unit determines that there is a defect, a variation calculation unit that calculates the variation of the boundary variation degree using the boundary variation degree of each group calculated by the boundary variation degree calculation unit as a variable;
The variation of the boundary fluctuation degree calculated by the variation calculating means is compared with a preset set value to determine whether the defect is an uneven defect continuously existing in a predetermined range or more. An inspection apparatus for translucent objects, comprising: 2 determination means.   Creating differential image data obtained by converting the differential value for each light receiving location calculated by the differential data calculation means into lightness, and creating and displaying an image created from the differential image data. The translucent object inspection apparatus according to claim 1.   The translucent object is a touch panel in which a transparent electrode layer is formed on at least a transparent substrate, and the defect to be inspected is unevenness continuously existing in a predetermined range or more between the transparent substrate and the transparent electrode layer. The translucent object inspection apparatus according to claim 1, wherein the inspection apparatus is a transparent object. Light is applied to a translucent object formed by laminating thin plate-like objects or a plate-like translucent object formed of a plurality of layers so that reflected light is generated from the surface and inside of the translucent object. A light irradiation step for irradiating;
Receiving the reflected light from the translucent object with a light receiver, and creating received light image data in which the received light intensity of each light receiving location is light, and an image data creation step,
Translucent object inspection method for inspecting for presence or absence of defects existing between thin plate-like objects or between layers of the translucent object by analyzing the received light image data created in the image data creating step In
The light applied to the translucent object in the light irradiation step is interfered before being applied to the translucent object, and an image is created from the received light image data created in the image data creation step. Is light that causes interference fringes to appear in the image,
Differential data calculation step of calculating a differential value obtained by differentiating the received light image data at each light receiving location of the light receiver using the received light image data at the light receiving locations around the respective light receiving locations for each of the light receiving locations. When,
In the change curve created from the differential value calculated by the differential data calculation step for each of the light receiving points of a plurality of lines perpendicular to the line of the interference fringe determined at a predetermined interval in the line direction of the interference fringe, Detecting a point that crosses a set value line as a boundary point, collecting the detected boundary point into groups that are the same boundary in the line direction of the interference fringes, and straight lines of the boundary points for each group A boundary variation degree calculation step for calculating a deviation from the boundary variation degree;
A first determination step of comparing the boundary variation degree for each group calculated in the boundary variation degree calculation step with a preset allowable value and determining the presence or absence of the defect;
When it is determined that there is a defect in the first determination step, a variation calculation step for calculating a variation in the boundary variation degree using the boundary variation degree for each group calculated in the boundary variation degree calculation step as a variable;
The variation of the boundary variation degree calculated in the variation calculation step is compared with a preset set value to determine whether or not the defect is an uneven defect continuously existing in a predetermined range or more. 2 judgment steps
A method for inspecting a translucent object, comprising: