JP2004109106A - Method and apparatus for inspecting surface defect - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To simultaneously detect ruggedness defects and undulations by using a time-delay and integration (TDI) line CCD. <P>SOLUTION: While a surface to be inspected, which is regarded as a mirror, is moved in a direction perpendicular to the TDI line CCD 26, a check pattern 23 is projected onto the surface of an object to be inspected 31 and is reflected, and the reflected image is observed by the CCD 26. The check pattern is projected onto the surface to be inspected such that the striped pattern of the reflected image becomes oblique to the moving direction of the surface to be inspected. When a minimum unit of the repetition of alternating bright and dark parts repeated in the moving direction of the surface to be inspected is set to be one period, a luminous amount accumulating range of the CCD 26 is set to be, for example, equivalent to one and a half periods. By comparing a signal extracted from the CCD 26 to a threshold value, convex defects are detected. By comparing a signal intensity pattern extracted from the CCD 26 to a reference pattern, undulations and recess defects are detected. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a surface defect inspection method and a surface defect inspection apparatus for inspecting defects such as minute irregularities existing on a flat surface and a wide range of thickness abnormalities, and particularly to an appearance inspection of a display panel during the production of a plasma display device. Also, the present invention relates to a technique suitable for being applied to a substrate inspection of an intermediate layer during the production of a ceramic substrate.
[0002]
The manufacturing process of the plasma display device includes a process of applying a paste-like glass powder to a glass plate. The surface to which the glass powder is applied must be flat, but small convex portions (hereinafter, referred to as convex defects) and concave portions (hereinafter, referred to as concave defects), or a wide range of thickness abnormalities (hereinafter, referred to as undulations) Defects may exist. Such a defect eventually becomes a defect of the display element. Therefore, it is necessary to inspect the coated surface immediately after the application of the glass powder to eliminate the substrate having the above-described defect. If the substrate removed at this stage is reprocessed, it can be commercialized, and the yield is improved.
[0003]
[Prior art]
2. Description of the Related Art Conventionally, a defect inspection on a surface to which glass powder is applied has been performed visually. Since no pattern is formed on the surface to which the glass powder is applied, it is relatively easy to find a large flaw having a depth of about several hundred μm.
[0004]
By the way, as a method of inspecting unevenness of a flat surface, as shown in FIG. 31, a method of uniformly irradiating light of an illumination light source 2 on an inspection target surface 1 and observing a shadow formed at that time by a television camera 3 or the like. is there. In addition, instead of applying uniform illumination light, a method is also known in which a light beam is scanned on the inspection target surface 1 and the distribution of the amount of reflected light is observed. In such a method, in the case of the convex defect 4, the shadow 5 is formed on the side opposite to the light irradiation direction, so that it can be detected. In addition, as an example, the defect 4 shown as “convex defect” in FIG. 31 is one in which the center of the recessed portion 6 is raised in a convex shape, and a shadow 7 is also formed on the recessed portion 6.
[0005]
In addition, as a method of measuring the undulation, there is a moire interference method, an optical interference method, or the like. In these methods, the lattice fringes projected on the inspection target surface are observed by a television camera or the like. However, as shown in FIG. it can. When a line CCD is used in place of a television camera or the like, as shown in FIG. 33, a portion 11 having a high signal intensity corresponding to a bright portion of a grid pattern and a portion having a low signal intensity corresponding to a dark portion of a grid pattern are provided. 12 is obtained.
[0006]
In this signal pattern, in a portion corresponding to the undulating portion 9 (see FIG. 32), as shown by reference numeral 13 in FIG. 33, the interval of the portion having the higher signal strength is wider than the original interval shown by the dotted line. (May be narrower). In addition, since the convex defect indicated by reference numeral 14 in FIG. 32 is located in the bright part of the lattice fringe, the signal pattern shown in FIG. As described above, the moire interferometry or the like can detect a swell or a convex defect in a bright portion of a lattice fringe. In addition, a method of projecting lattice fringes perpendicular to the moving direction on a flat surface of the inspection object to inspect the surface for irregularities (for example, see Patent Document 1), or rotating a cylindrical inspection object by rotating the inspection object 2. Description of the Related Art A method has been known in which an outer peripheral surface is rotated around an axis, and lattice fringes perpendicular to the rotation axis are projected on the outer peripheral surface to check for irregularities on the outer peripheral surface (for example, see Patent Document 2).
[0007]
[Patent Document 1]
JP 2001-349716 A
[0008]
[Patent Document 2]
JP-A-2002-148029
[0009]
[Problems to be solved by the invention]
However, in the above-described visual inspection, there is a problem in that the detection of a concave defect having a depth of several μm varies due to a difference in the skill of the inspector. Further, in the method of observing a shadow formed on the inspection target surface as shown in FIG. 31, even if light is applied to a swell having a small change in the surface angle (indicated by reference numeral 8 in FIG. 31) or a gentle concave defect, the shadow is not affected. However, there is a problem that the detection cannot be performed because the detection does not occur.
[0010]
Further, when a television camera is used in the moire interference method or the like, the number of pixels of the camera is about 500 × 500 pixels, and the detectable area is too small. Not suitable. When a line CCD is used in the moiré interferometry or the like, a defect in a dark portion of a lattice fringe as indicated by reference numeral 16 in FIG. 32 has a signal intensity originally obtained from the dark portion. There is a problem that it is hidden and cannot be detected. In FIG. 32, since the dark portions of the checkerboard are black, the defects 16 in the dark portions are shown in white.
[0011]
In addition, when the line CCD is used in the moiré interferometry or the like, if the moving speed of the inspection object changes or the position of the inspection object surface changes in the vertical direction due to vibration during movement, as shown in FIG. At the time of the change, the whole shape of the lattice fringe 10 changes. If a local distortion 19 of the lattice fringe shape due to a defect is absorbed in the entire shape change portion 18 (a part surrounded by a broken line) of the lattice fringe 10, it is difficult to extract only the lattice fringe shape change due to the defect. Therefore, there is a problem that the defect cannot be detected.
[0012]
The present invention has been made in view of the above problems, and it is possible to simultaneously detect unevenness defects and undulations using a storage type (TDI) line CCD in which a plurality of line sensors are arranged in parallel. It is an object of the present invention to provide a possible surface defect inspection method and a surface defect inspection device. Another object of the present invention is to use a line CCD having a configuration in which a plurality of line sensors are arranged in parallel, and to detect surface defects without being affected by the moving speed or vertical movement of the inspection object. It is an object of the present invention to provide a possible surface defect inspection method and a surface defect inspection device.
[0013]
[Means for Solving the Problems]
In order to achieve the above object, a surface defect inspection method and a surface defect inspection apparatus according to the present invention project a lattice fringe on an inspection target surface while moving an inspection target surface, which looks like a mirror, in a direction orthogonal to a TDI line CCD. When the light quantity of the reflected image is accumulated by the TDI line CCD, a lattice fringe is projected on the inspection target surface such that the stripe pattern of the reflection image is oblique to the moving direction of the inspection target surface, for example, 45 °. I do.
[0014]
When the minimum unit of repetition of a bright portion and a dark portion repeated in the direction of movement of the inspection target surface of the reflected image is one cycle, the accumulation range of the TDI line CCD is n times as large as one cycle (n is a natural number). Is added to the phase amount α of less than one cycle, for example, 1.5 cycles. Then, a threshold value of the signal intensity is set in advance, and this threshold value is compared with the signal intensity actually obtained from each pixel of the TDI line CCD.
[0015]
In addition, a standard signal intensity pattern is set in advance, and this standard pattern is compared with a signal intensity pattern actually obtained from each pixel of the TDI line CCD. When the inspection target surface is not a reflection surface such as a mirror, the lattice fringes may be directly projected on the inspection target surface, and the lattice fringes reflected on the inspection target surface may be observed by the TDI line CCD.
[0016]
According to the present invention, a signal in which an AC component is superimposed on a DC component is obtained from the TDI line CCD. If a convex defect or a minute concave defect is present, the DC component of the signal obtained from each pixel of the TDI line CCD is detected. The intensity drops below the threshold. Also, if there is undulation or a gentle concave defect, the phase of the signal intensity pattern obtained from each pixel of the TDI line CCD will be out of phase with the phase of the standard pattern.
[0017]
In the above-described invention, n general line CDDs may be used instead of the TDI line CCD. In this case, if the periodic signals corresponding to the repetition of the bright portion and the dark portion of the grid pattern obtained from each line CCD are arranged close to and parallel to each other so as to be shifted by a phase difference of 360 / n degrees. Good.
[0018]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
[0019]
(Embodiment 1)
FIG. 1 is a schematic diagram showing a configuration of an inspection apparatus used for performing a surface defect inspection method according to the first embodiment of the present invention. As shown in FIG. 1, the inspection apparatus includes a moving stage (not shown) on which an inspection object 31 is placed, an illumination light source 22, a lattice plate 24 on which lattice fringes 23 are formed, an imaging lens 25, and a TDI line CCD 26. I have.
[0020]
The lattice fringes 23 of the lattice plate 24 are projected on the surface of the inspection object 31 by the illumination light source 22. When the surface of the inspection object 31 is a surface to which the glass paste is applied, the surface of the inspection object 31 becomes a mirror, so that the projected lattice fringes are reflected on the surface of the inspection object 31. The reflected light passes through the imaging lens 25 and forms an image on the sensor surface of the TDI line CCD 26. In this imaging optical system, the object point is the lattice fringe 23 of the lattice plate 24, and the image point is the sensor surface of the TDI line CCD 26.
[0021]
Although not particularly shown, the moving stage moves at a constant speed in a direction orthogonal to the TDI line CCD 26 by a transport mechanism. Along with this, the inspection object 31 on the moving stage also moves at a constant speed in a direction orthogonal to the TDI line CCD 26 (the direction indicated by the arrow A in FIG. 1), whereby the surface of the inspection object 31 is moved. The whole is inspected.
[0022]
The illumination light source 22 is a linear light source, and its length corresponds to the width of the inspection object 31. Here, the width of the inspection object 31 is the length of a side of the rectangular inspection object 31 in a direction orthogonal to the moving direction. As the illumination light source 22, for example, a fluorescent lamp, a rod lens with a reflective tape, an optical fiber bundle, or the like is used.
[0023]
On the lattice plate 24, the stripe pattern when the lattice fringes 23 are projected on the surface of the inspection target 31 is oblique to the moving direction of the inspection target 31, and is not particularly limited, but forms an angle of, for example, 45 °. , A lattice fringe 23 is formed. The lattice fringes 23 are formed, for example, corresponding to the width of the inspection object 31.
[0024]
The magnification of the imaging lens 25 is appropriately selected according to the light amount accumulation range of the TDI line CCD 26 described later. Then, in this imaging optical system, the angle formed by the object point, that is, the optical axis before reflection from the lattice fringe 23 of the grid plate 24 to the surface of the inspection object 31 with respect to the surface of the inspection object 31 (hereinafter, illumination angle) ), And the angle formed by the optical axis after reflection from the surface of the inspection object 31 to the image point, that is, the TDI line CCD 26 (hereinafter referred to as an observation angle) is the same. And the angle is set as small as possible. For example, the set angle is desirably 15 ° or less with respect to the surface of the inspection object 31. As described above, when the illumination angle and the observation angle are small, the influence of the depth of focus on the vertical movement of the inspection target 31 decreases.
[0025]
As shown in FIG. 2, the TDI line CCD 26 has a configuration in which a plurality of line sensors are arranged in parallel, and in synchronization with the moving speed of the object, here, the inspection object 31, the detected charge is transferred to the next line. It has a function to transfer to the line. That is, referring to FIG. 2, when a portion to be inspected on the surface of the inspection object 31 moves from A to B and C as indicated by an arrow B in FIG. It moves from A to B and C as indicated by arrow C on the TDI line CCD 26. In accordance with the moving speed of the image on the TDI line CCD 26, the electric charge accumulated in each line sensor moves to the rear line as shown by the arrow d. As a result, the integrated charge is output from the last line sensor as indicated by the arrow E, so that the detection sensitivity is improved.
[0026]
Here, as shown in FIG. 3, in the stripe pattern 27 when the lattice fringes 23 are projected on the surface of the inspection target 31, a bright pattern repeated in the moving direction of the inspection target 31 (the direction of arrow A in FIG. 3). Assuming that the minimum unit of repetition of the section 28 and the dark section 29 is one cycle (corresponding to λ in FIG. 3), the range in which the TDI line CCD 26 accumulates the amount of reflected light is n times that cycle (n is a natural number) Is a range in which a phase component α of less than one cycle is added to. Preferably, n is 1 and α is approximately 周期 period. In the example shown in FIG. 3, n is 1 and α is a half cycle. In this case, the relationship between the number of storage stages of the TDI line CCD 26 and the resolution and phase is expressed by the following equation (1).
[0027]
[Number of accumulation stages] × [Resolution]> 1 + α (period) (1)
[0028]
By making the light quantity accumulation range of the TDI line CCD 26 a range to which a phase component α of less than one cycle is added, a signal in which an AC component is superimposed on a DC component is output from the TDI line CCD 26 as shown in FIG. You. This is because, in the striped pattern 27 shown in FIG. 3, the pixels of the TDI line CCD 26 that accumulate the light amount along the DD ′ cross the dark area 29 more than the light area 28, Since the pixel that accumulates the light amount along −E ′ crosses the light portion 28 more than the dark portion 29, as shown in FIG. This is because the sum of the light amounts accumulated along EE ′ is higher.
[0029]
As described above, the sum of the accumulated light amounts varies periodically depending on the position crossing the striped pattern 27 shown in FIG. 3, so that a signal having periodicity as shown in FIG. 4 is output from the TDI line CCD 26. . When a signal having periodicity is obtained, information relating to the phase can be obtained. Since this phase changes depending on the height and surface angle of the inspection target surface, the shape of a minute defect on the inspection target surface can be observed by analyzing the phase information.
[0030]
On the other hand, when the light quantity accumulation range of the TDI line CCD 26 is set to one cycle, the TDI line CCD 26 outputs a signal of only a DC component as shown in FIG. This is because, as shown in FIG. 6, in the stripe pattern 27 when the lattice fringe 23 is projected on the surface of the inspection object 31, even if the pixel of the TDI line CCD 26 that accumulates the amount of light along FF ′, G This is because the sum of the accumulated light amounts becomes the same as shown in FIG. 8, although the pixels that accumulate the light amount along −G ′ have different phases of the start of the light amount accumulation.
[0031]
In this case, if there is no defect on the inspection target surface, even if the inspection target object 31 is moved, a uniform amount of reflected light can be obtained over the entire inspection target surface. However, if there is a defect on the surface to be inspected, the amount of reflected light becomes weaker at the location of the defect. Therefore, the defect can be detected by observing a point where the signal intensity has decreased. However, since a signal of an AC component cannot be obtained from the TDI line CCD 26, information relating to the phase cannot be obtained, so that it is not possible to observe the shape of a minute defect on the inspection target surface.
[0032]
For the reasons described above, in the first embodiment, the light quantity accumulation range of the TDI line CCD 26 is set to 1 + α cycle. An example of detecting a defect in this case will be described. FIG. 9 shows a case where the surface of the inspection object 31 has a portion that is not reflected by the convex defects 32 and 33. In this case, as shown in FIG. 10, the signal intensity becomes lower than the threshold value at the positions corresponding to the convex defects 32 and 33, so that the defect can be detected. Here, the threshold value is set to an appropriate value in advance.
[0033]
FIG. 11 shows a case where the waviness 34 is present on the surface of the inspection object 31. In this case, as shown in FIG. 12, the obtained signal intensity pattern (shown by a solid line) is changed to a standard signal intensity pattern (shown by a dotted line) which is expected to be obtained in a normal case without undulation. Compared to (), the phase is shifted at the position corresponding to the undulation 34, so that a defect can be detected. Further, by obtaining the phase difference signal, it is possible to know the shape such as the height and size of the defect as described below. Here, the standard signal strength pattern is set in advance.
[0034]
Next, the flow of image processing will be described with reference to FIG. The image detected by the TDI line CCD 26 has a stripe pattern as shown in FIG. If there is a defect on the inspection target surface, the interval between stripes at a location corresponding to the defect changes. If there is no defect on the surface to be inspected, the interval between stripes is constant. Subsequently, as shown in FIG. 13B, the detected stripe pattern (shown by a solid line) is compared with a standard stripe pattern (shown by a dotted line). The standard stripe pattern is known in advance. Then, as shown in FIG. 13C, the phase difference is detected from the difference between the detected fringe pattern and the standard fringe pattern. Based on this phase difference, as shown in FIG. 13D, the actual size of the irregularities on the inspection target surface is obtained.
[0035]
As shown in FIG. 14, the signal processing system for performing the above-described image processing includes an image memory 41, a standard pattern generation circuit 42, a threshold generation circuit 43, a moving stage speed detection circuit 44, a phase comparison circuit 45, and an intensity comparison circuit. 46, first and second shape calculation circuits 47 and 48, and a result output circuit 49. The standard pattern generation circuit 42 stores the standard stripe pattern described above, and supplies the standard stripe pattern to the phase comparison circuit 45. The threshold value generation circuit 43 stores the above-described threshold value of the signal strength, and supplies the threshold value to the strength comparison circuit 46.
[0036]
The analog signal output from the TDI line CCD 26 is stored in the image memory 41 and supplied to the phase comparison circuit 45 and the intensity comparison circuit 46. The phase comparison circuit 45 compares the image supplied from the image memory 41 with the standard stripe pattern supplied from the standard pattern generation circuit 42 to detect a phase difference. The first shape calculation circuit 47 calculates and obtains the shape of the defect based on the phase difference detected by the phase comparison circuit 45.
[0037]
Further, the intensity comparison circuit 46 compares the image supplied from the image memory 41 with the threshold value supplied from the threshold value generation circuit 43. The second shape calculation circuit 48 calculates and obtains the shape of the defect based on the comparison result by the intensity comparison circuit 46. The results obtained by the first and second shape calculation circuits 47 and 48 are output from a result output circuit 49. When calculating the shape of the defect in the first and second shape calculation circuits 47 and 48, the size of the defect is corrected based on the moving speed of the stage detected by the moving stage speed detecting circuit 44.
[0038]
According to the first embodiment described above, a signal in which an AC component is superimposed on a DC component is obtained from the TDI line CCD 26, and if there is a convex defect or a minute concave defect on the surface of the inspection object 31, the TDI line CCD 26 The intensity of the DC component of the signal obtained from each pixel becomes lower than the threshold. If the surface of the inspection object 31 has undulation or a gentle concave defect, the phase of the signal intensity pattern obtained from each pixel of the TDI line CCD 26 is shifted from the phase of the standard pattern. Therefore, unevenness defects and undulations can be detected simultaneously.
[0039]
(Embodiment 2)
The surface defect inspection method according to the second embodiment of the present invention is a method for inspecting the surface of an inspection object 61 whose surface is diffusible. Hereinafter, the description overlapping with the first embodiment will be omitted, and only different points from the first embodiment will be described.
[0040]
The diffusible surface does not become a mirror as in the first embodiment. For this reason, as shown in FIG. 15, in the second embodiment, a lens 51 is inserted between the lattice plate 24 and the inspection object 61, and the lattice 51 of the lattice plate 24 is By directly projecting the light onto the surface, a stripe pattern 62 formed by the lattice stripes 23 is projected on the surface of the inspection object 61. Therefore, in the second embodiment, the focal point of the imaging lens 25 is adjusted so as to match the stripe pattern 62 on the surface of the inspection object 61. Then, the TDI line CCD 26 is configured to observe the striped pattern 62 on the surface of the inspection object 61.
[0041]
Here, the angle at which the lattice fringes 23 are projected onto the surface of the inspection target 61 is not particularly limited, but is preferably, for example, 45 °. The sensor surface of the TDI line CCD 26, the surface of the imaging lens 25, and the inspection target surface are arranged so as to satisfy the Scheimpflug relationship. That is, a straight line including the sensor surface of the TDI line CCD 26 (FIG. 15, two-dot chain line H), a straight line including the surface of the imaging lens 25 (FIG. 15, two-dot chain line J), and including the surface of the inspection object 61. It is arranged so that a straight line (FIG. 15, two-dot chain line K) intersects at one point. By satisfying the Scheimpflug relationship, the surface of the inspection object 61 and the sensor surface of the TDI line CCD 26 match.
[0042]
According to the above-described second embodiment, even when the surface of the inspection object 61 is diffusible, the intensity of the signal obtained from each pixel of the TDI line CCD 26 is set to the threshold value as in the first embodiment. By comparing the phase of the signal intensity pattern obtained from each pixel of the TDI line CCD 26 with the phase of the standard pattern, unevenness defects and undulations on the diffusible surface can be simultaneously detected. .
[0043]
(Embodiment 3)
The third embodiment of the present invention is different from the above-described first embodiment in that, instead of the TDI line CCD, a general line CCD is not particularly limited in number as a line sensor. That is, they were arranged in parallel with each other. Hereinafter, the same components as those of the first embodiment are denoted by the same reference numerals as those of the first embodiment, and the duplicate description will be omitted.
[0044]
FIG. 16 is a perspective view schematically showing a configuration of an inspection apparatus used for performing the surface defect inspection method according to the third embodiment of the present invention, and FIG. 17 is a side view thereof. As shown in FIGS. 16 and 17, the inspection apparatus includes a moving stage (not shown) on which an inspection object 31 is mounted, an illumination light source 22, a lattice plate 24 on which lattice fringes 23 are formed, an imaging lens 25, and especially the number thereof. However, for example, three line CCDs 71, 72, 73 are provided. The three line CCDs 71, 72, and 73 are connected to a signal processing circuit 100 that processes signals obtained from each line CCD.
[0045]
In the third embodiment, as in the first embodiment, the surface of the inspection object 31 is mirror-shaped. Therefore, the lattice fringes projected on the surface of the inspection target 31 by the illumination light source 22 are reflected on the surface of the inspection target 31, pass through the imaging lens 25, and be placed on the line CCDs 71, 72, 72, 72. The image is formed at 73. In this way, even if the moving stage moves to the vertical position of the inspection object 31 during the movement, the influence of the vertical movement is reduced.
[0046]
In FIG. 16, reference numeral 81 denotes an image of a lattice fringe image formed on the line CCDs 71, 72, and 73 (hereinafter, referred to as a lattice fringe image). In FIG. 17, reference numerals 22 ′ and 24 ′ are mirror images of the illumination light source 22 and the lattice plate 24, respectively.
[0047]
As in the first embodiment, the lattice fringe 23 is formed such that the stripe pattern when projected on the surface of the inspection object 31 is oblique to the moving direction of the inspection object 31. As the object 31 moves up and down, loose undulations can be detected. The pitch of the lattice is constant, and is sufficiently smaller than the defect existing on the surface of the inspection object 31. The pitch of the grating is not particularly limited, but is, for example, about １ ／ of the size of a defect to be detected.
[0048]
As the three line CCDs 71, 72, 73, for example, commercially available line sensors for detecting color images can be used. Generally, a line sensor for color image detection has three line sensors arranged in parallel at a predetermined interval (up to 100 μm), and each line sensor has a color filter of three primary colors (or complementary colors) such as red, blue, and green. It has an attached structure. In the third embodiment, since the color of the inspection target is white, the output of each line sensor is regarded as having passed through a certain ND (Neutral Density) filter. Note that a sensor without a color filter may be used.
[0049]
In the third embodiment, the light reflected on the surface of the inspection object 31, that is, the reflection image of the lattice fringe 23 is detected simultaneously by the three line CCDs 71, 72, and 73. Then, in the signal processing circuit 100, a defect existing on the surface of the inspection object 31 is determined based on the lattice fringe images 81 simultaneously detected by the three line CCDs 71, 72, 73. With such a configuration, even if the moving speed of the moving stage fluctuates or the moving stage moves up and down, a defect can be detected without being greatly affected by the fluctuation.
[0050]
Next, the relationship between the detection sensitivity of the concave defect and the optical system will be described. FIG. 18 is a diagram showing a part of the lattice fringes 23 on the lattice plate 24. In FIG. 18, reference numerals 74 and 75 are a dark portion and a bright portion of the lattice fringe 23, respectively. A dashed line MM ′ indicates a plane (horizontal plane) parallel to the surface of the inspection object 31. As shown in FIG. 18, the inclination angle of the lattice fringes 23 with respect to the horizontal plane is defined as ψ, and the lattice spacing in the direction perpendicular to the inclination of the dark part 74 of the lattice fringes 23 (that is, the actual lattice spacing) is P. ₀ Then, the lattice interval p in the direction perpendicular to the horizontal plane MM ′, that is, the direction corresponding to the moving direction of the inspection object 31 (this is defined as an apparent lattice interval) p _v Is represented by the following equation (2).
[0051]
p _v = P ₀ / Cosψ (2)
[0052]
Here, as shown in FIG. 19, it is assumed that the dark portion 74 of the lattice fringe 23 is observed when the flat surface of the inspection object 31, that is, the surface having no defect is observed. The size of the observation range on the surface of the inspection object 31 is D, the angle between the optical axis and the surface of the inspection object 31, that is, the illumination angle is θ, the distance between the lattice 23 and the observation range is 1 and the lattice Let L be the distance from 23 to the imaging lens 25, a be the size of the aperture, and f be the focal length of the imaging lens 25. In the state shown in FIG. 19, the size i of the observation range in the direction orthogonal to the optical axis at the center of the observation range is expressed by the following equation (3). From the proportional relationship, the following equation (4) is established.
[0053]
i = Dsin θ (3)
1 / L = i / a (4)
[0054]
Further, the F value of the imaging lens 25 is represented by the following equation (5), and the following equation (6) is derived from the above equations (3), (4) and (5). From this equation (6), the F value of the imaging lens 25 for the size D of the observation range is determined.
[0055]
F = f / a (5)
F = f × (1 / L) × (1 / i) = fl / (LD sin θ) (6)
[0056]
Also, as shown in FIG. 20, a mortar-shaped concave portion 76 having a depth h is present in the observation range on the surface of the inspection object 31, whereby the lattice fringe image is bent, and the dark part 74 of the lattice fringe 23 was originally observed. It is assumed that the optical path (see FIG. 19) has changed so as to observe the bright part 75 of the lattice fringe 23. At this time, it is apparent that the concave portion 76 can be detected. Since the size (distance) at which the checkerboard image bends is 2 × distance × angle, the apparent grid interval p _v The following equation (7) holds. By transforming this equation (7), p _v By substituting the above equation (2) into the following equation, the following equation (8) is obtained.
[0057]
p _v / 2 = 2l (2h / D) (7)
h = p _v D / 8l = P ₀ D / (8lcosψ) (8)
[0058]
From this, it can be seen that in order to find a defect having a size of about several mm in diameter, the actual lattice spacing should be about 0.5 mm. Therefore, in order to find a defect having a diameter D, the F-number of the imaging lens 25 may be selected so as to satisfy the above equation (6). Then the actual grid spacing is P ₀ Thus, when the distance between the lattice fringe 23 and the observation range is 1, a defect having a depth h can be detected. Here, h is expressed by the following equation (9).
[0059]
h = P ₀ D / (8lcosψ) = P ₀ f / (8FLcos @ sinθ) (9)
[0060]
When the size of the concave portion (defect) 76 becomes smaller than (the size of) the observation range D, in addition to the concave portion 76, a flat surface portion without any defect exists in the observation range D. Therefore, not only the light reflected from the concave portion 76 but also the light reflected from the flat surface portion is included, and the S / N ratio of the defect signal is degraded. Therefore, it is preferable to adjust the aperture of the imaging lens 25 so that the size D of the observation range exactly matches the size of the concave portion (defect) 76. In other words, by adjusting the squeezing of the imaging lens 25, the minimum size of the defect to be detected can be changed.
[0061]
Next, the relationship between the lattice fringe image 81 on the imaging plane and the line CCDs 71, 72, 73 will be described. FIG. 21 is a diagram showing a lattice fringe image 81 and line CCDs 71, 72, and 73. For convenience of description, in FIG. 21, a first line CCD 71, a second line CCD 72, and a third line CCD 73 are referred to in order from the top. The three line CCDs 71, 72, and 73 are parallel to the surface of the inspection object 31, that is, parallel to the horizontal plane MM '(see FIG. 18), and include a first line CCD 71 and a second line CCD 72. And the distance between the second line CCD 72 and the third line CCD 73 are the same.
[0062]
As shown in FIG. 21, the lattice fringe image 81 is formed obliquely on the three line CCDs 71, 72, and 73. Therefore, in the first line CCD 71, at the position advanced by Δp from the second line CCD 72, and at the second line CCD 72, the position advanced by the distance Δp from the third line CCD 73, The dark part 82 or the light part 83 is generated. Then, the pitch of the lattice fringe image 81 in the direction parallel to the line CCDs 71, 72, and 73 is constant p at the portion reflected by the surface having no defect. Here, as shown in FIG. 21, the pitch of the lattice fringe image 81 in the direction parallel to the line CCDs 71, 72, and 73 changes to p 'at the portion reflected by the defective surface. In the example shown in FIG. 21, the second dark portion 82 from the left is deformed due to a defect, and the deformed portion is just observed by the second line CCD 72.
[0063]
FIG. 22 is a diagram showing waveforms of sensor signals when the lattice fringe image 81 shown in FIG. 21 is observed by three line CCDs 71, 72, and 73. As shown in FIG. 22, the sensor signals S1, S2, S3 obtained from the three line CCDs 71, 72, 73 are shifted by ΔP pixels corresponding to the distance Δp. Although the pitch of the cycle of each sensor signal is P pixels, the pitch of the cycle of the sensor signal changes to P ′ in the portion reflected by the defective surface.
[0064]
FIG. 23 is a diagram showing waveforms in which the phase difference between the sensor signals obtained from the three line CCDs 71, 72, 73 has been corrected. The moving distance ΔP on the line CCD is determined by the configuration of the optical system. Therefore, as shown in FIG. 23, the sensor signal S2 from the second line CCD 72 is moved to the left by ΔP pixels. Further, the sensor signal S1 from the first line CCD 71 is shifted to the left by 2ΔP pixels. In this way, the three sensor signals S1, S2, and S3 match in the normal portion, but the sensor signal S2 from the second line CCD 72 including the defect indicates that the light reflected on the defective surface is detected. A shift of ΔP ′ pixels occurs. By detecting a signal difference due to this shift, it is possible to detect a defect existing on the surface of the inspection object 31.
[0065]
Next, an example of a specific method for detecting a shift in the cycle of the sensor signal obtained from the line CCD will be described. For example, in the case of a sensor signal as shown in FIG. 11, first, a position where the maximum point (hereinafter, referred to as a peak) and a position where the minimum point (hereinafter, referred to as a bottom) are obtained. Although there is no particular limitation on how to find it, for example, a peak threshold and a bottom threshold are determined as shown in FIG. 24, and in the case of a peak position, the maximum value among signals larger than the peak threshold is determined. On the other hand, in the case of the bottom position, the position having the minimum value among the signals smaller than the bottom threshold value may be obtained. Further, the maximum value and the minimum value obtained as described above are set as a temporary maximum value and a temporary minimum value, respectively, and the signal intensities of pixels around (for example, both sides of) the pixel at which the temporary maximum value is obtained are calculated. The peak position may be obtained by interpolation by obtaining the center of gravity of 3 to 5 pixels, and similarly, the bottom position may be obtained by interpolation based on the temporary minimum value. This increases the accuracy of the peak position and the bottom position.
[0066]
When the peak position and the bottom position are determined, as shown in FIG. 25, each peak is assigned a number counted from the start position. Each bottom is also given a number counted from the start position. For convenience, in the example shown in FIG. 25, the peak numbers are [1], [2],..., And the bottom numbers are <1>, <2>,. As a result, data indicating the position of the peak or bottom counted from the start position is obtained.
[0067]
Then, the numbers of the peaks and the numbers of the bottoms are made to correspond to the positions obtained earlier. For example, as shown in FIG. 26, when the peak and bottom numbers are plotted on the horizontal axis, and the peak and bottom positions are plotted on the vertical axis, and the positions corresponding to the numbers are plotted, when there is no defect on the surface of the inspection object 31, The plot corresponds to a predetermined straight line. However, when a defect exists on the surface of the inspection object 31, a difference ΔP ′ is generated between the plot position and the straight line corresponding to the defect position. Then, the difference ΔP ′ is compared with a preset threshold value, and if the difference ΔP ′ is larger than the threshold value, it is determined that there is a defect, and a defect signal is generated. If the difference ΔP ′ is smaller than the threshold value, it is determined that there is no defect because the size of the defect is within the allowable range.
[0068]
Although a graph is shown in FIG. 26 for convenience of description, the same is actually performed by calculation instead of creating a graph. At this time, the expression of the straight line of the peak position and the expression of the straight line of the bottom position in the normal case can be obtained by calculation, or can be obtained by averaging based on the actual sensor signal of the normal part. When it is determined from an actual signal, it is possible to correct a fixed distortion including an optical system such as a distortion of the imaging lens 25 and a variation of the lattice fringe 23. As shown in FIG. 21, when a plurality of line CCDs 71, 72, and 73 are used, a plurality of differences ΔP ′ may occur. In this case, the absolute value of each difference ΔP ′ is calculated. , May be calculated as a difference signal.
[0069]
In the case of detecting the deviation of the period of the sensor signal by the above-described method, even if the defect on the surface of the inspection object 31 is too large and the lattice fringe image 81 disappears halfway, the peak or bottom Since the difference between the number and the detected position becomes significantly large (a difference of n times the grating pitch is obtained), the detection is possible.
[0070]
Next, the advantage of simultaneously detecting the surface of the inspection object 31 by the plurality of line CCDs 71, 72, 73 will be described. When the defect on the surface of the inspection object 31 is large, as shown in FIG. 27, the deformation of the lattice fringe image 81 becomes large, and the deformation of the lattice fringe image 81 over a plurality of line CCDs 72 and 73 may be detected. In such a case, there is an advantage that there is less room for an error in the detected result. In particular, even if the moving speed of the moving stage fluctuates or the moving stage moves up and down, it is possible to detect a defect without being affected by the fluctuation, thereby reducing errors caused by these.
[0071]
Also, as shown in FIG. 28, when the defect on the surface of the inspection object 31 is sufficiently smaller than the lattice fringe, the defect 84 is located at the intersection of the dark part 82 of the lattice fringe image 81 and the second line CCD 72, and 82. In such a case, for example, assuming that there is only the second line CCD 72 in FIG. 28, this defect 84 cannot be detected. However, in the third embodiment, since the first and third line CCDs 71 and 73 are provided in addition to the second line CCD 72, when the defect 84 moves as shown by an arrow in FIG. This defect 84 can be detected by the first line CCD 71 and the third line CCD 73.
[0072]
In this case, it is necessary to shift the phases of the sensor signals S1, S2, S3 obtained from the first to third line CCDs 71, 72, 73 by 120 °. In general, if the number of line CCDs is n, the phase of the sensor signal obtained from each line CCD is shifted by 360 / n degrees. Therefore, as shown in FIG. 29, when two line CCDs 71 and 72 are provided, the inclination of the grid pattern and the distance between the line CCDs 71 and 72 are adjusted to adjust the first and second line CCDs 71 and 72. It is necessary to make the phases of the sensor signals S1 and S2 obtained from the phase shift by 180 °. In this way, when one line CCD corresponds to the dark portion 82 of the checkerboard image 81, the other line CCD corresponds to the brighter portion 83 of the checkerboard image 81, so that the undetectable area can be reduced. Can be. Here, for the sake of convenience, the description has been made such that the defect 84 moves with respect to the lattice fringe image 81, but the defect 84 actually moves together with the inspection object 31.
[0073]
Next, the signal processing circuit 100 that performs the above-described signal processing will be described with an example in which the number of line CCDs is three. FIG. 30 is a block diagram illustrating an example of a configuration of the signal processing circuit 100. As shown in FIG. 30, the signal processing circuit 100 includes three peak / bottom detection circuits 101, 102, 103 corresponding to the first to third line CCDs 71, 72, 73, a peak / bottom position storage circuit 104, A peak / bottom reference position calculation circuit 105, a position difference calculation circuit 106, a difference determination circuit 107, a defect signal generation circuit 108, and a defect map creation circuit 109 are provided.
[0074]
The peak / bottom detection circuits 101, 102, and 103 respectively record the sensor signals S1, S2, and S3, which are analog signals supplied from the corresponding line CCDs 71, 72, and 73, and respectively detect the sensor signals S1, S2, and S3. Are detected at the peak position and the bottom position. The peak / bottom position storage circuit 104 stores the peak position and the bottom position detected by the peak / bottom detection circuits 101, 102, and 103. The peak / bottom reference position calculation circuit 105 generates normal peak positions and bottom positions serving as references. Here, the peak / bottom reference position calculation circuit 105 calculates the reference normal peak position and bottom position by calculation based on the data of the peak position and bottom position obtained from the actual sensor signal of the normal portion. I have.
[0075]
The position difference calculation circuit 106 compares the peak position and the bottom position stored in the peak / bottom position storage circuit 104 with the reference peak position and the bottom position obtained by the peak / bottom reference position calculation circuit 105, Find the difference. The difference determination circuit 107 compares the difference obtained by the position difference calculation circuit 106 with a preset threshold value. The defect signal generation circuit 108 generates a defect signal based on the comparison result in the difference determination circuit 107. The defect map creation circuit 109 maps a defect position based on the defect signal.
[0076]
According to the third embodiment described above, the plurality of line CCDs 71, 72, and 73 are arranged close to and parallel to each other such that the period of the sensor signal is shifted by a phase difference of 360 / n degrees with respect to the number n of the line CCDs. And a plurality of line CCDs 71, 72, 73 simultaneously detect surface defects of the inspection object 31, thereby detecting surface defects without being affected by the moving speed or vertical movement of the inspection object 31. can do.
[0077]
In the above, the present invention is not limited to the above-described first to third embodiments, and can be variously modified.
[0078]
(Supplementary Note 1) While moving the inspection target surface in a direction orthogonal to the accumulation type line CCD, projecting and reflecting lattice fringes on the inspection object surface and reflecting the image, and observing the reflected image by the accumulation type line CCD, A surface defect inspection method for inspecting a defect on a surface to be inspected,
The lattice fringes are projected on the inspection target surface so that the stripe pattern of the reflected image is oblique to the moving direction of the inspection target surface, and the reflection image is repeated in the moving direction of the inspection target surface. When the minimum unit of repetition of the bright part and the dark part is one cycle, the light amount of the reflected image in a range obtained by adding a phase component α of less than one cycle to n times (n is a natural number) the one cycle is calculated as the accumulation type line. Accumulating by CCD;
Detecting a defect on the inspection target surface based on a signal intensity and a signal intensity pattern corresponding to the amount of light accumulated in each pixel of the accumulation type line CCD;
A surface defect inspection method comprising:
[0079]
(Supplementary Note 2) In the case where the inspection target surface is a glass powder application surface obtained by applying a glass powder paste on a flat substrate surface in a planar manner, the projection source of the lattice fringe and the accumulation type line CCD are respectively set as object points. And an imaging optical system as an image point, and an optical axis before reflection from the projection source to the inspection target surface and an optical axis from the inspection target surface after reflection to the accumulation type line CCD, respectively. The surface defect inspection method according to claim 1, wherein the angle formed between the surface defect and the inspection target surface is 15 ° or less.
[0080]
(Supplementary Note 3) By moving the inspection target surface in a direction orthogonal to the storage type line CCD, projecting the lattice fringes on the inspection target surface, and observing the lattice fringes reflected on the inspection target surface by the storage type line CCD. In inspecting the inspection target surface for defects,
The lattice pattern is projected on the inspection target surface so that the stripe pattern of the lattice pattern reflected on the inspection target surface is oblique to the moving direction of the inspection target surface, and the lattice pattern reflected on the inspection target surface is When a minimum unit of repetition of a bright portion and a dark portion repeated in the moving direction of the inspection target surface is one cycle, a range in which a phase component α of less than one cycle is added to n times (n is a natural number) that one cycle. Accumulating the light quantity of the lattice fringes by the accumulation type line CCD;
Detecting a defect on the inspection target surface based on a signal intensity and a signal intensity pattern corresponding to the amount of light accumulated in each pixel of the accumulation type line CCD;
A surface defect inspection method comprising:
[0081]
(Supplementary Note 4) The signal intensity obtained from the accumulation type line CCD is compared with a preset threshold value, and an area of the inspection target surface where the signal intensity less than the threshold value is obtained is detected. The surface defect inspection method according to any one of supplementary notes 1 to 3, characterized in that:
[0082]
(Supplementary Note 5) The phase of the signal intensity pattern obtained from the accumulation type line CCD is compared with the phase of a preset standard pattern, and the phase of the signal intensity pattern on the surface to be inspected is the phase of the standard pattern. The surface defect inspection method according to any one of supplementary notes 1 to 4, wherein a region deviated from the surface defect is detected.
[0083]
(Supplementary note 6) The surface defect inspection method according to supplementary note 5, wherein a size of a defect existing on the inspection target surface is obtained based on a phase difference between the signal intensity pattern and the standard pattern.
[0084]
(Supplementary Note 7) An imaging optical system having an object point and an image point using the lattice fringes reflected on the inspection target surface and the accumulation type line CCD, respectively, is used, and an imaging lens of the imaging optical system, the inspection target surface, and 4. The surface defect inspection method according to claim 3, wherein the sensor surface of the storage type line CCD satisfies the Scheimpflug relationship.
[0085]
(Supplementary Note 8) The surface defect inspection method according to any one of Supplementary Notes 1 to 7, wherein n is 1 and α is １ ／.
[0086]
(Supplementary Note 9) A surface defect inspection apparatus for inspecting a defect on the inspection target surface by projecting and reflecting a lattice fringe on the inspection target surface while moving the inspection target surface, and observing a reflected image thereof. ,
Projecting means for projecting the lattice fringes such that the fringe pattern of the reflection image is oblique to the moving direction of the inspection target surface;
Accumulation means for accumulating the amount of light of the reflected image for each pixel using one or a plurality of accumulation type line CCDs;
Detecting means for detecting a defect on the inspection target surface based on a signal intensity and a pattern of the signal intensity corresponding to the amount of light accumulated in each pixel of the accumulation means;
A surface defect inspection device comprising:
[0087]
(Supplementary Note 10) While moving the inspection target surface in a direction orthogonal to one or a plurality of line sensors arranged in parallel with each other, a grid pattern is projected and reflected on the inspection target surface, and the reflected image is subjected to pluralities. A surface defect inspection method for inspecting a defect on the inspection target surface by observing with a line sensor group including the line sensor,
The lattice fringes are projected on the inspection target surface so that the stripe pattern of the reflection image is oblique to the moving direction of the inspection target surface, and any one of at least one line sensor in the line sensor group is used. The remaining lines that detect light in the dark part of the reflected image at the light detection point and pass through the light detection point that is detecting the light in the dark part, and are located on a straight line perpendicular to the line sensor group Among the light detection points of the sensor, a step of detecting light of a bright portion of the reflected image at at least one light detection point,
Based on a signal intensity and a signal intensity pattern corresponding to the light detected by the line sensor group, a step of detecting a defect on the inspection target surface,
A surface defect inspection method comprising:
[0088]
(Supplementary note 11) The surface defect inspection according to supplementary note 10, wherein the presence or absence of a defect on the inspection target surface is determined based on a signal corresponding to light detected simultaneously by each line sensor of the line sensor group. Method.
[0089]
(Supplementary Note 12) The surface defect inspection method according to Supplementary Note 10 or 11, wherein an angle formed between the inspection target surface and an optical axis of light reflected by the inspection target surface is 15 ° or less.
[0090]
(Supplementary Note 13) For each of the line sensors, a maximum point and a minimum point of a signal corresponding to the light detected by each line sensor are obtained, and the maximum point and the minimum point are deviated from the original appearance position based on the deviation amount. 13. The surface defect inspection method according to any one of supplementary notes 10 to 12, wherein the presence or absence of a defect on the inspection target surface is determined.
[0091]
(Supplementary Note 14) The relationship between the actual appearance position and the number assigned to the actual appearance position of each of the maximum point and the minimum point and the original appearance position of each of the maximum point and the minimum point 14. The surface defect inspection method according to supplementary note 13, wherein the presence or absence of a defect on the inspection target surface is determined by comparing the relationship between the assigned number and the original appearance position.
[0092]
(Supplementary Note 15) Assuming that the number of the line sensors is n (n is 2 or more), a periodic signal obtained from each line sensor due to repetition of a bright portion and a dark portion of the reflection image is 360 / n degrees each other. 15. The surface defect inspection method according to any one of supplementary notes 10 to 14, wherein the surface defect is shifted by a phase difference.
[0093]
(Supplementary note 16) The surface defect inspection method according to supplementary note 15, wherein n is 3.
[0094]
(Supplementary Note 17) A surface defect inspection apparatus that inspects a defect of the inspection target surface by projecting and reflecting a grid pattern on the inspection target surface while moving the inspection target surface, and observing a reflected image thereof. ,
Projecting means for projecting the lattice fringes such that the fringe pattern of the reflection image is oblique to the moving direction of the inspection target surface;
It has one or a plurality of line sensors arranged in parallel to each other, detects light of a dark part of the reflected image at any light detection point of at least one line sensor, and also detects light of the dark part The light of the bright part of the reflected image is detected by at least one of the light detection points of the remaining line sensors that pass through the light detection point and are located on a straight line perpendicular to the line sensor group. Observation means
Based on a signal intensity and a signal intensity pattern corresponding to the light detected by the observation unit, a detection unit that detects a defect on the inspection target surface,
A surface defect inspection device comprising:
[0095]
(Supplementary Note 18) The three line sensors are arranged so that periodic signals obtained from each line sensor due to repetition of a bright portion and a dark portion of the reflection image are shifted from each other by a phase difference of 120 °. 18. The surface defect inspection apparatus according to claim 17, wherein
[0096]
【The invention's effect】
According to the present invention, a signal in which an AC component is superimposed on a DC component is obtained from the TDI line CCD, and if there is a convex defect or a minute concave defect, the DC component of the signal obtained from each pixel of the TDI line CCD is detected. If the intensity is lower than the threshold value, and if there is undulation or a gentle concave defect, the phase of the signal intensity pattern obtained from each pixel of the TDI line CCD and the phase of the standard pattern are shifted, so that unevenness defects and Swell can be detected at the same time.
[Brief description of the drawings]
FIG. 1 is a schematic diagram showing a configuration of an inspection apparatus used for performing a surface defect inspection method according to a first embodiment of the present invention.
FIG. 2 is a diagram schematically showing a relationship between a storage type line CCD and an inspection object of the inspection apparatus having the configuration shown in FIG.
FIG. 3 is a diagram showing a part of a fringe pattern when a lattice fringe is projected on a surface to be inspected in the surface defect inspection method according to the first embodiment of the present invention;
FIG. 4 is a diagram showing output signals when the light quantity accumulation range of the accumulation type line CCD is set to 1 + α cycle.
FIG. 5 is a diagram for explaining that the amount of accumulated light differs for each pixel when the accumulated amount of light of the accumulation type line CCD is 1 + α cycle.
FIG. 6 is a diagram for explaining that only a DC component output signal can be obtained when the light amount accumulation range of the accumulation type line CCD is one cycle.
FIG. 7 is a diagram showing output signals when the light quantity accumulation range of the accumulation type line CCD is set to one cycle.
FIG. 8 is a diagram for explaining that the accumulated light amounts of the respective pixels are equal when the light amount accumulation range of the accumulation type line CCD is one cycle.
FIG. 9 is a diagram showing a state where a convex defect is present on the inspection target surface.
FIG. 10 is a diagram for explaining a change in signal intensity when there is a convex defect in the surface defect inspection method according to the first embodiment of the present invention;
FIG. 11 is a diagram showing a state where there is undulation on the inspection target surface.
FIG. 12 is a diagram for explaining a change in signal intensity when there is undulation in the surface defect inspection method according to the first embodiment of the present invention;
FIG. 13 is a diagram for explaining a flow of image processing of the surface defect inspection method according to the first embodiment of the present invention;
FIG. 14 is a block diagram illustrating a configuration of a signal processing system of the inspection apparatus used for performing the surface defect inspection method according to the first embodiment of the present invention;
FIG. 15 is a schematic diagram showing a configuration of an inspection apparatus used for performing a surface defect inspection method according to the second embodiment of the present invention.
FIG. 16 is a schematic diagram showing a configuration of an inspection apparatus used for performing a surface defect inspection method according to a third embodiment of the present invention.
17 is a side view of a main part of the inspection device shown in FIG.
FIG. 18 is a diagram showing a part of a lattice fringe in an enlarged manner in order to explain a relationship between a detection sensitivity of a concave defect and an optical system in the surface defect inspection method according to the third embodiment of the present invention.
FIG. 19 is a schematic diagram showing an optical system when there is no defect in the surface defect inspection method according to the third embodiment of the present invention.
FIG. 20 is a schematic diagram showing an optical system when there is a concave defect in the surface defect inspection method according to the third embodiment of the present invention.
FIG. 21 is a diagram showing a relationship between a lattice fringe image and a line sensor in the surface defect inspection method according to the third embodiment of the present invention.
FIG. 22 is a diagram showing a waveform of a sensor signal when the lattice fringe image shown in FIG. 21 is observed.
FIG. 23 is a diagram showing a waveform in which the phase difference between the sensor signals shown in FIG. 22 has been corrected.
FIG. 24 is a diagram showing waveforms of sensor signals for describing a method of detecting a shift in the period of a sensor signal in the surface defect inspection method according to the third embodiment of the present invention;
FIG. 25 is a diagram showing a state where numbers are assigned to respective peaks and bottoms of the sensor signal shown in FIG. 24;
FIG. 26 is a graph showing the relationship between the numbers assigned to the peaks and bottoms of the sensor signal and the positions of the peaks and bottoms.
FIG. 27 is a diagram illustrating a relationship between a lattice fringe image and a line sensor in the surface defect inspection method according to the third embodiment of the present invention;
FIG. 28 is a diagram illustrating a relationship between a lattice fringe image and a line sensor in the surface defect inspection method according to the third embodiment of the present invention;
FIG. 29 is a diagram showing another example of the relationship between the lattice fringe image and the line sensor in the surface defect inspection method according to the third embodiment of the present invention.
FIG. 30 is a block diagram showing a configuration of a signal processing circuit of an inspection device used for performing a surface defect inspection method according to a third embodiment of the present invention.
FIG. 31 is a view for explaining a conventional surface inspection method.
FIG. 32 is a view for explaining the relationship between lattice fringes and defects in a conventional moire interferometry or the like.
FIG. 33 is a diagram for explaining a relationship between a signal pattern of a line CCD obtained by a conventional moire interference method or the like and a defect.
FIG. 34 is a diagram for explaining a relationship between a signal pattern of a line CCD obtained by a conventional moire interference method or the like and a defect.
[Explanation of symbols]
23 Plaid
26 Storage type line CCD
28 Akebe
29 Dark part
31, 61 Inspection object

Claims (5)

While moving the inspection target surface in a direction perpendicular to the storage type line CCD, the inspection target surface is projected and reflected by a lattice fringe, and the reflected image is observed by the storage type line CCD. A surface defect inspection method for inspecting defects,
The lattice fringes are projected on the inspection target surface so that the stripe pattern of the reflected image is oblique to the moving direction of the inspection target surface, and the reflection image is repeated in the moving direction of the inspection target surface. When the minimum unit of repetition of the bright part and the dark part is one cycle, the light amount of the reflected image in a range obtained by adding a phase component α of less than one cycle to n times (n is a natural number) the one cycle is calculated as the accumulation type line. Accumulating by CCD;
Detecting a defect on the inspection target surface based on a signal intensity and a signal intensity pattern corresponding to the amount of light accumulated in each pixel of the accumulation type line CCD;
A surface defect inspection method comprising: When the inspection target surface is a glass powder application surface obtained by applying a glass powder paste on a flat substrate surface in a planar manner, the projection source of the lattice fringe and the accumulation type line CCD are respectively set as an object point and an image point. And an optical axis before reflection from the projection source to the inspection target surface, and an optical axis from the inspection target surface to the accumulation type line CCD after reflection, and the inspection. The surface defect inspection method according to claim 1, wherein an angle between the target surface and the target surface is set to 15 ° or less. By moving the inspection target surface in a direction orthogonal to the storage type line CCD, projecting a grid pattern on the inspection target surface, and observing the grid pattern reflected on the inspection target surface by the storage type line CCD, When inspecting surface defects,
The lattice pattern is projected on the inspection target surface so that the stripe pattern of the lattice pattern reflected on the inspection target surface is oblique to the moving direction of the inspection target surface, and the lattice pattern reflected on the inspection target surface is When a minimum unit of repetition of a bright portion and a dark portion repeated in the moving direction of the inspection target surface is one cycle, a range in which a phase component α of less than one cycle is added to n times (n is a natural number) that one cycle. Accumulating the light quantity of the lattice fringes by the accumulation type line CCD;
Detecting a defect on the inspection target surface based on a signal intensity and a signal intensity pattern corresponding to the amount of light accumulated in each pixel of the accumulation type line CCD;
A surface defect inspection method comprising: Comparing the signal intensity obtained from the accumulation type line CCD with a preset threshold value, and detecting an area of the inspection target surface where the signal intensity equal to or less than the threshold value is obtained. The surface defect inspection method according to claim 1. A surface defect inspection apparatus that inspects a defect of the inspection target surface by projecting and reflecting a grid pattern on the inspection target surface while reflecting the inspection target surface while moving the inspection target surface,
Projecting means for projecting the lattice fringes such that the fringe pattern of the reflection image is oblique to the moving direction of the inspection target surface;
Accumulation means for accumulating the amount of light of the reflected image for each pixel using one or a plurality of accumulation type line CCDs;
Detecting means for detecting a defect on the inspection target surface based on a signal intensity and a pattern of the signal intensity corresponding to the amount of light accumulated in each pixel of the accumulation means;
A surface defect inspection device comprising:

Images