JP4108829B2 - Thickness defect inspection apparatus and inspection method thereof - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a thickness defect inspection apparatus that detects a case where the thickness of an object to be inspected such as a transparent film lacks uniformity, and an inspection method thereof.
[0002]
[Prior art]
Conventionally, the quality of a product is determined by inspecting an inspection object such as a sheet-like film for defects such as foreign matter, holes, and dirt. In this case, while irradiating the inspection object with light from the light source, the inspection object is imaged by an area sensor, a line sensor or the like while the inspection object is sent at a constant speed, and this imaging information is image-processed to obtain a foreign object. Inspect for defects such as holes and dirt.
[0003]
Incidentally, for example, an inspection apparatus disclosed in Japanese Patent Publication No. 7-99355 is known as a technique related to an apparatus for inspecting the quality of an inspection object such as a sheet-like film. This is for inspecting spots such as color filters for liquid crystal displays, and performing arithmetic processing to emphasize the spots from the luminance information of the object to be inspected obtained by the sensor. In response to this, the spots are inspected.
[0004]
[Problems to be solved by the invention]
By the way, when inspecting the quality of the object to be inspected such as a sheet-like film by the inspection method shown in the above-described prior art, the brightness information of the brightness information imaged by the sensor is in the case of a semi-transparent one. It is possible to inspect the thickness of the object to be inspected from the change. In this case, since the light transmittance varies depending on the thickness by applying light to the object to be inspected from the back surface, the quality of the object to be inspected can be inspected according to the change in the transmittance.
[0005]
However, when inspecting the thickness of a transparent object to be inspected, even if light is applied to the object to be inspected from the back, the change in the light transmittance according to the thickness is extremely small, so the image is picked up by a sensor. Therefore, it is impossible to inspect the uniformity of the thickness of the object to be inspected.
[0006]
By the way, it is possible to inspect the thickness of the inspection object by shining light on the transparent inspection object from an oblique direction. There is a problem that the change in the reflectance of light according to the above is affected by the fluttering of the inspection object.
[0007]
In addition, there is a method of visually confirming the reflected light irradiated to the object to be inspected, but this not only requires skill, but also causes oversight, so it is necessary to inspect for defects in the thickness of the object under inspection. Is impossible.
[0008]
The present invention has been made in order to solve such a conventional problem, and a thickness defect inspection apparatus capable of easily and reliably inspecting defects regarding the uniformity of thickness of a transparent object to be inspected and its The purpose is to provide an inspection method.
[0009]
[Means for Solving the Problems]
In order to solve the above-described problem, the thickness defect inspection apparatus according to the present invention has a reference pattern, an imaging unit that images the reference pattern through a transparent inspection object, and a deformation amount of the imaging pattern captured by the imaging unit. And a defect inspection means for performing a defect inspection on the uniformity of the thickness of the object to be inspected.
[0010]
Here, as a transparent to-be-inspected object, it is a flat transparent body which permeate | transmits light like a film, a glass plate, etc., for example. In addition, the distance between the reference pattern and the transparent inspection object is not particularly limited, but the thickness of the inspection object generated in the image information of the inspection object depends on the distance between the reference pattern and the transparent inspection object. For example, when the distance between the sensor and the reference pattern is L, the distance between the reference pattern and the transparent inspection object is 1/3 · L. It is preferable to do.
[0011]
In such a configuration, when the focus position of the imaging means (sensor) is set as a reference pattern, and the reference pattern is imaged through a transparent inspection object, the imaging pattern is distorted according to a change in the thickness of the inspection object. By detecting the degree of deformation (distortion), it is possible to perform a defect inspection on the uniformity of the thickness of the inspection object. The degree of deformation (distortion) can be detected by detecting the deformation or displacement of a line provided on the reference pattern, or the change in the area of a predetermined region on the reference pattern. Also, pattern matching such as correlation can be applied to the pattern.
In addition, when performing a defect inspection of the thickness of a strip-shaped transparent inspection object, the inspection object is efficiently sent from the front end to the rear end of the inspection object sequentially by sending the inspection object in one direction at a predetermined speed. Thickness inspection can be performed.
[0012]
In the thickness defect inspection apparatus according to the present invention, a light source is disposed on the back side of the reference pattern, and irradiation light from the light source is directed to the imaging means (sensor) side through the reference pattern and the transparent inspection object. It is characterized by progress.
Here, the light source may be any light source that emits visible light. For example, a fluorescent lamp or an incandescent lamp can be used.
[0013]
In such a configuration, by arranging the light source on the back side of the reference pattern, the sensor can capture the light that has passed through the reference pattern and the transparent inspection object, and the transparent inspection object has a thickness defect. Inspection can be performed reliably. The reason for this is that reflected light from a transparent inspection object is not captured. For example, when performing a thickness defect inspection while feeding a strip-shaped inspection object in one direction at a predetermined speed, the inspection object is fed. This is because it is not affected by the flutter of time.
[0014]
In the thickness defect inspection apparatus according to the present invention, a light source is disposed between the reference pattern and the transparent inspection object, and reflected light from the reference pattern is passed through the transparent inspection object side to the imaging means side. It is characterized by making it progress.
[0015]
In such a configuration, similarly to the above, since the reflected light of the transparent inspection object is not taken into the imaging means, the transparent inspection object is not affected by the fluttering caused when the inspection object is fed. The defect inspection of the thickness can be reliably performed.
[0016]
In the thickness defect inspection apparatus according to the present invention, the reference pattern is a lattice pattern that is two-dimensionally arranged in a matrix at equal intervals, and the imaging means is an area sensor.
[0017]
Here, it is possible to inspect the change in the thickness of a minute object to be inspected by reducing the size of each lattice pattern of the reference pattern. However, if the size is too small, the time required for pattern matching is reduced. In order to extend the length, for example, when the resolution of the sensor is 0.1 to 0.5 mm, about 5 × 5 mm is preferable.
[0018]
In such a configuration, since the imaging means can be an area sensor, the imaging range can be widened, so the reference pattern capture time through the transparent inspection object can be shortened and the reference pattern can be captured. The change in the thickness of the object to be inspected can be captured by the distortion of the edge portion of the lattice pattern of the captured image pattern.
[0019]
Moreover, in the thickness defect inspection apparatus according to the present invention, the reference pattern is a lattice pattern arranged in a line orthogonal to the scanning direction of the imaging unit, and the imaging unit is a line sensor. .
[0020]
In such a configuration, by using the line sensor as the image pickup unit, the width of the detection area in the image pickup pattern obtained by picking up the reference pattern can be changed according to the visual field of the line sensor, and the distortion of the lattice pattern of the image pickup pattern can be changed. Changes in the thickness of the object to be inspected can be captured.
[0021]
In the thickness defect inspection apparatus according to the present invention, the reference pattern is a figure, and the deformation amount is detected by pattern matching. Here, the figure can be a geometric pattern, a human face, a picture, or the like.
[0022]
In such a configuration, a distortion according to the change in the thickness of the inspection object is generated in the imaging pattern obtained by imaging the reference pattern which is a figure, so that the defect inspection of the thickness of the transparent inspection object is reliably performed as described above. be able to.
[0023]
Further, in the defect inspection means of the thickness defect inspection apparatus according to the present invention, the defect inspection means divides the grid-like imaging patterns arranged at equal intervals imaged by the imaging means into a plurality of detection areas having the same area. In addition, a luminance information integration unit that integrates the luminance information in each of the divided detection areas, a luminance information subtraction unit that sequentially subtracts the integrated luminance information of a plurality of detection areas integrated by the luminance information integration unit, And a defect detection unit that detects a defect with respect to the uniformity of the thickness of the inspection object based on the subtraction result subtracted by the luminance information subtraction unit.
[0024]
Here, the luminance information integration unit can obtain the luminance information of a plurality of detection areas divided into equal areas on the image by integration processing. The area (size) of the detection area is arbitrary, but it is preferable to reduce the detection area in order to increase the accuracy of detecting the defect in uniformity of the thickness of the inspection object. The luminance information subtraction unit can detect the magnitude of the deformation amount by sequentially subtracting the luminance information of each detection area integrated by the luminance information integration unit and performing differential processing. By performing this differential processing, even if the degree of the defect in the thickness of the inspection object is extremely small, the defect can be reliably detected. Further, the defect detection unit that detects a defect regarding the uniformity of the thickness of the inspection object has a threshold value indicating an allowable range of the deformation amount, and compares this threshold value with the subtraction result of the luminance information subtraction unit. . This threshold value can be changed according to an allowable range for determining whether each product is good or bad.
[0025]
In such a configuration, the luminance information integration unit obtains the luminance information for each detection area by integrating from the imaging pattern imaged by the imaging means. Then, when the luminance information subtraction unit performs differentiation processing by subtraction processing of luminance information for each adjacent detection area, the defect detection unit compares the value obtained by the differentiation processing with the threshold value, and the differential processing result When the thickness exceeds a threshold value, a thickness defect (defect regarding uniformity) is judged, and a thickness defect detection (non-uniformity detection) of a transparent inspection object is performed. As a result, automatic defect detection becomes possible, so that not only skill by visual inspection as in the prior art is required, but also oversight of defects can be prevented.
[0026]
Further, the thickness defect inspection method according to the present invention images a reference pattern through a transparent inspection object, and performs defect inspection on the thickness uniformity of the inspection object based on the deformation amount of the imaged pattern. It is characterized by doing.
[0027]
In such an inspection method, the focus position of the imaging means (sensor) is used as a reference pattern, and the reference pattern is imaged through the transparent inspection object, and the imaging pattern is distorted according to the change in the thickness of the transparent inspection object. By detecting the amount of deformation of the imaging pattern, it is possible to perform defect inspection on the thickness uniformity of the inspection object.
[0028]
In the thickness defect inspection method according to the present invention, the reference pattern is a lattice pattern that is two-dimensionally arranged in a matrix at equal intervals, and the imaging pattern captured by the imaging unit is divided into a plurality of detection areas having the same area. In addition, the luminance information in each of the divided detection areas is integrated, and the integrated luminance information of the integrated detection areas is sequentially subtracted, and the thickness uniformity of the inspection object is based on the subtraction result. It is characterized by detecting defects.
[0029]
In such an inspection method, as described above, since automatic defect detection becomes possible, it is necessary to reliably inspect defects in the thickness of the object to be inspected without requiring skill by visual inspection as in the past. Can do.
[0030]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
1 is a view showing an embodiment of a thickness defect inspection apparatus according to the present invention, FIG. 2 is a plan view showing an example of a reference pattern, FIG. 3 is a block diagram showing defect inspection means, and FIG. It is a flowchart which shows the inspection method of the thickness defect by the thickness defect inspection apparatus of 1.
[0031]
The thickness defect inspection apparatus shown in FIG. 1 includes a reference pattern 1 and an area sensor 2, and a transparent film 3 as a transparent inspection object that is sent between the reference pattern 1 and the area sensor 2 at a constant speed. The thickness inspection is performed.
[0032]
The area sensor 2 is focused on the reference pattern 1. A light source 4 is disposed on the back side of the reference pattern 1, and transmitted light from the light source 4 travels to the area sensor 2 side through the reference pattern 1 and the transparent film 3. As a result, the area sensor 2 is prevented from receiving reflected light from the transparent film 3.
[0033]
Here, the interval between the reference pattern 1 and the transparent film 3 is not particularly limited, but is preferably 1/3 · L, where L is the interval between the reference pattern 1 and the area sensor 2. . This is an interval for obtaining an appropriate degree of distortion. When the interval between the reference pattern 1 and the transparent film 3 is smaller than 1/3 · L, the change in the thickness of the transparent film 3 in the imaging pattern 1a described later. This is because the degree of distortion corresponding to the thickness increases, and conversely, if the interval is larger than 1/3 · L, the degree of distortion corresponding to the change in the thickness of the transparent film 3 decreases.
[0034]
As shown in FIG. 2, the reference pattern 1 is a lattice pattern that is two-dimensionally arranged in a matrix at equal intervals. In addition, each dimension in the hatched portion A and the white portion B of each lattice pattern is not particularly limited. However, for example, an imaging pattern 1a described later corresponding to the resolution in the area sensor 2 is set to 5 × 5 mm. Is obtained.
[0035]
Here, the allowable range of the thickness of the transparent film 3 is ± 10% in the present embodiment when the thickness of the transparent film 3 is 1 mm. A portion exceeding this range is detected as a thickness defect, details of which will be described later. Moreover, in the imaging pattern 1a in the same figure, what is shown with the wavy lines a and b shows distortion (deformation) corresponding to the change in the thickness of the transparent film 3. This distortion is caused by the difference in the refractive index of light according to the change in the thickness of the transparent film 3, and is particularly likely to occur at the edge portion of the hatched portion A and the white portion B of the imaging pattern 1a. . A wavy line a indicates horizontal distortion, and a wavy line b indicates vertical distortion.
[0036]
Further, the resolution in the area sensor 2 is not particularly limited. When the hatched portion A and the white portion B of each lattice pattern are defined as one detection area, the resolution is 0.1 to 0.5 mm. Is desirable.
[0037]
As shown in FIG. 3, the defect inspection means for performing the thickness defect inspection based on the imaging pattern 1a imaged by the area sensor 2 includes a luminance information integrating unit 5a, a luminance information subtracting unit 5b, and a defect detecting unit 5c. It is configured.
[0038]
The luminance information integration unit 5a partitions the imaging pattern 1a of FIG. 2 into a plurality of detection areas having the same area, and integrates the luminance information for each of the plurality of detection areas. That is, when (x1, y1 + y2) corresponding to the hatched portion A and the white portion B of the lattice pattern of the reference pattern 1 is defined as one detection area, the luminance information is integrated for each detection area to obtain an integral value Σ11. , Σ21, Σ31, Σ41... Σmn.
[0039]
The luminance information subtracting unit 5b sequentially subtracts the luminance information integrated in each detection area by the luminance information integrating unit 5a, and the subtraction values (Σ11−Σ21), (Σ21−Σ31), (Σ31−Σ41). .. Differential processing for obtaining (Σm-1, n-1−Σmn) is performed. In addition, even if it is a case where the degree of the defect of the thickness of the transparent film 3 is very small, the defect can be reliably detected by integration and differentiation processes.
[0040]
The defect detection unit 5c determines whether the magnitude of the deformation amount is larger than a predetermined value (threshold value) from the subtraction value (differential processing result) obtained by the luminance information subtraction unit 5b, and the thickness of the transparent film 3 The defect is detected with respect to the uniformity (whether or not the thickness is uniform). That is, if the absolute value of the value (corresponding to the thickness change ratio) obtained by the luminance information subtraction unit 5b is equal to or less than the threshold value (± 10% in the change ratio), the thickness of the transparent film 3 is within the allowable range. When the error exceeds ± 10%, it is determined that the thickness of the transparent film 3 is outside the allowable range, that is, NG.
[0041]
Then, the thickness defect detection method of the transparent film 3 by the thickness defect inspection apparatus of the above structures is demonstrated using FIG.
[0042]
First, the reference pattern 1 is imaged by the area sensor 2 through the transparent film 3, and after obtaining the imaging pattern 1a shown in FIG. 2, luminance information of each detection area is obtained by integration (step 401). That is, as described above, when (x1, y1 + y2) corresponding to the hatched portion A and the white portion B of the lattice pattern of the reference pattern 1 is set as one detection area, the respective detection areas are integrated to obtain an integrated value Σ11. , Σ21, Σ31, Σ41... Σmn.
[0043]
Next, for each adjacent detection area, subtraction processing is performed with (Σ11−Σ21), (Σ21−Σ31), (Σ31−Σ41)... (Σm−1, n−1−Σmn), and the differential value Is obtained (step 402). Next, the absolute value of the obtained differential value (subtraction value) is compared with a threshold value to detect a thickness defect of the transparent film 3 (step 403).
[0044]
Here, when the absolute value of the differential value obtained in (Step 402) is smaller than the threshold value, it is determined that the thickness of the transparent film 3 is within an allowable range (for example, the thickness change ratio is ± 10%). When the threshold value is exceeded, it is determined that the thickness of the transparent film 3 is outside the allowable range, that is, NG.
[0045]
FIG. 5 is a diagram showing another embodiment in which the area sensor 2 of FIG. 1 is changed to a line sensor 2A.
[0046]
The reference pattern 1A shown in the figure is a lattice pattern in which hatched portions A and white portions B are alternately arranged vertically. The dimensions of the hatched portion A and the white portion B of each lattice pattern are not particularly limited, but are, for example, 5 × 5 mm as described above.
[0047]
Further, the image pickup pattern 1b picked up by the line sensor 2A through the transparent film 3 is a stripe pattern as shown in FIG. In the imaging pattern 1b in the figure, what is indicated by a wavy line a indicates distortion according to a change in the thickness of the transparent film 3. As described above, this distortion is caused by the difference in the refractive index of light according to the change in the thickness of the transparent film 3, and the wavy line a indicates the distortion in the horizontal direction.
[0048]
Then, based on the imaging pattern 1b captured by the line sensor 2A, the defect inspection means 5 including the luminance information integrating unit 5a, the luminance information subtracting unit 5b, and the defect detecting unit 5c shown in FIG. A defect inspection of the thickness of the film 3 is performed.
[0049]
That is, after obtaining the imaging pattern 1b of the reference pattern 1A through the transparent film 3 by the line sensor 2A, the luminance information of each detection area is obtained by integration. Here, in order to set each detection area in the imaging pattern 1b, the scan width V is set by the scan rate of the line sensor 2A, for example. As described above, (x1, y1 + y2) corresponding to the hatched portion A and the white portion B of the lattice pattern of the reference pattern 1A is set as one detection area, and the respective detection areas are integrated to obtain integrated values Σ11, Σ21. , Σ31, Σ41... Σmn.
[0050]
Next, for each adjacent detection area, subtraction processing is performed with (Σ11−Σ21), (Σ21−Σ31), (Σ31−Σ41)... (Σm−1, n−1−Σmn), and the subtraction is performed. The thickness defect of the transparent film 3 is detected by comparing the value with the threshold value.
[0051]
That is, if the differentiated value is equal to or less than the threshold value, as described above, the thickness of the transparent film 3 is determined to be within the allowable range, and when the value exceeds the threshold value, The thickness of the transparent film 3 is determined to be outside the allowable range, that is, NG. The threshold value is set to a value at which the thickness non-uniformity becomes ± 10%, for example.
[0052]
As described above, in each of the above embodiments, the focus positions of the area sensor 2 and the line sensor 2A are the reference patterns 1 and 1A, and the reference patterns 1 and 1A are imaged through the transparent film 3 that is a transparent inspection object. At the same time, since the degree of distortion of the imaging pattern corresponding to the change in the thickness of the transparent film 3 is detected by the deformation of the imaging pattern, a defect inspection of the thickness of the transparent film 3 can be performed easily and reliably.
[0053]
The light source 4 is arranged at the back side of the reference patterns 1 and 1A or between the reference patterns 1 and 1A and the transparent film 3, and the area sensor 2 and the line sensor 2A receive reflected light from the transparent film 3. Since it was made not to be inserted, it is not influenced by the fluttering at the time of feeding the transparent film 3, and the defect inspection of the thickness of the transparent film 3 can be reliably performed.
[0054]
In the thickness defect inspection apparatus shown in FIG. 1, the case where the light source 4 is arranged on the back side of the reference pattern 1 has been described. However, the present invention is not limited to this example. For example, as shown in FIG. You may arrange | position in the position which irradiates the surface of the reference pattern 1 on the back side.
[0055]
Also in this case, similarly to the above, since the reflected light from the transparent film 3 is not taken into the area sensor 2 and the line sensor 2A, it is not affected by the fluttering during the feeding of the transparent film 3, and the transparent film 3 Thickness defect inspection can be performed reliably.
[0056]
Further, as apparent from the description of each of the above embodiments, when the thickness of the transparent film 3 changes, the imaging patterns 1a and 1b described in FIGS. 2 and 5 are distorted. Not only the lattice pattern described above but also a geometric pattern reference pattern 1B as shown in FIG. 7 can be used, for example. Further, the thickness is not limited to such a geometric pattern reference pattern 1B, but by using other reference patterns based on figures such as human faces and pictures, and pattern matching between the pattern and the image pattern by correlation or the like. It is also possible to perform an inspection for the uniformity.
[0057]
【The invention's effect】
As described above, according to the thickness defect inspection apparatus of the present invention, it is possible to easily perform defect inspection on the uniformity of the thickness of the inspection object.
[Brief description of the drawings]
FIG. 1 is a diagram showing an embodiment of a thickness defect inspection apparatus of the present invention.
FIG. 2 is a plan view showing an example of a reference pattern used in the thickness defect inspection apparatus of FIG.
FIG. 3 is a block diagram showing defect inspection means provided in the thickness defect inspection apparatus of FIG. 1;
4 is a flowchart showing a thickness defect inspection method by the thickness defect inspection apparatus of FIG. 1; FIG.
FIG. 5 is a diagram showing another embodiment in which the area sensor of FIG. 1 is changed to a line sensor.
6 is a diagram showing another embodiment in the case where the arrangement position of the light source is changed in the thickness defect inspection apparatus of FIG. 1. FIG.
7 is a diagram showing another example of the reference pattern shown in FIGS. 2 and 5. FIG.
[Explanation of symbols]
1, 1A, 1B Reference pattern 1a, 1b Imaging pattern 2 Area sensor 2A Line sensor 3 Transparent film 5 Defect inspection means 5a Luminance information integration unit 5b Luminance information subtraction unit 5c Defect detection unit

Claims (6)

A grid reference pattern arranged in a two-dimensional matrix , and
Imaging means for imaging the reference pattern through a transparent object;
A defect inspection means for performing a defect inspection on the uniformity of the thickness of the object to be inspected based on the deformation amount of the imaging pattern imaged by the imaging means ,
The defect inspection means divides the imaging pattern into a plurality of detection areas having the same area, and a luminance information integration unit that integrates luminance information in each of the divided detection areas;
A luminance information subtraction unit that sequentially subtracts the integrated luminance information of the plurality of detection areas integrated by the luminance information integration unit;
Based on the subtraction result subtracted by the luminance information subtraction unit, a defect detection unit that detects a defect about the uniformity of the thickness of the inspection object;
A thickness defect inspection apparatus comprising: The thickness defect inspection apparatus according to claim 1 , wherein a light source is disposed on a back side of the reference pattern, and the light source advances light to the imaging unit side through the reference pattern and the transparent inspection object. A light source is disposed between the reference pattern and the transparent inspection object, and the light source irradiates the reference pattern with light and reflects the reflected light from the reference pattern through the transparent inspection object. The thickness defect inspection apparatus according to claim 1, wherein the thickness defect inspection apparatus is advanced to the imaging means side.   The thickness defect inspection apparatus according to claim 1, wherein the imaging unit is an area sensor. The imaging means is a line sensor that scans perpendicularly or in parallel to a lattice of lattice patterns arranged in a two-dimensional matrix in the reference pattern, and performs imaging for each column of the lattice patterns orthogonal to the scanning direction. The thickness defect inspection apparatus according to claim 1, wherein: A grid pattern reference pattern that is two-dimensionally arranged at regular intervals is imaged through a transparent inspection object, and the captured image pattern is divided into a plurality of detection areas having the same area, and each of the divided detections is detected. The luminance information in the area is integrated, the integrated luminance information of a plurality of integrated detection areas is sequentially subtracted, and a defect in the thickness uniformity of the inspection object is detected based on the subtraction result. Thickness defect inspection method.

Images