TW200426497A - Method for detecting the position of a defect in a glass substrate in a depth direction - Google Patents
Method for detecting the position of a defect in a glass substrate in a depth direction Download PDFInfo
- Publication number
- TW200426497A TW200426497A TW092137456A TW92137456A TW200426497A TW 200426497 A TW200426497 A TW 200426497A TW 092137456 A TW092137456 A TW 092137456A TW 92137456 A TW92137456 A TW 92137456A TW 200426497 A TW200426497 A TW 200426497A
- Authority
- TW
- Taiwan
- Prior art keywords
- glass substrate
- defect
- camera
- focal plane
- gradient index
- Prior art date
Links
- 239000000758 substrate Substances 0.000 title claims abstract description 143
- 230000007547 defect Effects 0.000 title claims abstract description 142
- 239000011521 glass Substances 0.000 title claims abstract description 142
- 238000000034 method Methods 0.000 title claims abstract description 42
- 239000000463 material Substances 0.000 claims 1
- 238000001514 detection method Methods 0.000 description 29
- 239000012535 impurity Substances 0.000 description 14
- 238000010586 diagram Methods 0.000 description 10
- 238000004088 simulation Methods 0.000 description 8
- 238000002474 experimental method Methods 0.000 description 6
- 238000007689 inspection Methods 0.000 description 4
- 238000004519 manufacturing process Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000005286 illumination Methods 0.000 description 3
- 238000011179 visual inspection Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005401 electroluminescence Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 239000013078 crystal Substances 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 230000003631 expected effect Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000010309 melting process Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G02—OPTICS
- G02F—OPTICAL DEVICES OR ARRANGEMENTS FOR THE CONTROL OF LIGHT BY MODIFICATION OF THE OPTICAL PROPERTIES OF THE MEDIA OF THE ELEMENTS INVOLVED THEREIN; NON-LINEAR OPTICS; FREQUENCY-CHANGING OF LIGHT; OPTICAL LOGIC ELEMENTS; OPTICAL ANALOGUE/DIGITAL CONVERTERS
- G02F1/00—Devices or arrangements for the control of the intensity, colour, phase, polarisation or direction of light arriving from an independent light source, e.g. switching, gating or modulating; Non-linear optics
- G02F1/01—Devices or arrangements for the control of the intensity, colour, phase, polarisation or direction of light arriving from an independent light source, e.g. switching, gating or modulating; Non-linear optics for the control of the intensity, phase, polarisation or colour
- G02F1/13—Devices or arrangements for the control of the intensity, colour, phase, polarisation or direction of light arriving from an independent light source, e.g. switching, gating or modulating; Non-linear optics for the control of the intensity, phase, polarisation or colour based on liquid crystals, e.g. single liquid crystal display cells
- G02F1/1306—Details
- G02F1/1309—Repairing; Testing
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/89—Investigating the presence of flaws or contamination in moving material, e.g. running paper or textiles
- G01N21/892—Investigating the presence of flaws or contamination in moving material, e.g. running paper or textiles characterised by the flaw, defect or object feature examined
- G01N21/896—Optical defects in or on transparent materials, e.g. distortion, surface flaws in conveyed flat sheet or rod
Abstract
Description
200426497 玖、發明說明： 【發明戶斤屬之技術領域1 發明領域 本發明涉及一種檢測存在於玻璃基板内的内部缺陷的 5 深度方向位置的方法，更詳細地說，涉及一種玻璃基板内 缺陷的深度方向位置檢測方法，可以使用在使攝影機 (camera)的焦點面（focal plane)從玻璃基板的表面向玻璃基 板内移動的同時，處理由攝影機拍攝的缺陷的圖像而計算 出的梯度指標(Gradient Indicator)，與照明的亮度、缺陷的 10 大小、形狀、邊界、厚度等無關地正確計算缺陷的深度方 向位置，而且，對微細的缺陷也能通過該正確的位置獲得 鮮明的缺陷圖像。 I：先前技術3 發明背景 15 一般在TFT-LCD、PDP、EL (Electro-Luminescence)等 平板顯示器的製造領域中使用的玻璃基板上，若其中存在 微小空洞、微小裂痕、雜質等微細的缺陷，則在製品的使 用環境中，容易由衝擊、熱應變等造成破損。爲此，在用 於具有高可靠性的玻璃基板的生産中，該缺陷的檢測是非 20 常重要的。 作爲玻璃基板的缺陷檢測方法，依靠檢測者感覺的肉 眼檢測法被廣泛使用。隨著玻璃基板的大型化，該肉眼檢 測法顯現出與檢測的正確性和檢測所用時間相關的界限。 因此，爲了克服檢測玻璃基板缺陷的肉眼檢測法的界 5 限，有必要開發自動檢测法。 作爲自動檢測法，可以例舉在使用CCD(電荷耦合器件) 攝影機檢測車輛用玻璃、玻璃瓶等的檢測步驟中使用的機 械影像(Machine Vision)技術。 這樣的使用CCD攝影機的機械影像檢測法具有能應用 於透明且表面光滑的玻璃、能進行非接觸檢測、成本比較 低的優點。 但疋’使用CCD攝影機的機械影像檢測法還存在只能 應用於比較大的缺陷的檢測、不能鮮明地檢測玻璃基板記 憶體在的數百微米以下的缺陷的缺點。 因此，有必要開發一種自動檢測法，能通過檢測玻璃 基板s己憶體在的微細缺陷，來正確判斷該玻璃基板的好壞。 【發明内容】 發明概要 本發明是鑒於上述問題而提出來的，其目的在於提供 一種玻璃基板内缺陷的基板深度方向位置檢測方法，通過 使攝影機的焦點面從玻璃基板的表面向内部移動，並處理 由攝影機拍攝的圖像計算出的亮度梯度指標，與照明的亮 度、缺陷的大小、形狀、邊界、厚度等無關地正確計算缺 陷的深度方向位置，對微細的缺陷也能通過該計算出的缺 陷深度方向的正確位置獲得鮮明的缺陷圖像，由此可以正 確且立即判斷玻璃基板的好壞。 爲了達成上述目的，根據本發明的較佳實施例，本發 明的玻璃基板内缺陷的深度方向位置的檢測方法，使用攝 影機檢測玻璃基板内缺陷的深度方向位置，其特徵在於， 具有以下步驟：第1步驟，使上述攝影機的焦點面與上述缺 陷位於的上述玻璃基板的一個面一致；第2步驟，使上述攝 影機的焦點面從上述玻璃基板的一個面向另一個面移動_ 疋距離里，第3步驟，用使上述焦點面移動了 一定距離的上 述攝影機拍攝上述缺陷；第4步驟，使用由上述攝影機拍攝 的圖像，計算缺陷和背景的邊界的亮度梯度，並計算與上 述攝影機的焦點面從上述玻璃基板的表面移動的距離相對 應的梯度指標GI的值；第5步驟，對上述攝影機的焦點面從 上述玻璃基板的表面移動的距離和上述玻璃基板的厚度進 订比較；第6步驟，在上述第5步驟，當上述攝影機的焦點 面距離上述玻璃基板表面的移動距離大於上述玻璃基板的 厚度時’將與上述梯度指^GI的值之巾的最大值相對應的 距離上述玻璃基板表面的移動距離，決定爲缺陷的深度方 向位置。 很骒丰發明 J Μ運到如下效果，使攝影機的焦點面 f賴基板的表面向内部移動，處理由攝影機拍攝的圖像 "弟又才曰;^ ’使用该梯度指標與照明的亮度、缺陷的 大小、形狀、厚度等無_正確計算缺陷的深度方向位置， 而對微細的缺陷也能通過該計算出的缺陷隸方向的 *减其*獲侍鮮明的缺陷圖像，由此可以正確且立即判斷 玻璃基板的好壞。 圖式簡單說明 第圖疋表不本發明涉及的破璃基板内缺陷的深度方 200426497 向位置檢測方法的流程圖。 第2圖是表示在本發明涉及的玻璃基板内缺陷的深度 方向位置檢測方法中使用的裝置的示意圖。 第3圖是表示在本發明涉及的玻璃基板内缺陷的深度 5 方向位置檢測方法中，使攝影機焦點面與缺陷位於的面一 致時的示意圖。 第4圖是表示在本發明涉及的玻璃基板内缺陷的深度 方向位置檢測方法中，用於說明輪廓線檢測的存在於玻璃 基板内的氣泡的圖像，第4a圖是原圖像，第4b圖是應用了 10 索貝爾濾波器（Sobel Filter)的圖像。 第5圖是表示在本發明涉及的玻璃基板内缺陷的深度 方向位置檢測方法中，用於說明輪扉線檢測的存在於玻璃 基板内的雜質的圖像，第5a圖是原圖像，第5b圖是應用了 索貝爾濾波器的圖像。 15 第6圖由第6a至6e圖組成，是表示在本發明涉及的玻璃 基板内缺陷的深度方向位置檢測方法中，用於說明梯度指 標的圖，是使橢圓形缺陷邊界厚度改變的仿真模型。 第7圖由第7a至7f圖組成，是表示在本發明涉及的玻璃 基板内缺陷的深度方向位置檢測方法中，用於說明梯度指 20 標的圖，是使圓形缺陷邊界厚度改變的仿真模型。 第8圖是表示在本發明涉及的玻璃基板内缺陷的深度 方向位置檢測方法中，與第6a至6e圖的橢圓形缺陷邊界厚 度變化相對應的梯度指標的分佈的曲線圖。 第9圖是表示在本發明涉及的玻璃基板内缺陷的深度 8 200426497 方向位置檢測方法中，與第7a至7f圖的圓形缺陷邊界厚度 變化相對應的梯度指標的分佈的曲線圖。 第10圖是表示在本發明涉及的玻璃基板内缺陷的深度 方向位置檢測方法中，與第6c圖的橢圓形缺陷的背景亮度 5 變化相對應的梯度指標的分佈的曲線圖。 第11圖是表示在本發明涉及的玻璃基板内缺陷的深度 方向位置檢測方法中，與第7c圖的圓形缺陷的背景亮度變 化相對應的梯度指標的分佈的曲線圖。 第12圖是表示在本發明涉及的玻璃基板内缺陷的深度 10 方向位置檢測方法中，梯度指標值(縱軸)與基於攝影機焦點 面z軸移動距離(橫軸）的氣泡的變化的曲線圖。 第13圖由第13a至13c圖組成，是表示在本發明涉及的 玻璃基板内缺陷的深度方向位置檢測方法中，與第12圖所 示的梯度指標值的變化相對應的氣泡的圖像。 15 第14圖是表示在本發明涉及的玻璃基板内缺陷的深度 方向位置檢測方法中，梯度指標值(縱軸)與基於攝影機焦點 面z軸移動距離(橫軸）的雜質的變化的曲線圖。 第15圖由第15a至15c圖組成，是表示在本發明涉及的 玻璃基板内缺陷的深度方向位置檢測方法中，與第14圖所 20 示的梯度指標值的變化相對應的雜質的圖像。 【實施方式3 較佳實施例之詳細說明 以下，基於附圖詳細地說明本發明的較佳實施例。 第1圖是表示本發明涉及的玻璃基板内缺陷的深度方 9 向位置檢測方法的流程圖，第2圖是表示本發明涉及的方法 中使用的裝置的示意圖，第3圖是表示在本發明涉及的方法 中’使攝影機焦點面與缺陷位於的面一致時的操作的示意 圖。 本發明涉及的玻璃基板内缺陷的深度方向位置檢測方 法，具備以下步驟：使攝影機11(例如CCD攝影機）的焦點面 與玻璃基板1的表面lb—致的步驟(s 10)(即，基板1的表面lb 和攝影機11之間的間隔成爲焦距fc);通過朝z方向移送攝影 機11一定距離，例如ΙΟΟμηι，使攝影機丨丨的焦點面向玻璃基 板1的内部移動一定距離的步驟(S20);用攝影機η拍攝缺陷 la的步驟(S30);從用攝影機11拍攝的圖像，計算出攝影機 11的焦點面從玻璃基板1的表面lb移動的距離對應的梯度 指標(Gradient Indicatior:GI)的步驟(S40);對攝影機11的焦 點面從玻璃基板1的表面lb移動的距離和玻璃基板1的厚度 tG進行比較的步驟(S50);以及，作爲該比較結果，在焦點 面向玻璃基板1的表面lb移動的距離在玻璃基板1的厚度tG 以下的情況，反覆進行步驟S20、S30、S40、S50，攝影機 11的焦點面從玻璃基板1的表面lb移動的距離在玻璃基板1 的厚度tG以上的情況，將與最大梯度指標GI的值相對應的 攝影機11的焦點面距離玻璃基板1的表面lb的移動距離Λζ， 定爲缺陷la的深度方向位置d的步驟(S60)。 第2圖是表示在本發明涉及的玻璃基板内缺陷的深度 方向位置檢測方法中使用的裝置10的示意圖，玻璃基板1受 支援台12垂直支援，將攝影機11設置成從玻璃基板1的一側 200426497 通過直線運動裝置13沿著玻璃基板1的深度方向，即z軸方 向自由移動，安裝與攝影機11成同軸的照明裝置14，具備 用於控制攝影機11和直線運動裝置13的電腦15。 攝影機11通過由電腦15控制的直線運動裝置13移動， 5 攝影機11獲得的圖像由電腦15進行圖像處理。 使攝影機11的焦點面與玻璃基板1的表面1 b —致的步 驟（S10)如第3圖所示，使攝影機11的焦點面與已知X軸和y 軸上位置的玻璃基板1的内部缺陷la的位置相對應的玻璃 基板1的表面lb—致。 10 使攝影機11的焦點面與玻璃基板1的表面lb—致之 後，使攝影機11的焦點面向玻璃基板1内移動一定距離 (S20) 〇 使攝影機11的焦點面向玻璃基板1内移動一定距離的 步驟(S20)，攝影機11通過直線運動裝置13的驅動向玻璃基 15 板1側移動一定距離，由此，攝影機11的焦點面向玻璃基板 1内移動一定距離量。 使攝影機11的焦點面向玻璃基板1内移動一定距離 (S20)後，用攝影機11拍攝缺陷la(S30)。 從由攝影機11拍攝的缺陷la的圖像，計算出攝影機11 20 的焦點面從玻璃基板1的表面lb移動的距離相對應的梯度 指標GI。 算出梯度指標GI的步驟(S40)，從由攝影機11拍攝的圖 像計算缺陷la和背景的邊界上的亮度梯度，並算出攝影機 11的焦點面從玻璃基板1的表面lb移動的距離相對應的梯 11 200426497 度指標GI。 爲了算出梯度指標GI ’首先需要從拍攝的圖像分離缺 陷la和背景。該分離方法使用的演算歧輪廓線(_)檢測 法0 5 該輪廓線檢測法是將缺陷1a的輪廓線部分、或缺陷1& 的表面和背景的亮度差變換成1次或2次梯度值來識別輪廓 線的方法，也可以使用索貝爾濾波器（s〇belFilter)、拉普拉200426497 (1) Description of the invention: [Technical Field of the Inventor 1 Field of the Invention The present invention relates to a method for detecting 5 depth-direction positions of internal defects existing in a glass substrate, and more particularly, to a method for detecting defects in a glass substrate. The position detection method in the depth direction may use a gradient index calculated by processing the image of a defect captured by the camera while moving the focal plane of the camera from the surface of the glass substrate into the glass substrate. Gradient Indicator), which accurately calculates the depth direction position of the defect regardless of the brightness of the illumination, the size of the defect, the shape, the boundary, the thickness, etc. In addition, it is possible to obtain a clear defect image from the correct position even for fine defects. I: Prior Art 3 Background of the Invention 15 If a glass substrate generally used in the field of manufacturing a flat panel display such as a TFT-LCD, PDP, or EL (Electro-Luminescence) has minute defects such as tiny voids, tiny cracks, and impurities, In the use environment of the product, it is easy to be damaged by impact, thermal strain, and the like. For this reason, the detection of this defect is very important in the production of glass substrates with high reliability. As a defect detection method of a glass substrate, a visual inspection method relying on the sense of the inspector is widely used. With the increase in the size of glass substrates, this visual inspection method has shown limits related to the accuracy of the inspection and the time required for inspection. Therefore, in order to overcome the limitations of the visual inspection method for detecting glass substrate defects, it is necessary to develop an automatic inspection method. As the automatic detection method, a machine vision (Machine Vision) technique used in a detection step for detecting glass, glass bottles, and the like of a vehicle using a CCD (Charge Coupled Device) camera can be exemplified. Such a mechanical image detection method using a CCD camera has the advantages of being applicable to transparent and smooth glass, non-contact detection, and relatively low cost. However, the mechanical image detection method using a CCD camera has the disadvantages that it can only be applied to the detection of relatively large defects, and cannot detect the defects of the glass substrate memory below a few hundred microns. Therefore, it is necessary to develop an automatic inspection method that can accurately judge the quality of the glass substrate by detecting the micro defects in the glass substrate s. SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and an object thereof is to provide a method for detecting a position in a substrate depth direction of a defect in a glass substrate by moving a focal plane of the camera from the surface of the glass substrate to the inside, and The brightness gradient index calculated by processing the image captured by the camera is used to correctly calculate the depth direction position of the defect regardless of the brightness of the illumination, the size, shape, border, and thickness of the defect. The fine defect can also be calculated through this calculation. By obtaining a clear defect image at the correct position in the defect depth direction, the quality of the glass substrate can be accurately and immediately determined. In order to achieve the above object, according to a preferred embodiment of the present invention, a method for detecting a position in a depth direction of a defect in a glass substrate according to the present invention uses a camera to detect a position in the depth direction of a defect in a glass substrate, which is characterized by having the following steps: 1 step, make the focal plane of the camera coincide with one surface of the glass substrate where the defect is located; 2 step, move the focal plane of the camera from one surface of the glass substrate to the other surface In step 4, the defect is captured by the camera with the focus surface moved a certain distance. In the fourth step, the brightness gradient of the boundary between the defect and the background is calculated using the image captured by the camera, and the focus surface from the camera is calculated. The value of the gradient index GI corresponding to the distance moved by the surface of the glass substrate; in the fifth step, the distance moved by the focal plane of the camera from the surface of the glass substrate and the thickness of the glass substrate are compared and compared; in the sixth step, In the fifth step, when the focal plane of the camera is away from the When the moving distance is larger than the thickness of the glass substrate surface of the glass substrate 'means the moving distance above the maximum gradient napkin ^ GI value of a distance corresponding to the glass substrate surface, is determined as a depth direction position of the defect. It was discovered by JF that the following effects were achieved by moving the focal plane of the camera to the inside of the substrate, and processing the images captured by the camera. "I use this gradient index and the brightness of the lighting, The size, shape, and thickness of the defect are not correctly calculated in the depth direction of the defect, and for fine defects, a sharp defect image can be obtained by * minus * the calculated direction of the defect. And immediately judge the quality of the glass substrate. BRIEF DESCRIPTION OF THE DRAWINGS The figure shows a flowchart of a method for detecting the depth of a defect in a broken glass substrate according to the present invention. Fig. 2 is a schematic view showing an apparatus used in a method for detecting a position in the depth direction of a defect in a glass substrate according to the present invention. Fig. 3 is a schematic diagram showing a method for detecting the depth 5-direction position of a defect in a glass substrate according to the present invention, when the focal plane of the camera is aligned with the plane on which the defect is located. FIG. 4 is an image showing a bubble existing in the glass substrate for detecting contour lines in a method for detecting a depth direction position of a defect in a glass substrate according to the present invention. FIG. 4a is an original image, and FIG. 4b The picture is an image with a 10 Sobel Filter applied. FIG. 5 is an image showing a depth direction position detection method of a defect in a glass substrate according to the present invention, and is used to explain an impurity existing in the glass substrate detected by a wheel line; FIG. 5a is an original image; Figure 5b is an image with a Sobel filter applied. 15 Fig. 6 is composed of Figs. 6a to 6e, and is a diagram for explaining a gradient index in a method for detecting a depth direction position of a defect in a glass substrate according to the present invention, and is a simulation model for changing the thickness of an elliptical defect boundary. . Fig. 7 is composed of Figs. 7a to 7f, and is a diagram illustrating a gradient index of 20 in the depth direction position detection method of a defect in a glass substrate according to the present invention. It is a simulation model that changes the thickness of a circular defect boundary. . Fig. 8 is a graph showing the distribution of a gradient index corresponding to a change in the thickness of the elliptical defect boundary in Figs. 6a to 6e in the depth direction position detection method of a defect in a glass substrate according to the present invention. Fig. 9 is a graph showing the distribution of the gradient index corresponding to the change in the thickness of the circular defect boundary in Figs. 7a to 7f in the method for detecting the depth of a defect in a glass substrate according to the present invention. Fig. 10 is a graph showing the distribution of the gradient index corresponding to the change in the background brightness 5 of the elliptical defect in Fig. 6c in the method for detecting the depth direction position of a defect in a glass substrate according to the present invention. Fig. 11 is a graph showing the distribution of the gradient index corresponding to the change in the background brightness of the circular defect in Fig. 7c in the method for detecting the position in the depth direction of a defect in a glass substrate according to the present invention. FIG. 12 is a graph showing changes in a gradient index value (vertical axis) and bubbles based on a z-axis movement distance (horizontal axis) of a focal plane of a camera in a method for detecting a depth 10 direction of a defect in a glass substrate according to the present invention; . Fig. 13 is composed of Figs. 13a to 13c and is an image showing bubbles corresponding to a change in the gradient index value shown in Fig. 12 in the depth direction position detection method of a defect in a glass substrate according to the present invention. 15 FIG. 14 is a graph showing changes in a gradient index value (vertical axis) and impurities based on a z-axis movement distance (horizontal axis) of a focal plane of a camera in a method for detecting a depth direction position of a defect in a glass substrate according to the present invention. . Fig. 15 is composed of Figs. 15a to 15c, and is an image showing impurities corresponding to changes in the gradient index value shown in Fig. 14 and Fig. 20 in the depth direction position detection method of a defect in a glass substrate according to the present invention. . [Embodiment 3 Detailed Description of Preferred Embodiments] Hereinafter, preferred embodiments of the present invention will be described in detail based on the drawings. FIG. 1 is a flowchart showing a method for detecting a depth 9-direction position of a defect in a glass substrate according to the present invention, FIG. 2 is a schematic diagram showing an apparatus used in the method according to the present invention, and FIG. 3 is a diagram showing the device used in the present invention In the method involved, a schematic diagram of the operation when the focal plane of the camera is aligned with the plane on which the defect is located. The method for detecting a position in the depth direction of a defect in a glass substrate according to the present invention includes the following steps: a step (s 10) of matching the focal surface of the camera 11 (for example, a CCD camera) with the surface lb of the glass substrate 1 (that is, the substrate 1) The distance between the surface lb and the camera 11 becomes the focal length fc); by moving the camera 11 a certain distance in the z direction, for example 100 μηι, the step of moving the focus of the camera 丨 丨 toward the inside of the glass substrate 1 by a certain distance (S20); Step (S30) of the camera η capturing the defect la (S30); Step of calculating a gradient index (Gradient Indicatior: GI) corresponding to the distance moved by the focal plane of the camera 11 from the surface lb of the glass substrate 1 from the image captured by the camera 11 ( S40); a step of comparing the distance moved by the focal plane of the camera 11 from the surface lb of the glass substrate 1 and the thickness tG of the glass substrate 1 (S50); and as a result of this comparison, the focal point faces the surface lb of the glass substrate 1 When the moving distance is less than the thickness tG of the glass substrate 1, steps S20, S30, S40, and S50 are repeatedly performed. The focal plane of the camera 11 moves from the surface lb of the glass substrate 1. When the thickness tG of the glass substrate 1 or more, the moving distance Λζ of the focal plane of the camera 11 corresponding to the value of the maximum gradient index GI from the surface lb of the glass substrate 1 is determined as the depth d of the defect la Step (S60). FIG. 2 is a schematic view showing a device 10 used in a method for detecting a position in the depth direction of a defect in a glass substrate according to the present invention. The glass substrate 1 is vertically supported by a support table 12, and the camera 11 is disposed from one side of the glass substrate 1. 200426497 The linear motion device 13 is free to move along the depth direction of the glass substrate 1, that is, the z-axis direction. An illumination device 14 coaxial with the camera 11 is mounted, and a computer 15 for controlling the camera 11 and the linear motion device 13 is provided. The camera 11 is moved by a linear motion device 13 controlled by a computer 15, and the image obtained by the camera 11 is image processed by the computer 15. The step (S10) of matching the focal plane of the camera 11 with the surface 1 b of the glass substrate 1 is as shown in FIG. 3, and the focal plane of the camera 11 is aligned with the interior of the glass substrate 1 with known positions on the X and y axes. The surface lb of the glass substrate 1 corresponding to the position of the defect la is the same. 10 After the focal surface of the camera 11 is aligned with the surface lb of the glass substrate 1, the focal point of the camera 11 is moved a certain distance toward the glass substrate 1 (S20). The step of moving the focal point of the camera 11 toward the glass substrate 1 a certain distance. (S20), the camera 11 is moved to the glass base 15 plate 1 side by a certain distance by the drive of the linear motion device 13, so that the focal point of the camera 11 is moved toward the glass substrate 1 by a certain distance. After moving the focal point of the camera 11 toward the glass substrate 1 by a certain distance (S20), the camera 11 captures a defect la (S30). From the image of the defect la captured by the camera 11, a gradient index GI corresponding to the distance the focal plane of the camera 11 20 moves from the surface 1 b of the glass substrate 1 is calculated. In the step (S40) of calculating the gradient index GI, the brightness gradient on the boundary between the defect la and the background is calculated from the image captured by the camera 11, and the distance corresponding to the focal plane of the camera 11 moved from the surface 1b of the glass substrate 1 is calculated. Ladder 11 200426497 degree indicator GI. In order to calculate the gradient index GI ', it is first necessary to separate the defect la and the background from the captured image. Computational ambiguity contour detection (_) detection method used by this separation method 0 5 This contour detection method is to convert the brightness difference between the contour portion of defect 1a, or the surface and background of defect 1 & into a gradient value of one or two times. To identify the contour line, you can also use Sobel Filter, Laplacian
斯濾波器（Laplacian Filter)、普拉衛特梯度法（Prewitt Gradient Method)、線段增強（Line Segment Enhancement)等 10 多種技術中的任意一種。特別優選使用縱橫方向的輪廓線 檢測特性優異的索貝爾濾波器。 圖像G(x，y)的點(x，y)處的亮度梯度可以用下式的向量 表示。 【數1】 15 VG =Laplacian Filter, Prewitt Gradient Method, Line Segment Enhancement and any of more than 10 technologies. It is particularly preferable to use a Sobel filter having excellent contour detection characteristics in the vertical and horizontal directions. The brightness gradient at the point (x, y) of the image G (x, y) can be represented by a vector of the following formula. [Number 1] 15 VG =
Gx Gy dG dx ^ dGJy 式⑴Gx Gy dG dx ^ dGJy formula ⑴
在式（1)中，VG是點(X,y)處的梯度向量，表示該點處 的亮度的最大變化率方向，輪廓線檢測中最重要的值是該 向量的大小，如果單純地將其作爲梯度，則可以用下式表 示。 20【數2】 VG = mag(VG) = [g2 + G2 ί 式(2)In equation (1), VG is the gradient vector at point (X, y), which indicates the direction of the maximum rate of change of brightness at that point. The most important value in contour detection is the size of this vector. As a gradient, it can be expressed by the following formula. 20 [Number 2] VG = mag (VG) = [g2 + G2 ί formula (2)
Lx y J 12 200426497 在上述式⑺中，VG與每個向…方向的單位長度的 G(x，y)的最大增加率相同。具體實現實際演算法時，爲 省計算時間並節省硬體，多使用下式作爲梯度的計算式。 【數3】 上式(3)是用絕對值表示梯度的近似式。Lx y J 12 200426497 In the above formula (i), the maximum increase rate of VG and G (x, y) per unit length in the direction of ... is the same. When implementing the actual algorithm, in order to save calculation time and save hardware, the following formula is used as the gradient calculation formula. [Equation 3] The above formula (3) is an approximation formula in which the gradient is expressed by an absolute value.
第4a和5a圖分別表示玻璃基板以憶體在的缺陷说 中的氣泡和雜質，應用索貝爾據波器將各圖元的▽⑽變換 成256灰度級再構成該兩圖的圖像，則如第扑和_所示。 1〇同圖所示，由缺陷1冰背景的亮度差，f料認邊界 浮出的情況。如上所述使用索貝爾據波器獲得的輪廓線靈 活運用於缺陷U和背景的分離、及自動難（—ng)。 隨著攝影機11的焦點面與缺陷la的位置X軸和y軸所成 的水平面-致，圖像中缺陷la的輪廊變得明確，缺陷⑽ 15背景的邊界處的亮度梯度增加。因此，著重這點由下式算 出梯度指標GI。Figures 4a and 5a show the bubbles and impurities in the glass substrate based on the defect theory of the memory. The Sober wave device is used to transform the ▽ ⑽ of each pixel into 256 gray levels to form the two images. As shown in the first flutter and _. 10 As shown in the figure, from the difference in brightness of the ice background of defect 1 and f, it is recognized that the boundary is floating. As described above, the contour line obtained using the Sobel wave device is flexibly applied to the separation of the defect U and the background, and automatically difficult (-ng). With the horizontal plane formed by the focal plane of the camera 11 and the position of the X axis and the y axis of the defect la, the contour of the defect la in the image becomes clear, and the brightness gradient at the boundary of the background of the defect ⑽ 15 increases. Therefore, focusing on this point, the gradient index GI is calculated by the following formula.
【數4】 GI(z) = Σ image atz |VG|[Number 4] GI (z) = Σ image atz | VG |
-VG 式（4) 在上述式⑷中，可以由式(3)求出VG的啦，y)位置的圖 20兀和8方向相鄰的圖元的亮度梯度，VG腿、VG*是一 圖像内梯度值的最大、最小值。 ▽G自身可受到周圍平面背景亮度的影響，所以，應用 ▽ G麵與VGmin的差值來減少對亮度絕對值的影響，求出 13 、樣冲算的各值的合計，是爲了合計各圖元的各梯度值來 得到對圖像全體的梯度值。 5十算出與攝影機11的焦點面從玻璃基板1的表面lb移 動的距離相對應的梯度指標⑽)之後，對攝影機η的焦點 面從玻璃基板1的表面u移動的距離和玻璃基板1的厚度tG 進行比較(S50)。 在對攝影機11的焦點面從玻璃基板1的表面lb移動的 距離和玻璃基板1的厚度tG進行比較的步驟(⑽），攝影機^ 的焦點面從玻璃基板丨的表面师動的距離在玻璃基板W 厚度tG以下的情況，反覆進行使攝影機n的焦點面向玻璃基 板1内移動一定距離的步驟(S2〇)，對每個攝影機η的焦點面 從玻璃基板1的表面1 b移動的距離算出梯度指標GI(S4〇)。 在對攝影機11的焦點面從玻璃基板i的表面lb移動的 距離和玻璃基板1的厚度。進行比較的步驟(S50)，攝影機U 的焦點面從玻璃基板1的表面lb移動的距離超過玻璃基板i 的厚度tG時，將與梯度指標⑺之中最大的值相對應的攝影 機11的焦點面距離玻璃基板1的表面lb的移動距離^2；，定 爲缺陷la的深度方向位置d(S60)。 這樣一來，最大梯度指標GI對應的攝影機11焦點面距 離玻璃基板1的表面lb的移動距離Az相當於缺陷la的深度 方向位置，即距離玻璃基板1的表面lb的深度d，所以，使 攝影機11的焦點面與梯度指標GI中最大值對應的攝影機” 的焦點面距離玻璃基板1的表面lb的移動距離進行位置 配合，通過最佳閾值化和附帶標記(Labeling)，可以向檢測 200426497 者提供鮮明的缺陷la的圖像。 在本發明涉及的玻璃基板内缺陷的深度方向位置檢測 方法中，爲了決定玻璃基板1内的缺陷la的位置，將梯度指 標GI作爲對焦（in focus)判定指標使用，但爲了表示作爲玻 5 璃基板内的缺陷檢測指標是妥當的，進行了以下試驗。 (實驗1) 生産玻璃基板1時發生最多的缺陷la的代表例有氣泡 (blister)和雜質（inclusion)等。氣泡是在玻璃基板1的製造過 程中的熔融過程中空氣混合而表現的形態，雜質不是非晶 10 質狀態的玻璃，是指結晶性物質存在於玻璃基板1内而言。 該實驗使用電腦，以類似實際缺陷la的形態，使用了第6和 7圖那樣生成的仿真模型。 第6a〜6e圖分別表示具有1、5、10、15的邊界厚度t、 黑色的橢圓形仿真模型，第7a〜7f圖分別表示邊界厚度t爲 15 1、5、10、15、20、黑色的圓形仿真模型。 如第8圖所示，可知背景亮度一定的情況下，對應於第 6a〜6e圖的橢圓形缺陷邊界厚度t的變化，梯度指標GI的值 的大小幾乎不變。通過從焦點偏移狀態向對焦狀態的聚焦 程度的變化，可以確認梯度指標(^的值的變化的大小明確 20 不同。 如第9圖所示的第7a〜7f圖的圓形缺陷的邊界厚度的 變化對應的梯度指標GI的值的分佈那樣，圓形缺陷的情況 也可以得到這樣的結果。 由於缺陷邊界厚度固定爲一定值，所以背景亮度爲256 15 200426497 灰度級變化爲50、100、150、200、250，而且，對梯度指 標GI的值的變化進行仿真。其結果，可知缺陷邊界厚度t爲 1〇，表示第6c圖的橢圓形缺陷背景亮度變化相對應的梯度 指標GI的值的分佈的第1〇圖，其缺陷邊界厚度t爲10，如表 5 示第7c圖的圓形缺陷背景亮度變化相對應的梯度指標GI的 值的分佈的第11圖那樣，即使背景亮度增加，梯度指標GI 也幾乎不變化。 另外，可知對應於從焦點偏移狀態向對焦狀態的聚焦 程度的變化，可以明確地體現梯度指標GI的值的大小的變 10 化。 這樣的結果是表示在檢測玻璃基板1内的缺陷時使用 梯度指標GI是妥當的。梯度指標GI有最大值時，通過使攝 影機11的焦點面與缺_la存在的缺陷面一致，可以對缺陷 1 a獲得鮮明的圖像。 15 (實驗2) 爲了調查本發明涉及的玻璃基板内缺陷深度方向位置 檢測方法的實際應用及作用，使用第2圖的裝置對玻璃基板 1進行缺陷la的檢測的實驗。 圖像獲得所使用的影像板(vision board)是邁創(Matrox) 20 公司制的Meteor Π，攝影機11使用三星BE360ED單色 (Monochrome)的CCD攝影機，照明裝置14使用與攝影機同 軸的超高亮度LED，電腦15使用AMD Duron 1GHz。 第12圖表示對數百微米左右的氣泡的梯度指標gi的值 (縱軸）的z軸方向移動距離（橫軸）相對應的變化，在z軸每 16 200426497 ΙΟΟμιη進行移動來獲得圖像，對各個獲得的圖像計算梯度 指標GI的值。 如第12圖所示，可知梯度指標⑺的值的大小在某—特 定位置急劇增大而減少，距離原點的距離在llmm附近有 5 最大值。 第13a〜13c圖是第12圖中梯度指標⑺有最大值時的圖 像和有接近最大值的梯度指標GI的值時的圖像，分別是距 離原點的距離爲1.0mm、1.1mm、i.2mm時的氣泡的圖像。 如第13a〜13c圖所示，可知在梯度指標以有最大值的 10情況、距離原點的距離爲1.lmm時，攝影機丨丨的焦點面與氣 泡存在的平面-致，而且可以使用梯度指標GI的值判定玻 璃基板1内像氣泡這樣的缺陷la存在的平面。 第14圖表不對數十微米左右的微細雜質的實驗結果。 詳細進行說明，就是表示梯度指標GI的值對攝影機u的焦 15點面的Z轴方向移動距離(橫轴)對應的雜質的變化，雜質的 大小比在上述實驗中使用的氣泡小很多，樣子也不同，但 已知與氣泡一樣，攝影機Η的焦點面到達缺陷面’而且存 在梯度指標GI的值比周圍的梯度指標GI的值大很多的部 分。 第⑸15’是第14圖中梯度指標GI有最大值時的圖 像矛有最大值附近的梯度指標以的值時的圖像，分別是距 離原點的距離爲0.9mm , Λ 11、1.0mm、1.1mm時的雜質的圖像。 如第15a〜15c圖恥一 斤不’可知在梯度指標GI有最大值的 情況、距離原點的矩齙这η ^ 雕為l_〇mm時，攝影機11的焦點面與雜 17 質存在的平面一致，而且可以使用梯度指標GI的值判定破 璃基板1内像雜質這樣的缺陷la存在的平面。 如上所述，本發明涉及的玻璃基板内缺陷的深度方向 位置檢測方法，爲了檢測内部存在缺陷h的玻璃基板1的表 面lb到深度方向的位置，使用利用攝影機u的自動調焦技 術，將梯度指標GI作爲對缺陷la存在面的對焦判定基準使 用。 在本發明涉及的玻璃基板内缺陷的深度方向位置檢測 方法中’爲了驗證使用梯度指標GI的妥當性，進行仿真， 來檢測與缺陷la的邊界厚度的變化和背景亮度的變化相對 應的梯度指標GI的值的變化，並說明其結果。另外，本發 明中提示過的算法可以應用於實際步驟中，而且，對是否 具有預期的作用也進行了實驗，並詳細說明了其結果。 因此，隨著從焦點偏移狀態轉移至對焦狀態，梯度指 標GI的值增大，梯度指標GI是最大值時，將攝影機u的焦 點面距離玻璃基板1的表面lb的移動距離Δζ判定爲缺陷ia 的深度d，由此可以正確判定是否存在微細的缺陷u，而且 可以獲得缺陷la的鮮明的圖像。 以上說明了本發明的較佳實施方式，但本發明的權利 要求範圍不限於此，本領域的技術人員可以進行各種改變。 C圖式簡單:¾¾明；j 第1圖是表示本發明涉及的玻璃基板内缺陷的深度方 向位置檢測方法的流程圖。 第2圖是表不在本發明涉及的玻璃基板内缺陷的深度 200426497 方向位置檢測方法中使用的裝置的示意圖。 第3圖是表示在本發明涉及的玻璃基板内缺陷的深度 方向位置檢測方法中，使攝影機焦點面與缺陷位於的面一 致時的示意圖。 5 第4圖是表示在本發明涉及的玻璃基板内缺陷的深度 方向位置檢測方法中，用於說明輪廓線檢測的存在於玻璃 基板内的氣泡的圖像，第4a圖是原圖像，第4b圖是應用了 索貝爾濾波器（Sobel Filter)的圖像。 第5圖是表示在本發明涉及的玻璃基板内缺陷的深度 10 方向位置檢測方法中，用於說明輪廓線檢測的存在於玻璃 基板内的雜質的圖像，第5a圖是原圖像，第5b圖是應用了 索貝爾濾波器的圖像。 第6圖由第6a至6e圖組成，是表示在本發明涉及的玻璃 基板内缺陷的深度方向位置檢測方法中，用於說明梯度指 15 標的圖，是使橢圓形缺陷邊界厚度改變的仿真模型。 第7圖由第7a至7f圖組成，是表示在本發明涉及的玻璃 基板内缺陷的深度方向位置檢測方法中，用於說明梯度指 標的圖，是使圓形缺陷邊界厚度改變的仿真模型。 第8圖是表示在本發明涉及的玻璃基板内缺陷的深度 20 方向位置檢測方法中，與第6a至6e圖的橢圓形缺陷邊界厚 度變化相對應的梯度指標的分佈的曲線圖。 第9圖是表示在本發明涉及的玻璃基板内缺陷的深度 方向位置檢測方法中，與第7a至7f圖的圓形缺陷邊界厚度 變化相對應的梯度指標的分佈的曲線圖。 19 200426497 第10圖是表示在本發明涉及的玻璃基板内缺陷的深度 方向位置檢測方法中，與第6c圖的橢圓形缺陷的背景亮度 變化相對應的梯度指標的分佈的曲線圖。 第11圖是表示在本發明涉及的玻璃基板内缺陷的深度 5 方向位置檢測方法中，與第7c圖的圓形缺陷的背景亮度變 化相對應的梯度指標的分佈的曲線圖。 第12圖是表示在本發明涉及的玻璃基板内缺陷的深度 方向位置檢測方法中，梯度指標值(縱軸)與基於攝影機焦點 面z軸移動距離(橫軸）的氣泡的變化的曲線圖。 10 第13圖由第13a至13c圖組成，是表示在本發明涉及的 玻璃基板内缺陷的深度方向位置檢測方法中，與第12圖所 示的梯度指標值的變化相對應的氣泡的圖像。 第14圖是表示在本發明涉及的玻璃基板内缺陷的深度 方向位置檢測方法中，梯度指標值(縱軸)與基於攝影機焦點 15 面z軸移動距離(橫軸）的雜質的變化的曲線圖。 第15圖由第15a至15c圖組成，是表示在本發明涉及的 玻璃基板内缺陷的深度方向位置檢測方法中，與第14圖所 示的梯度指標值的變化相對應的雜質的圖像 【圖式之主要元件代表符號表】 15…電腦 d...深度 t...邊界厚度 t〇...厚度 GI...梯度指標 △z...移動距離 1.. .玻璃基板 la. ..内部缺陷 lb. ..表面 11.. .攝影機 12.. .支援台 13.. .直線運動裝置 14…照明裝置 20-VG formula (4) In the above formula (3), the brightness gradient of the adjacent picture elements in the 20th and 8th directions of the VG, y) position can be obtained from the formula (3). The VG legs and VG * are the same. The maximum and minimum value of the gradient value in the image. ▽ G itself can be affected by the brightness of the background of the surrounding plane. Therefore, the difference between the ▽ G surface and VGmin is used to reduce the effect on the absolute value of the brightness. The total of the values calculated by 13 and the sample is calculated in order to total the figures. Element to obtain the gradient value for the entire image. After calculating the gradient index ⑽) corresponding to the distance that the focal plane of the camera 11 moves from the surface lb of the glass substrate 1, the distance of the focal plane of the camera η from the surface u of the glass substrate 1 and the thickness of the glass substrate 1 are calculated. tG is compared (S50). In the step (i) of comparing the distance moved by the focal plane of the camera 11 from the surface 1b of the glass substrate 1 and the thickness tG of the glass substrate 1, the distance of the focal plane of the camera ^ from the surface of the glass substrate When the thickness is less than or equal to tG, the step of moving the focal point of the camera n toward the glass substrate 1 by a certain distance is repeated (S20), and the gradient is calculated for each focal point of the camera η by the distance moved from the surface 1 b of the glass substrate 1. Index GI (S40). The distance moved from the surface lb of the glass substrate i at the focal plane to the camera 11 and the thickness of the glass substrate 1. In a comparison step (S50), when the focal plane of the camera U moves from the surface lb of the glass substrate 1 by a distance exceeding the thickness tG of the glass substrate i, the focal plane of the camera 11 corresponding to the largest value of the gradient index ⑺ The moving distance ^ 2 from the surface lb of the glass substrate 1 is determined as the depth direction position d of the defect la (S60). In this way, the moving distance Az of the focal plane of the camera 11 corresponding to the maximum gradient index GI from the surface lb of the glass substrate 1 corresponds to the position in the depth direction of the defect la, that is, the depth d from the surface lb of the glass substrate 1. The focal plane of 11 corresponds to the camera corresponding to the maximum value in the gradient index GI. The focal plane is positioned to move distance from the surface lb of the glass substrate 1. Through the optimal thresholding and labeling, it can be provided to those who detect 200426497. A sharp image of the defect la. In the depth direction position detection method of a defect in a glass substrate according to the present invention, in order to determine the position of the defect la in the glass substrate 1, the gradient index GI is used as an in focus determination index However, in order to show that the defect detection index in the glass substrate is appropriate, the following tests were performed. (Experiment 1) Representative examples of defects la which occur most during the production of glass substrate 1 are blister and inclusion. Etc. Air bubbles are morphologies that are manifested by air mixing during the melting process in the manufacturing process of the glass substrate 1, and impurities are not non- The glass in the qualitative state of crystal 10 means that a crystalline substance exists in the glass substrate 1. In this experiment, a computer was used, in a form similar to the actual defect la, using the simulation model generated in Figures 6 and 7. Sections 6a ~ Figure 6e shows the oval elliptical simulation model with a border thickness t of 1, 5, 10, and 15, respectively. Figures 7a to 7f show the circles with a border thickness t of 1, 1, 5, 10, 15, 20, and black, respectively. Simulation model. As shown in Fig. 8, it can be seen that when the background brightness is constant, the value of the gradient index GI is almost unchanged corresponding to the change in the thickness t of the elliptical defect boundary in Figs. 6a to 6e. The change in the degree of focus from the moving state to the in-focus state can confirm that the magnitude of the change in the value of the gradient index (^ is clearly 20 different. As shown in FIG. 9, the change in the boundary thickness of the circular defect in FIGS. 7a to 7f corresponds to the change The distribution of the value of the gradient index GI can also be obtained in the case of circular defects. Since the thickness of the defect boundary is fixed to a certain value, the background brightness is 256 15 200426497 The gray level change is 50, 100, 150, 200 250. Furthermore, the change in the value of the gradient index GI was simulated. As a result, it was found that the thickness t of the defect boundary was 10, which represented the first distribution of the value of the gradient index GI corresponding to the change in the brightness of the background of the oval defect in FIG. In figure 10, the defect boundary thickness t is 10, as shown in Table 5 which shows the distribution of the value of the gradient index GI corresponding to the background brightness change of the circular defect in FIG. 7c. Even if the background brightness increases, the gradient index GI It also hardly changes. In addition, it can be seen that a change in the magnitude of the value of the gradient index GI is clearly reflected in accordance with a change in the degree of focus from the focus shift state to the in-focus state. This result indicates that it is appropriate to use the gradient index GI when detecting defects in the glass substrate 1. When the gradient index GI has a maximum value, a sharp image can be obtained for the defect 1 a by making the focal plane of the camera 11 coincide with the defect surface in which the defect _la exists. 15 (Experiment 2) In order to investigate the practical application and function of the method for detecting the position in the depth direction of a defect in a glass substrate according to the present invention, an experiment of detecting a defect 1a was performed on the glass substrate 1 using the apparatus of FIG. 2. The vision board used for image acquisition is a Meteor Π made by Matrox 20. The camera 11 uses a Samsung BE360ED Monochrome CCD camera, and the lighting device 14 uses an ultra-high brightness coaxial with the camera. LED, computer 15 uses AMD Duron 1GHz. FIG. 12 shows the change in the z-axis direction moving distance (horizontal axis) of the value of the gradient index gi (vertical axis) of a bubble of about several hundred micrometers. The z-axis is moved every 16 200426497 100 μm to obtain an image. The value of the gradient index GI is calculated for each obtained image. As shown in Figure 12, it can be seen that the magnitude of the value of the gradient index ⑺ increases sharply at a certain position and decreases, and the distance from the origin has a maximum value of 5 around llmm. Figures 13a to 13c are the images when the gradient index ⑺ in Figure 12 has a maximum value and the image when the gradient index GI is close to the maximum value in Figure 12. The distances from the origin are 1.0mm, 1.1mm, Image of bubbles at i.2mm. As shown in Figures 13a to 13c, it can be seen that when the gradient index is 10 with a maximum value and the distance from the origin is 1.1 mm, the focal plane of the camera and the plane where the bubbles exist are consistent, and the gradient can be used. The value of the index GI determines the plane where defects la such as bubbles exist in the glass substrate 1. Fig. 14 does not show the experimental results for fine impurities in the range of tens of microns. To explain in detail, it means that the value of the gradient index GI changes the impurity corresponding to the Z-axis movement distance (horizontal axis) of the 15-point focal plane of the camera u. The size of the impurity is much smaller than the bubbles used in the above experiment. It is also different, but it is known that, like the bubbles, the focal plane of the camera 到达 reaches the defective surface, and there is a portion where the value of the gradient index GI is much larger than the value of the surrounding gradient index GI. Figure ⑸15 'is the image in Figure 14 when the gradient index GI has a maximum value. The image when the spear has a value near the maximum value of the gradient indicator. The distance from the origin is 0.9mm, Λ 11, 1.0mm. , Image of impurities at 1.1 mm. As shown in Figures 15a to 15c, it can be seen that when the gradient index GI has a maximum value and the moment from the origin is η ^, the focal plane of the camera 11 and the miscellaneous 17 exist. The planes are the same, and the value of the gradient index GI can be used to determine the plane where defects la such as impurities are present in the broken glass substrate 1. As described above, in the depth direction position detection method of a defect in a glass substrate according to the present invention, in order to detect the position in the depth direction of the surface lb of the glass substrate 1 with the defect h inside, an automatic focusing technology using a camera u is used to change the gradient The index GI is used as a focus determination criterion for the surface where the defect la exists. In the method for detecting a depth direction position of a defect in a glass substrate according to the present invention, 'in order to verify the validity of using the gradient index GI, a simulation is performed to detect a gradient index corresponding to a change in the boundary thickness of the defect la and a change in the background brightness. Changes in the value of GI and the results. In addition, the algorithm suggested in the present invention can be applied to actual steps, and experiments have been performed on whether it has the expected effect, and the results are explained in detail. Therefore, the value of the gradient index GI increases as the focus shifts from the focus shift state to the focus state. When the gradient index GI is the maximum value, the moving distance Δζ of the focal plane of the camera u from the surface lb of the glass substrate 1 is determined as a defect The depth d of ia can be used to correctly determine whether there is a fine defect u, and a sharp image of the defect la can be obtained. The preferred embodiments of the present invention have been described above, but the scope of claims of the present invention is not limited thereto, and those skilled in the art can make various changes. The diagram of C is simple: ¾¾; j FIG. 1 is a flowchart showing a method for detecting a position in the depth direction of a defect in a glass substrate according to the present invention. FIG. 2 is a schematic diagram showing an apparatus used in a method for detecting a position of a defect in a glass substrate according to the present invention. Fig. 3 is a schematic view showing a method for detecting a depth direction position of a defect in a glass substrate according to the present invention, when a focal plane of a camera is aligned with a surface on which the defect is located. 5 FIG. 4 is an image showing bubbles in the glass substrate for detecting contour lines in the depth direction position detection method of a defect in a glass substrate according to the present invention, and FIG. 4a is an original image. Figure 4b is an image with a Sobel Filter applied. FIG. 5 is an image showing a depth 10-direction position detection method of a defect in a glass substrate according to the present invention, and is used to explain the contour line detection of impurities existing in the glass substrate. FIG. 5a is an original image. Figure 5b is an image with a Sobel filter applied. Fig. 6 is composed of Figs. 6a to 6e. It is a diagram for explaining the gradient index 15 in the depth direction position detection method of a defect in a glass substrate according to the present invention. It is a simulation model that changes the thickness of an elliptical defect boundary. . Fig. 7 is composed of Figs. 7a to 7f, and is a diagram for explaining a gradient index in a method for detecting a depth direction position of a defect in a glass substrate according to the present invention, and is a simulation model for changing a thickness of a circular defect boundary. Fig. 8 is a graph showing the distribution of a gradient index corresponding to a change in the thickness of the elliptical defect boundary in Figs. 6a to 6e in the method for detecting the depth 20 direction of a defect in a glass substrate according to the present invention. Fig. 9 is a graph showing the distribution of the gradient index corresponding to the change in the thickness of the circular defect boundary in Figs. 7a to 7f in the depth direction position detection method of a defect in a glass substrate according to the present invention. 19 200426497 Fig. 10 is a graph showing the distribution of the gradient index corresponding to the background luminance change of the elliptical defect in Fig. 6c in the method for detecting the depth direction position of a defect in a glass substrate according to the present invention. Fig. 11 is a graph showing the distribution of the gradient index corresponding to the change in the background brightness of the circular defect in Fig. 7c in the method for detecting the depth 5-direction position of a defect in a glass substrate according to the present invention. Fig. 12 is a graph showing changes in a gradient index value (vertical axis) and bubbles based on a z-axis movement distance (horizontal axis) of a focal plane of a camera in a method for detecting a depth direction position of a defect in a glass substrate according to the present invention. 10 FIG. 13 is composed of FIGS. 13a to 13c, and is an image showing bubbles corresponding to changes in the gradient index value shown in FIG. 12 in the depth direction position detection method of a defect in a glass substrate according to the present invention. . FIG. 14 is a graph showing changes in a gradient index value (vertical axis) and impurities based on a 15-plane z-axis movement distance (horizontal axis) of a defect in a glass substrate according to the present invention in a method for detecting a position in the depth direction of a defect in a glass substrate. . FIG. 15 is composed of FIGS. 15a to 15c, and is an image showing impurities corresponding to changes in the gradient index value shown in FIG. 14 in the depth direction position detection method of defects in a glass substrate according to the present invention [ The main components of the figure represent the symbol table] 15 ... computer d ... depth t ... boundary thickness t0 ... thickness GI ... gradient index △ z ... moving distance 1 ... glass substrate la. .. Internal defect lb. .. Surface 11......................
Claims (1)
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| KR10-2003-0035066A KR100535569B1 (en) | 2003-05-31 | 2003-05-31 | Method for detecting the position of the depth direction of the defect of a glass substrate |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| TW200426497A true TW200426497A (en) | 2004-12-01 |
| TWI311687B TWI311687B (en) | 2009-07-01 |
Family
ID=34056780
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| TW092137456A TWI311687B (en) | 2003-05-31 | 2003-12-30 | Method for detecting the position of a defect in a glass substrate in a depth direction |
Country Status (3)
| Country | Link |
|---|---|
| KR (1) | KR100535569B1 (en) |
| CN (1) | CN100483117C (en) |
| TW (1) | TWI311687B (en) |
Families Citing this family (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100642500B1 (en) * | 2005-03-02 | 2006-11-06 | (주)쎄미시스코 | Apparatus for edge defect and discolor testing of glass board |
| JP5418983B2 (en) * | 2010-02-26 | 2014-02-19 | 旭硝子株式会社 | Method and apparatus for inspecting cracks in rectangular plate |
| JPWO2012077683A1 (en) * | 2010-12-09 | 2014-05-19 | 旭硝子株式会社 | Glass ribbon defect measuring method and glass ribbon defect measuring system |
| US9488597B2 (en) * | 2011-11-30 | 2016-11-08 | Corning Incorporated | Apparatus and methods for determining surface compliance for a glass surface |
| CN105842885B (en) * | 2016-03-21 | 2018-11-27 | 凌云光技术集团有限责任公司 | A kind of liquid crystal display defect Hierarchical Location method and device |
| CN106353900B (en) * | 2016-08-30 | 2019-08-13 | 武汉精测电子集团股份有限公司 | A kind of picture signal generation method and device obtaining function with coordinate |
| CN106908449B (en) * | 2017-02-17 | 2019-09-06 | 福州东旭光电科技有限公司 | A method of finding the workshop section for generating defect by measurement liquid-crystalline glasses board defect depth |
| CN106996937B (en) * | 2017-06-15 | 2019-09-06 | 福州东旭光电科技有限公司 | Defect inspection method and device in a kind of glass substrate |
| CN110998298B (en) * | 2017-08-24 | 2023-01-06 | 日本电气硝子株式会社 | Method for manufacturing plate-shaped glass |
| KR101867015B1 (en) | 2017-10-24 | 2018-06-14 | (주) 엠브이텍 | Device and method for inspecting defect of glass and inspection system |
| US10498948B1 (en) | 2018-06-05 | 2019-12-03 | Applied Materials, Inc. | Methods and apparatus for absolute and relative depth measurements using camera focus distance |
| CN110020992A (en) * | 2018-09-18 | 2019-07-16 | 永康市巴九灵科技有限公司 | Edge quality Identification Platform |
| KR102199313B1 (en) * | 2018-12-19 | 2021-01-06 | (주) 인텍플러스 | Apparatus for inspecting cover glass |
| CN110618141B (en) * | 2019-09-17 | 2022-08-05 | 深圳新视智科技术有限公司 | Method, system and device for detecting glass defects |
| CN113484333B (en) * | 2021-09-08 | 2021-12-14 | 苏州高视半导体技术有限公司 | Method for distinguishing foreign matter defect of multi-layer structure screen, electronic device and storage medium |
-
2003
- 2003-05-31 KR KR10-2003-0035066A patent/KR100535569B1/en active IP Right Grant
- 2003-12-30 TW TW092137456A patent/TWI311687B/en not_active IP Right Cessation
-
2004
- 2004-01-09 CN CNB2004100020444A patent/CN100483117C/en not_active Expired - Lifetime
Also Published As
| Publication number | Publication date |
|---|---|
| TWI311687B (en) | 2009-07-01 |
| JP2004361384A (en) | 2004-12-24 |
| KR20040103143A (en) | 2004-12-08 |
| CN100483117C (en) | 2009-04-29 |
| CN1573319A (en) | 2005-02-02 |
| JP4571401B2 (en) | 2010-10-27 |
| KR100535569B1 (en) | 2005-12-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| TW200426497A (en) | Method for detecting the position of a defect in a glass substrate in a depth direction | |
| KR101711073B1 (en) | Micro cracks detection device and method of flexible touch screen panel using deep learning algorithm | |
| JP5007979B2 (en) | Defect inspection method and defect inspection apparatus | |
| JP5198535B2 (en) | Glass substrate cut surface inspection system | |
| JP2008139203A (en) | Apparatus and method for detecting defect, apparatus and method for processing information, and its program | |
| CN103123313A (en) | Hardness tester and hardness test method | |
| JPS63244753A (en) | Defect recognition and processing device | |
| JP2019215321A (en) | System and method for inspecting refractive power and thickness of ophthalmic lens immersed in solution | |
| JP2010071845A (en) | Inspection device | |
| JP2010181328A (en) | Device, program and method for inspecting surface of solar battery wafer | |
| TW202129269A (en) | Laser based inclusion detection system and methods | |
| TW584708B (en) | Method and apparatus for measuring a line width | |
| CN101604080B (en) | Detection method of lens substrate and manufacturing method of display device using same | |
| JP4571401B6 (en) | Depth position detection method for defects in glass substrate | |
| JP2009079915A (en) | Method and device for measuring micro-dimension | |
| CN111521617B (en) | Optical detection device, control method for optical detection device, and storage medium | |
| JP2019066222A (en) | Visual inspection device and visual inspection method | |
| KR101111383B1 (en) | Denting inspecting system having 3D surface measuring instrument | |
| KR101563799B1 (en) | Method for estimating relative depth using focus measure | |
| JP2000146787A (en) | Measuring method for restriction in tensile test | |
| JP2878763B2 (en) | Appearance inspection device | |
| KR100505219B1 (en) | The apparatus of inspection for flat panel display | |
| JP2009036696A (en) | Image inspecting apparatus | |
| CN106248689B (en) | Method for judging defects of composite film | |
| JP2007199089A (en) | Surface inspection apparatus |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| MK4A | Expiration of patent term of an invention patent |