Device for inspecting glass substrates
Through digital image processing technology, the defects of the glass substrate are detected using two light sources and camera systems, which solves the problem of inaccurate detection in the prior art, improves detection accuracy and process efficiency, and reduces manufacturing costs.
Patent Information
- Application Number
- CN201911098938.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2018-11-12
- Filing Date
- 2019-11-12
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2039-11-12
AI Technical Summary
The prior art is difficult to accurately detect tiny defects of glass substrates, which leads to easily damage to the glass substrates or produce poor products during the manufacturing process, increasing manufacturing costs.
Using digital image processing technology, light scattered and reflected light of the glass substrate is captured through two light sources at different angles and camera systems, and combined with weight calculation and image correction, defects of the glass substrate are detected.
More accurate and efficient detection of glass substrate defects is achieved, manufacturing cost loss caused by wrong judgments is reduced, and process input rate and product quality are improved.
Smart Images

Figure CN111175227B_ABST
Abstract
Description
[0001] CROSS-REFERENCE TO RELATED APPLICATIONS
[0002] This application claims priority to and the benefit of Korean Patent Application No. 10-2018-0138320, filed on November 12, 2018, which is hereby incorporated by reference in its entirety. Technical Field
[0003] Exemplary embodiments of the present disclosure relate generally to an apparatus and method for inspecting a glass substrate, and more particularly, to an apparatus and method for inspecting a glass substrate that can detect defects of the glass substrate using a digital image of the glass substrate. Background Art
[0004] Generally, a glass substrate used to manufacture a flat panel display device such as a liquid crystal display (LCD), an organic light emitting display (OLED), etc. is very thin and may be easily damaged by a slight collision or temperature change.
[0005] For example, when the glass substrate has micro defects such as micro voids and micro cracks, the glass substrate can be easily damaged when processing or manufacturing a display device. In addition, when impurities penetrate into the glass substrate through the micro defects, defective products or quality degradation may occur due to contamination.
[0006] The above information disclosed in this section is only for understanding the background of the present inventive concept and therefore it may contain information that does not constitute prior art. Summary of the invention
[0007] One or more exemplary embodiments provide an apparatus and method for inspecting a glass substrate, which may detect defects of the glass substrate through a digital image processing technique.
[0008] One or more exemplary embodiments also provide an apparatus and method for inspecting a glass substrate that can detect defects of the glass substrate more accurately and easily than related art methods.
[0009] One or more exemplary embodiments also provide an apparatus and method for inspecting a glass substrate that can effectively prevent a loss of manufacturing cost due to an erroneous determination of a defect.
[0010] According to some exemplary embodiments, an apparatus for inspecting a glass substrate may include a platform, a first light source, a first camera, a second light source, a second camera, and a defect detection unit, wherein: the platform is configured to support the glass substrate; the first light source is used to irradiate light onto the surface of the glass substrate at a first angle; the first camera is used to capture scattered light of the light irradiated from the first light source; the second light source is used to irradiate light onto the surface of the glass substrate at a second angle greater than the first angle; the second camera is used to capture reflected light and scattered light of the light irradiated from the second light source; and the defect detection unit is used to detect defects of the glass substrate using a first image provided by the first camera and a second image provided by the second camera.
[0011] The stage may include a driving unit for moving the glass substrate in one direction.
[0012] The first light source and the first camera may be positioned ahead of the second light source and the second camera in a direction.
[0013] The first camera and the second camera may be arranged in a direction perpendicular to the platform. Each of the first camera and the second camera may be a line scan camera. Each of the first camera and the second camera may include a CMOS or a CCD.
[0014] The first light source and the second light source may be configured to irradiate light in a line form. Each of the first light source and the second light source may include a series of LEDs arranged in a bar shape.
[0015] The defect detection unit may include a weight calculation unit, an image generation unit, and a defect determination unit, wherein: the weight calculation unit is used to calculate the standard deviation of pixels in each area in a series of areas in the second image, and use the standard deviation to calculate the weight for each area in the series of areas; the image generation unit is used to apply the weight to the first image to generate a corrected first image; and the defect determination unit is used to determine the defects of the glass substrate using the corrected first image.
[0016] The weight calculation unit may be configured to divide the second image into a series of regions, and use the grayscale value and the average grayscale value of the pixels in each region of the series of regions to calculate the standard deviation. Each standard deviation in the standard deviation may be calculated by the following formula 1.
[0017] Formula 1
[0018]
[0019] Where s is the standard deviation of a region in the series, n is the number of pixels in a region, and X i is the grayscale value of the i-th pixel (i is a natural number from 1 to n) of a region, and is the average gray value of these pixels.
[0020] According to some exemplary embodiments, a method for inspecting a glass substrate may include: providing a glass substrate; irradiating light from a first light source onto a first area of the glass substrate and capturing a first image of the first area of the glass substrate using a first camera; irradiating light from a second light source onto the first area of the glass substrate and capturing a second image of the first area of the glass substrate using a second camera; calculating weights corresponding to pixels of the second image; applying the weights to the pixels of the first image to generate a corrected first image; and detecting defects of the glass substrate using the corrected first image.
[0021] While the glass substrate is being inspected, the glass substrate may be moved in one direction.
[0022] The first camera may capture scattered light of the light irradiated from the first light source.
[0023] The second camera may capture reflected light and scattered light of the light irradiated from the second light source.
[0024] Calculating the weights may include dividing the second image into a plurality of regions, calculating a standard deviation of the pixels using the grayscale values and the average grayscale value of the pixels of each of the plurality of regions, and converting the standard deviation of each pixel into a weight according to a reference value table. Each standard deviation may be calculated by the following formula 2.
[0025] Formula 2
[0026]
[0027] Where s is the standard deviation, n is the number of pixels, and X i is the grayscale value of the i-th pixel (i is a natural number from 1 to n), and is the average gray value of the pixels.
[0028] The corrected first image may be generated by multiplying the grayscale values of the pixels of the first image by the weights.
[0029] Detecting a defect of the glass substrate may include comparing a grayscale value of a pixel of each region of the corrected first image with a threshold value to binarize the grayscale value, connecting pixels having a binarized value of 1 to each other, and determining a defect when a cumulative length of pixels connected to each other is greater than a length reference value.
[0030] Connecting pixels may include connecting only pixels within a predetermined range. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] The accompanying drawings are included to provide a further understanding of the inventive concept and are incorporated in and constitute a part of this specification. The accompanying drawings illustrate exemplary embodiments of the inventive concept and together with the description serve to explain the principles of the inventive concept.
[0032] Figure 1 and Figure 2 is a schematic block diagram of an apparatus for inspecting a glass substrate according to an embodiment of the present disclosure.
[0033] Figure 3 is a flowchart illustrating a method for inspecting a glass substrate according to an embodiment of the present disclosure.
[0034] Figure 4 is a plan view of an example of a glass substrate that may be inspected according to an embodiment of the present disclosure.
[0035] Figure 5 It is shown in Figure 3 During the method Figure 1 and Figure 2 The operation of the defect detection unit is shown in a flowchart of the tasks performed.
[0036] Figure 6 is Figure 3 and Figure 5 A second image acquired during the method of claim 1 and a graphical example showing the calculation of grayscale values of pixels in a region of the second image.
[0037] Figure 7 is a diagram showing an example of a standard deviation of pixels in a region of a second image calculated according to an embodiment of the present disclosure.
[0038] Figure 8 The present invention is based on the embodiment of the present invention. Figure 7 0 is a graph of an example of the standard deviations of the weights calculated for the pixels of the second image.
[0039] Fig. 9 is a graph showing an example of a reference value table for calculating weights of pixels of the second image from standard deviations of the pixels of the second image according to an embodiment of the present invention.
[0040] Fig. 10A It is shown Figure 3 The first image is a photo of an example.
[0041] Fig. 10B It shows that according to Figure 3 Photograph of an example of a corrected first image acquired by the method shown.
[0042] Fig.11 It is shown in Figure 3Flowchart of tasks performed by the operation of a defect determination unit during the method. DETAILED DESCRIPTION
[0043] Hereinafter, the preferred embodiments of the present invention will be described in more detail with reference to the accompanying drawings. The following embodiments are provided to enable those skilled in the art to fully understand the present invention. The embodiments may be modified in various ways. The scope of the present invention is not limited to the embodiments described below.
[0044] In the embodiments, the terms first, second, etc. are not used in a limiting sense, and are used for the purpose of distinguishing one element from another element. In addition, an expression indicating the singular may include an expression indicating the plural unless it is obviously different in the context.
[0045] In addition, when layers, regions, elements, etc. are referred to as being "connected" in the embodiments, it will be understood that it may mean when the layers, regions, or elements are directly connected and when the layers, regions, or elements are indirectly connected to each other. For example, when layers, regions, elements, etc. are electrically connected, it may mean not only when the layers, regions, elements, etc. are directly connected to each other but also when another layer, region, element, etc. is interposed therebetween and is indirectly connected.
[0046] For convenience of explanation, the size of the elements shown in the drawings may be exaggerated or reduced. For example, for convenience of explanation, the size and thickness of each element are arbitrarily shown, and therefore the present invention is not necessarily limited to those shown in the drawings.
[0047] Figure 1 and Figure 2 is a schematic block diagram of an apparatus for inspecting a glass substrate according to an embodiment of the present disclosure.
[0048] refer to Figure 1 and Figure 2 , an apparatus 100 for inspecting a glass substrate may include a stage 10 on which a glass substrate 200 is placed, a first optical system disposed on one side (e.g., the upper side) of the stage 10 and including a first light source 20 and a first camera 30, a second optical system disposed on the same side (e.g., the upper side) of the stage 10 and including a second light source 40 and a second camera 50, and a defect detection unit 60 for detecting defects of the glass substrate 200 using a first image provided by the first camera 30 and a second image provided by the second camera 50. In the illustrated embodiment, the second optical system including the second light source 40 and the second camera 50 is spaced apart from the first optical system including the first light source 20 and the first camera 30 in a direction in which the apparatus 100 is configured to move the glass substrate 200 on the stage 10.
[0049] In one or more embodiments, the platform 10 may be, for example, in the form of a flat plate, so that a glass substrate 200 as an object to be inspected may be placed thereon. In addition, the platform 10 may include a driving unit 10a configured to move the glass substrate 200 in one direction at a predetermined speed while inspecting the glass substrate 200. The glass substrate 200 may be moved to different positions relative to the first optical system and the second optical system by the driving unit 10a.
[0050] Figure 1 shows the position of the glass substrate 200 when the inspection starts, and Figure 2 A state is shown in which the glass substrate 200 is moved in one direction by the driving unit 10 a.
[0051] although Figure 1 and Figure 2 , one glass substrate 200 is shown in FIG. 1 , but the inspection may be performed in a state where a plurality of glass substrates 200 are arranged in a line on the stage 10 .
[0052] The first light source 20 and the second light source 40 may irradiate light in a line form to (e.g., the first light source 20 and the second light source 40 may irradiate collimated or substantially collimated light to) a predetermined area of the glass substrate 200. The first light source 20 and the second light source 40 may include, for example, a plurality of light emitting diodes (LEDs) emitting white light. The light in a line form may be obtained by arranging the plurality of light emitting diodes (LEDs) in, for example, a bar shape.
[0053] In addition, the first light source 20 and the second light source 40 may further include optical elements (such as lenses, filters, etc.) for improving the efficiency and uniformity of light and for adjusting the light paths in parallel.
[0054] The first camera 30 and the second camera 50 may be arranged in a direction perpendicular to the platform 10 (e.g., the first camera 30 and the second camera 50 may be oriented to be orthogonal or substantially orthogonal to the platform 10). For example, the first camera 30 and the second camera 50 may be arranged to face the platform 10 within a range of 0° to 10° relative to the normal of the platform 10.
[0055] In one embodiment, each of the first camera 30 and the second camera 50 may be a line scan camera capable of sequentially photographing a direction perpendicular or substantially perpendicular to the moving direction of the glass substrate 200 on the platform 10 (eg, Figure 1 and Figure 2 The line scan camera may sequentially photograph the plurality of linear regions to obtain an image of the entire region of the glass substrate 200.
[0056] Each of the first camera 30 and the second camera 50 may include, for example, a charge coupled device (CCD) or a complementary metal oxide semiconductor (CMOS) as an imaging device for obtaining a digital image.
[0057] In one embodiment, the first light source 20 may be arranged to irradiate light at a first angle θ1 relative to a surface of the glass substrate 200 (e.g., an upper surface of the glass substrate 200 away from the platform 10), and the first camera 30 may be arranged to capture scattered light (scattered light of the light irradiated from the first light source 20) irregularly reflected on the surface of the glass substrate 200.
[0058] The scattered light may include light reflected from surface defects of the glass substrate 200 or the like.
[0059] The first camera 30 is arranged substantially perpendicular to the surface of the glass substrate 200, and the first light source 20 is arranged such that the optical axis of the first light source 20 is at a first angle θ1 relative to the surface of the glass substrate 200. In addition, in the illustrated embodiment, along the direction in which the apparatus 100 is configured to move the glass substrate 200 on the platform 10, the first light source 20 is positioned before the first camera 30 (i.e., in front of the first camera 30). In this state, when the light irradiated from the first light source 20 is totally reflected on the surface of the glass substrate 200, an image with a dark (black) background can be captured. In addition, when the light irradiated from the first light source 20 is irregularly reflected by a portion of the glass substrate 200 and generates scattered light, a partially bright image on a dark (black) background can be captured.
[0060] The first optical system configured as described above is used in a dark field method, which can facilitate binarization of image data because defective portions are captured as bright and normal portions are captured as dark.
[0061] Furthermore, the second light source 40 is arranged to irradiate light to the surface of the glass substrate 200 at a second angle θ2 greater than the first angle θ1 , and the second camera 50 is arranged to capture reflected light and scattered light of the light irradiated from the second light source 40 .
[0062] The reflected light may include light reflected on the surface of the glass substrate 200 , and the scattered light may include light reflected from a surface defect or the like of the glass substrate 200 .
[0063] The second camera 50 is arranged substantially perpendicular to the surface of the glass substrate 200, and the second light source 40 is arranged so that the optical axis is at a second angle θ2 relative to the surface of the glass substrate 200. In addition, in the illustrated embodiment, along the direction in which the apparatus 100 is configured to move the glass substrate 200 on the platform 10, the second light source 40 is positioned before the second camera 50 (i.e., in front of the second camera 50). In this state, when the light irradiated from the second light source 40 is totally reflected on the surface of the glass substrate 200, an image with a gray background can be captured. In addition, when the light irradiated from the second light source 40 is irregularly reflected by a portion of the glass substrate 200 and generates scattered light, an image partially shaded on a gray background can be captured.
[0064] The second optical system configured as described above serves as a modified dark field method, and may be advantageous for detecting defects having a height or depth, such as protrusions or jagged indentations.
[0065] The defect detection unit 60 may include: a weight calculation unit 60a, which is used to calculate the standard deviation of pixels in each area in the second image, and calculate the weight corresponding to each pixel using the standard deviation; an image generation unit 60b, which is used to apply the weight to the first image to generate a corrected first image; and a defect determination unit 60c, which is used to determine defects of the glass substrate 200 using the corrected first image.
[0066] The embodiments of the present disclosure will be described in more detail with reference to a method for inspecting a glass substrate using the apparatus 100 for inspecting a glass substrate. Figure 1 and Figure 2 The configuration of the apparatus 100 for inspecting a glass substrate is described.
[0067] Figure 3 is a flowchart illustrating a method for inspecting a glass substrate according to an embodiment of the present invention.
[0068] refer to Figure 3 First, a glass substrate 200 as an object to be inspected may be placed on the stage 10 ( S10 ).
[0069] Figure 4 is a plan view of an example of a glass substrate that may be inspected for defects (eg, voids, cracks, scratches, and / or foreign matter) according to an embodiment of the present disclosure.
[0070] Although Figure 4 A rectangular glass substrate 200 is shown in FIG. 2 , but the glass substrate 200 may be a square, a polygon, or a circle.
[0071] refer to Figure 4, the glass substrate 200 may be disposed such that the long sides are parallel to the moving direction of the stage 10, and the short sides intersect (e.g., are perpendicular to) the moving direction of the stage 10. In addition, the entire area of the glass substrate 200 may be divided into a plurality of areas defined in a linear shape in a direction substantially perpendicular to the moving direction of the stage 10. For example, the glass substrate 200 may include a first area 200-1 to an m-th area 200-m defined in a linear shape, where m is a natural number.
[0072] In the embodiments, for convenience, a plurality of regions are described as being linear, but the shape and number of the regions may be changed as needed.
[0073] refer to Figure 1 and Figure 3 , the first light source 20 may irradiate light to the first region 200 - 1 of the glass substrate 200, and the first camera 30 may capture a first image of the first region 200 - 1 of the glass substrate 200 (S20).
[0074] refer to Figure 2 and Figure 3 , the glass substrate 200 passes through the platform 10 in a direction (such as Figure 2 After moving upward (as indicated by the arrow in FIG. 2 ), the second light source 40 may irradiate light to the first area 200-1 of the glass substrate 200, and the second camera 50 may capture a second image of the first area 200-1 of the glass substrate 200 (S30).
[0075] In an embodiment, the first image and the second image may be continuously obtained in a state where both the first optical system and the second optical system are operated and the glass substrate 200 is moved by the stage 10 .
[0076] The first image and the second image may be provided to the defect detection unit 60 .
[0077] Figure 5 It shows that according to Figure 3 The inspection method shown in FIG. 1 is a flowchart of a task of calculating the weight of the second image through the operation of the defect detection unit 60 . Figure 6 is Figure 3 The second image 50a acquired during the method and showing the Figure 5 The task shown in FIG. 5 is a graphical example of calculating the grayscale values of pixels in a region of the second image 50 a .
[0078] refer to Figures 3 to 5 , the weight calculation unit 60a of the defect detection unit 60 may calculate weights corresponding to pixels using the second image 50a (S40).
[0079] First, during the task S40 of calculating weights, the weight calculation unit 60a may divide the second image 50a into a plurality of regions (eg Figure 6 As shown), and calculate the grayscale value of the pixels in each area (S41).
[0080] refer to Figure 6 , each pixel in each region can have a grayscale value X i (For example, i is 1 to 20). Gray value X i Can be a value in the range of 0 to 255.
[0081] exist Figure 6 , for ease of explanation, the number of pixels is shown as 20, but the number of pixels is not limited thereto.
[0082] Next, during the task S40 of calculating weights, the weight calculation unit 60 a may calculate the standard deviation of the pixels of each region using Formula 3 ( S42 ).
[0083] Formula 3
[0084]
[0085] Where s is the standard deviation of the pixels in a given area of the second image 50a, n is the number of pixels in a given area of the second image 50a, and X i is the grayscale value of the i-th pixel (i is a natural number from 1 to n) in a given area of the second image 50a, and is the average grey value of the pixels in a given area of the second image 50a.
[0086] Figure 7 3 is a diagram showing an example of the standard deviation of pixels in the area of the second image 50 a calculated by Formula 3.
[0087] refer to Figure 7 , portion A having a relatively low standard deviation may be a normal region having no defects, and portion B having a relatively high standard deviation may be a region having a high possibility of having defects.
[0088] Next, during task S40 of calculating weights, the weight calculation unit 60 a may calculate weights corresponding to pixels using standard deviations ( S43 ).
[0089] Figure 8 It is shown based on Figure 7 0 is a graph of an example of weights calculated for pixels of the second image 50a using standard deviations.
[0090] The standard deviation of the pixels may be converted into weights according to, for example, a reference value table.
[0091] Fig. 9 is a graph for calculating weights from standard deviations of pixels of the second image 50a drawn according to one embodiment of a reference value table.
[0092] exist Figure 7 In the example, the standard deviation 2.3 of the area determined to be free of defects may be defined as a reference value of 1.0. A corresponding weight may be assigned to each of the standard deviation lower than the standard deviation 2.3 and the standard deviation higher than the standard deviation 2.3. In an embodiment, the standard deviation lower than the standard deviation 2.3 and the standard deviation higher than the standard deviation 2.3 may be weighted at different ratios (slopes).
[0093] refer to Fig. 9 , the ratio (slope) can be expressed as a linear change according to an experimental value or an empirical value. For example, a standard deviation lower than the standard deviation of 2.3 can indicate that the possibility of a defect may be low, and a standard deviation higher than the standard deviation of 2.3 can indicate that the possibility of a defect may be high.
[0094] For example, a weight of a gentle slope may be assigned to a standard deviation where the likelihood of a defect is low, and a weight of a steep slope may be assigned to a standard deviation where the likelihood of a defect is high, thereby further emphasizing the size and contour of the defect.
[0095] Reference again Figure 2 and Figure 3 , the image generation unit 60b of the defect detection unit 60 may apply the weights to corresponding pixels of the first image to generate a corrected first image (S50).
[0096] For example, the corrected first image may be generated by multiplying the weights by the grayscale values of corresponding pixels of the first image, respectively.
[0097] Fig. 10A shows an example of a first image, and Fig. 10B An example of a corrected first image is shown.
[0098] exist Fig. 10A In the image, the outline of the defect (bright part) existing in a part of the area is somewhat unclear. Fig. 10B In the image, the size and outline of the defect (bright part) are clearer.
[0099] The first image obtained by the first optical system is conducive to binarization of image data, but may have varying brightness or low sharpness depending on the position and shape of the defect. On the other hand, the second image obtained by the second optical system can be conducive to detecting the outline of the defect.
[0100] By applying the weight calculated from the second image to the first image through the above process, the difference in sharpness between the portion without defects and the portion with defects can be clarified.
[0101] Next, the defect determination unit 60 c of the defect detection unit 60 may detect defects of the glass substrate 200 using the corrected first image ( S60 ).
[0102] Fig.11 It is shown by Figure 3 The operation of the defect determination unit 60c used in task S60 of the method shown in the flowchart of the tasks performed.
[0103] refer to Fig.11 The defect determination unit 60c may compare the grayscale value of the pixel of each region of the corrected first image with a threshold value to binarize the grayscale value (S61). Through binarization, each grayscale value may be converted into "0" or "1" or "true" or "false".
[0104] For example, grayscale values below a threshold may be converted to "0" or "true," and grayscale values above a threshold may be converted to "1" or "false."
[0105] The threshold value is a grayscale reference value used for defect determination and may be one or more values within the range of 0 to 255.
[0106] The defect determining unit 60c may connect pixels whose binarization values are "1" or "false" (S62), and compare the cumulative length of the pixels connected to each other with a length reference value (S63).
[0107] When pixels having a binary value of "1" or "false" are connected to each other, only pixels within a predetermined range (distance) may be connected.
[0108] When the cumulative length of pixels connected to each other is greater than the length reference value, the defect determining unit 60c may determine a defect (S64) and output the position and shape of the defect on the glass substrate 200 through, for example, a display unit.
[0109] For example, the defect determining unit 60c may not determine a defect when the cumulative length of pixels connected to each other is less than a length reference value, and may determine a defect when the cumulative length of pixels connected to each other is greater than the length reference value.
[0110] As another embodiment, a defect may be determined not only by comparing the cumulative length of pixels connected to each other with a length reference value but also by comparing the cumulative area (cumulative size) of pixels connected to each other with an area reference value.
[0111] As another embodiment, defects may be determined through visual inspection by a user.
[0112] The process input rate of glass substrates having defects that do not affect the process can be increased through the above process, and the loss of manufacturing costs due to erroneous determination can be effectively prevented or at least reduced.
[0113] By repeating the above process from the first region 200 - 1 to the m-th region 200 - m of the glass substrate 200 , the entire region of the glass substrate 200 may be inspected.
[0114] As described above, embodiments of the present invention use two different optical systems to obtain two different digital images. Weights are calculated from one image and applied to the other image, so that the size and contour of the defect can be determined.
[0115] Defects can be detected accurately and easily, and erroneous judgments based on the angle of the optical system and the position and shape of the defect can be reduced. Therefore, the process input rate of the glass substrate can be increased, the defect occurrence rate can be reduced, and the loss of manufacturing costs can be effectively prevented.
[0116] As described above, embodiments of the present invention have been disclosed by detailed description and accompanying drawings. It should be understood that the terms used herein are only for the purpose of describing the present invention and are not intended to limit the scope of the present invention described in the claims. Therefore, it will be understood by those skilled in the art that various modifications and equivalent embodiments are possible without departing from the scope of the present invention. Therefore, the true scope of the present invention should be determined by the technical ideas of the appended claims.
Claims
1. A device for inspecting a glass substrate, the device include: a platform configured to support the glass substrate; a first light source, configured to irradiate light onto the surface of the glass substrate at a first angle; a first camera for capturing scattered light of the light irradiated from the first light source; a second light source, configured to irradiate light onto the surface of the glass substrate at a second angle greater than the first angle; a second camera for capturing reflected light and scattered light of the light irradiated from the second light source; a defect detection unit, configured to detect defects of the glass substrate using a first image provided by the first camera and a second image provided by the second camera, Wherein, the defect detection unit comprises: a weight calculation unit, the weight calculation unit being used to calculate a standard deviation of pixels in each of a plurality of regions in the second image, and to calculate a weight for each of the plurality of regions using the standard deviation; an image generating unit, configured to apply the weights to the first image to generate a corrected first image; and a defect determining unit for determining a defect of the glass substrate using the corrected first image, and The weight calculation unit is configured to divide the second image into the plurality of regions, and calculate the standard deviation using the grayscale value and the average grayscale value of the pixels in each of the plurality of regions.
2. The device according to claim 1, in, The stage includes a driving unit for moving the glass substrate in one direction.
3. The device according to claim 1, in, The first camera and the second camera are arranged in a direction perpendicular to the platform.
4. The device according to claim 1, in, Each of the first camera and the second camera is a line scan camera.
5. The device according to claim 1, in, Each of the first camera and the second camera includes one of a CMOS and a CCD.
6. The device according to claim 1, in, The first light source and the second light source are configured to irradiate the light in a line form.
7. The device according to claim 6, in, Each of the first light source and the second light source includes a plurality of LEDs arranged in a bar shape.
8. The device according to claim 1, in, Each of the standard deviations is calculated by the following formula 1: Formula 1 Wherein, s is the standard deviation of one of the multiple regions, n is the number of pixels in the one region, and X i is the grayscale value of the i-th pixel of the region, i is a natural number from 1 to n, and is the average grayscale value of the pixels in the one area.
