201009324 六、發明說明：· 【發明所屬之技術領域】 本發明係關於一種對存在於具有透明性之玻璃板等之板 狀體之缺陷進行檢測的缺陷檢查系統及缺陷檢查方法。 【先前技術】 , 如今，玻璃板用於平板顯示器等之電子機器中，因此強 • 烈需要板厚較薄、氣泡等之缺陷極少或完全不存在之玻璃 板。 癱 雖可減少玻璃板中所含之氣泡或表面之劃痕等之缺陷， 但未必可完全除去該缺陷’因此，於檢查步驟中，需要除 去存在氣泡等之玻璃板之部分等之處理。因此，先前已提 出對存在於所製造之玻璃板等之具有透明性之板狀體中的 氣泡等之缺陷進行檢查之各種裝置。 例如’如圖6所示，於所搬送之玻璃板〇之一方之側設置 比玻璃板G之寬度長之線狀光源52，以特定之光強度朝玻 璃板G投光，利用設置於他方之側之線感測器型之相機54 _ 而拍攝透過玻璃板G之透射光之亮視場圖像，將該圖像發 送至處理單元56。利用處理單元56抽出亮視場圖像中所含 . 之成為暗部之區域作為缺陷區域。此時，為利用線感測器 . 型之相機54進行拍攝，朝玻璃基板G投光之線狀光源52係 使用沿玻璃板G之寬度方向延伸之細長之狭縫，而產生大 致平行光。藉此，可檢測所搬送之玻璃板G之缺陷。 另一方面，於下述專利文獻丨中已提出一種用以檢測透 明板狀體中之缺陷之缺陷檢測方法。於該檢測方法中，設 142304.doc 201009324 置以接近垂直之角度而對於柘壯栌夕 θ π敬狀體之面進行照明之照明 器、及以接近平行之角度進行照明之照明器，並對使用該 等照明器而獲得之圖像進行圖像處理，藉此，可檢測存在 於板狀體之表面及内部之缺陷。 [先行技術文獻] [專利文獻] [專利文獻1]日本專利特開2002-2141 58號公報 【發明内容】 [發明所欲解決之問題] 於上述專利文獻1中所揭示之缺陷檢測方法及上述圖6所 不之方法中，無法區分存在於玻璃板之面中之缺陷與存在 於内部之缺陷。進而，於如專利文獻丨般使用2個照明器， 並改變照明方法而進行拍攝之方法中，無法於線上自以固 足速度搬送之玻璃板中檢測出缺陷。因此，存在無法於玻 璃板之製造線上實施上述缺陷檢測方法之問題。 因此，為解決上述問題，本發明之目的在於提供一種缺 陷檢查系統及缺陷檢查方法，該缺陷檢查系統及缺陷檢查 方法可有效地用於玻璃板等之具有透明性之板狀體之製造 線等上，且可高效率地判別並檢測出板狀體中是否存在缺 陷’並且本發明之目的在於提供一種採用上述檢查方法之 板狀體之製造方法。 [解決問題之技術手段] 為解決上述問題，本發明提供一種缺陷檢查系統，其特 徵:在於：其係對存在於具有透明性之板狀體之缺陷進行檢 142304.doc 201009324 測者，其包含第i缺陷檢查裝置與處理裝置，上述第丨缺陷 檢查裝置包含··第w狀光源，其朝上述板狀體之面投 ::第1域，其㈣過上述板狀體之透射光聚光而拍攝 〜視場圖像，以及刀刃狀之光路遮蔽構件，其^置於上述 第1相機之透射光之光路中之位於上述第丨相機前面的位 置，上述處理裝置係自上述第丨相機所拍攝之亮視場圖像 中，將t匕亮視i易圖像之背景成分之信號值高之信號值作為 冑值，自冑視場圖像中搜纟亮部之區域，當搜t之結果為 ❿抽出亮部之區域時，使用該亮部之區域，判別於上述板狀 體上是否存在缺陷區域。 此時，較好的是上述第丨光源為線狀光源，於上述第1相 機之受光面之前面設置有用以使板狀體之像成像之成像透 鏡，以滿足第1條件與第2條件之方式，設定上述第i光 源上述成像透鏡以及上述第1相機，上述第1條件係當將 自上述第1光源經由上述成像透鏡而到達上述第丨相機之受 • 光面之上述第1光源之透射光之光束的擴散角之一半之角 度設為照明發光有效角，將自上述第i相機之受光面之位 置經由上述成像透鏡而到達上述第i光源之照射面之視野 範圍之視角的一半之角度設為視場角，將表示上述第1光 源之發光指向性之值設為α時，將表示上述第1光源之發光 指向性之值α、與上述視場角與上述照明發光有效角之比 率相乘所得之值大於2,上述第2條件係當於上述光路遮蔽 構件之上述透射光之經過位置，規定自形成於上述成像透 鏡之與光轴正交之面之上述受光面觀察時之視野範圍即模 142304.doc 201009324 糊圓時’藉由上述光路遮蔽構件遮斷上述模糊圓之部分之 面積成為上述模糊圓之面積的43〜57% »例如以滿足上述 第1條件及上述第2條件之方式，決定上述第1相機之光圈 值、上述第1相機與板狀體之間之距離、板狀體與上述第j 光源之間之距離、以及上述第1光源之發光幅度。 尤其好的是以第1相機之受光元件夾持板狀體而與第1光 源正對之方式配置第1相機。此處所謂以正對之方式配 置’係指以上述第1相機之受光元件夾持玻璃板G而位於第 1光源之最大光強度之方向的方式進行配置。 又，較好的是上述缺陷檢查系統除上述第1缺陷檢查裝 置外’還具有對於作為上述第1缺陷檢查裝置之檢查對象 之板狀體’接收自板狀體反射之照明光而檢查缺陷之第2 缺陷檢查裝置’上述第2缺陷檢查裝置包含：第2光源，其 將照明光照射至板狀體之面；以及第2相機，其將照射後 自板狀體之面出射之反射光聚光，拍攝亮視場反射圖像， 且其自板狀體觀察時係設置於與上述第2光源相同側；上 述處理裝置係自上述第2相機所拍攝之亮視場反射圖像 中，抽出由板狀體反射之圖像中之暗部之區域，於該暗部 之區域與上述亮部之區域接近並相對向之情形時，判別為 於板狀體上存在缺陷。 此時，較好的是上述處理裝置根據由板狀體反射之圖像 中之上述暗部之區域、與由板狀體之他方之面反射之圖像 中之暗部的位置偏差，求出板狀鱧之厚度方向之缺陷之位 置資訊。 142304.doc 201009324 再者’較好的是上述板狀體係被朝一個方向搬送而移 動，上述第1缺陷檢查裝置與上述第2缺陷檢查裝置中，上 述第1缺陷檢查裝置設置於上述第2缺陷檢查裝置之上游 侧。然而，上述第丨缺陷檢查裝置亦可設置於上述第2缺陷 檢查裝置之下游側。 此處’較好的是上述第丨缺陷檢查裝置與上述第2缺陷檢 查裝置以於搬送方向上相鄰之方式而設置。其原因在於： 當上述第1缺陷檢查裝置與上述第2缺陷檢查裝置之距離較 # 短時，與該距離較長之情形相比較，可於更短之時間内處 理缺陷之檢測與確定。 進而’本發明提供一種缺陷檢查方法，其特徵在於：其 係對存在於具有透明性之板狀體之缺陷進行檢測者，當自 第1線狀光源朝上述板狀體之面投光，將經過上述板狀體 之透射光聚光’利用第1相機拍攝亮視場圖像時，於上述 第1相機之透射光之光路中之位於上述第1相機前面的位置 • SX置刀刃狀之光路遮蔽構件並進行拍攝，自上述第1相機 所拍攝之亮視場圖像中，將比亮視場圖像之背景成分之信 號值高之信號值作為閾值，自亮視場圖像中搜索亮部之區 域，當搜索之結果為抽出亮部之區域時，使用該異常折射 所造成之亮部之區域，判別於上述板狀體上是否存在缺陷 區域。 再者’本發明對板狀體之厚度並無特別限制，例如可用 於氣造平板顯不(FPD ’ Flat Panel Display)中之用於包含 薄板類型之液晶顯示器（LCD，Liquid Crystal Display)用途 142304.doc 201009324[Technical Field] The present invention relates to a defect inspection system and a defect inspection method for detecting a defect existing in a plate-like body such as a glass plate having transparency. [Prior Art] Nowadays, glass plates are used in electronic devices such as flat panel displays, and therefore, there is a strong need for a glass plate having a thin plate thickness, little or no defects such as bubbles.缺陷 Although the defects such as the bubbles or the scratches on the surface of the glass sheet can be reduced, the defects are not necessarily completely removed. Therefore, in the inspection step, it is necessary to remove the portion of the glass sheet in which bubbles or the like are present. Therefore, various devices for inspecting defects such as bubbles existing in a transparent plate-like body such as a manufactured glass plate have been proposed. For example, as shown in FIG. 6, a linear light source 52 having a width longer than the width of the glass sheet G is provided on one side of the glass sheet to be conveyed, and light is projected toward the glass sheet G with a specific light intensity, and is disposed on the other side. The side line sensor type camera 54 _ captures a bright field of view image of the transmitted light transmitted through the glass sheet G, and sends the image to the processing unit 56. The processing unit 56 extracts a region which becomes a dark portion included in the bright field image as a defective region. At this time, in order to photograph the camera 54 using the line sensor type, the linear light source 52 projecting toward the glass substrate G uses an elongated slit extending in the width direction of the glass sheet G to generate substantially parallel light. Thereby, the defect of the conveyed glass plate G can be detected. On the other hand, a defect detecting method for detecting a defect in a transparent plate-like body has been proposed in the following patent document. In the detection method, 142304.doc 201009324 is provided with an illuminator that illuminates the surface of the θ π π π π π π π π π π π π π π π π π π π π The image obtained by using the illuminators is subjected to image processing, whereby defects existing on the surface and inside of the plate-like body can be detected. [PRIOR ART DOCUMENT] [Patent Document 1] Japanese Patent Laid-Open Publication No. 2002-2141 58 [Draft of the Invention] [Problems to be Solved by the Invention] The defect detecting method disclosed in the above Patent Document 1 and the above In the method of Fig. 6, it is impossible to distinguish the defects existing in the face of the glass sheet from the defects existing inside. Further, in the method of photographing using two illuminators as in the patent document and changing the illumination method, it is impossible to detect defects on the glass from the glass sheet conveyed at a fixed speed. Therefore, there is a problem that the above defect detecting method cannot be implemented on a manufacturing line of a glass plate. Therefore, in order to solve the above problems, an object of the present invention is to provide a defect inspection system and a defect inspection method which can be effectively used for a manufacturing line of a transparent plate-like body such as a glass plate. Further, it is possible to efficiently discriminate and detect whether or not a defect exists in the plate-shaped body, and an object of the present invention is to provide a method for producing a plate-shaped body using the above-described inspection method. [Means for Solving the Problems] In order to solve the above problems, the present invention provides a defect inspection system characterized in that it is a method for detecting a defect existing in a plate-like body having transparency, 142304.doc 201009324, which includes In the i-th defect inspection apparatus and the processing apparatus, the second defect inspection apparatus includes a w-shaped light source that projects toward a surface of the plate-shaped body: a first domain, and (4) a transmitted light concentrating through the plate-shaped body And capturing a field of view image, and a blade-shaped light path shielding member disposed at a position in front of the second camera in the light path of the first camera, wherein the processing device is from the camera In the bright field of view image, the signal value with the high signal value of the background component of the i-image is taken as the 胄 value, and the area of the bright portion is searched from the field of view image. When the region of the bright portion is extracted, the region of the bright portion is used to determine whether or not a defective region exists on the plate-like body. In this case, it is preferable that the second light source is a linear light source, and an imaging lens for imaging an image of the plate-shaped body is provided on a surface of the light receiving surface of the first camera to satisfy the first condition and the second condition. In the above aspect, the first i-th light source and the first camera are set, and the first condition is that the first light source is transmitted from the first light source to the light receiving surface of the second camera through the imaging lens. The angle of one half of the spread angle of the light beam is set as the effective angle of illumination illumination, and the angle from the position of the light receiving surface of the ith camera to the angle of view of the field of view of the illumination surface of the ith light source is transmitted through the imaging lens. When the value of the light-emitting directivity of the first light source is α, the value α indicating the light-emitting directivity of the first light source and the ratio of the angle of view to the effective angle of the illumination light are set to be the angle of view. The value obtained by the multiplication is greater than 2, and the second condition is defined as a position at which the transmitted light passes through the optical path shielding member, and is formed from a surface orthogonal to the optical axis of the imaging lens. The area of the field of view when the light surface is observed is modulo 142304.doc 201009324. When the paste is rounded, the area of the portion of the blur circle that is blocked by the light path shielding member is 43 to 57% of the area of the blur circle. In the first condition and the second condition, the aperture value of the first camera, the distance between the first camera and the plate body, the distance between the plate-shaped body and the j-th light source, and the first light source are determined. The range of illumination. In particular, the first camera is disposed such that the light receiving element of the first camera sandwiches the plate-shaped body and faces the first light source. Here, the arrangement is performed in such a manner that the glass plate G is sandwiched between the light receiving elements of the first camera and placed in the direction of the maximum light intensity of the first light source. In addition, it is preferable that the defect inspection system includes, in addition to the first defect inspection device, a defect in which the plate-shaped body that is the inspection target of the first defect inspection device receives the illumination light reflected from the plate-shaped body. The second defect inspection device 'the second defect inspection device includes: a second light source that illuminates the surface of the plate-shaped body; and a second camera that condenses the reflected light emitted from the surface of the plate-shaped body after the irradiation Light, capturing a bright field of view reflection image, which is disposed on the same side as the second light source when viewed from the plate body; and the processing device extracts from the bright field of view image captured by the second camera The region of the dark portion in the image reflected by the plate-like body is judged to have a defect on the plate-like body when the region of the dark portion is close to and opposed to the region of the bright portion. In this case, it is preferable that the processing device obtains a plate shape based on a positional deviation between a region of the dark portion in the image reflected by the plate-like body and a dark portion in an image reflected by the other surface of the plate-shaped body. Location information of defects in the thickness direction of the crucible. 142304.doc 201009324 Further, it is preferable that the plate-like system is moved and moved in one direction, and in the first defect inspection device and the second defect inspection device, the first defect inspection device is provided in the second defect Check the upstream side of the unit. However, the above-described second defect inspection device may be provided on the downstream side of the second defect inspection device. Here, it is preferable that the second defect inspection device and the second defect inspection device are disposed adjacent to each other in the transport direction. The reason for this is that when the distance between the first defect inspection device and the second defect inspection device is shorter than #, the detection and determination of the defect can be handled in a shorter time than when the distance is long. Further, the present invention provides a defect inspection method which is characterized in that a defect existing in a plate-like body having transparency is emitted from a first linear light source toward a surface of the plate-shaped body. When the bright field image is captured by the first camera by the first camera, the position of the optical path of the transmitted light of the first camera located at the front of the first camera • SX is a blade-shaped optical path The shielding member is photographed, and a signal value higher than a signal value of a background component of the bright field image is used as a threshold value from the bright field image captured by the first camera, and the brightness is searched for from the bright field image. In the region of the portion, when the search result is the region where the bright portion is extracted, the region of the bright portion caused by the abnormal refraction is used to determine whether or not the defective region exists on the plate-like body. Further, the present invention has no particular limitation on the thickness of the plate-like body, and can be used, for example, in a liquid crystal display (LCD) for use in a FPD 'Flat Panel Display. .doc 201009324
Plasma 的厚度為0_1〜0·7 mm者、或電漿顯示面板（pDp，Plasma thickness is 0_1~0·7 mm, or plasma display panel (pDp,
Dispiay Panel)等中之厚度為丨mm以上者、例如厚度為i 8 mm者或厚度為2.8爪爪者，進而，可用於製造建材用途之 板玻璃及採取板狀體之中間形態之透明之物品。 [發明之效果] 本發明之缺陷檢查系統及缺陷檢查方法中，於第丨相機 之透射光之光路中之第1相機前面的位置設置刀刃狀之光 路遮蔽構件，自第1相機所拍攝之亮視場圖像中，將比亮 視場圖像之背景成分之信號值高之信號值作為閾值，自亮 視場圖像中抽出亮部之區域，將該抽出之亮部之區域用於 檢測存在於板狀體上之缺陷區域，藉此可高效地判別並檢 測於板狀體上是否存在缺陷。 尤其，以滿足第1條件與第2條件之方式設定第i相機之 光學系統，藉此，可有效地使亮視場圖像產生上述亮部， 上述第1條件係指將表示第1光源之發光指向性之值α(0以 上且1以下之數值）、與視場角與照明發光有效角之比率相 乘所得的值大於2的條件，上述第2條件係以光路遮蔽構件 遮蔽上述模糊圓之部分之面積成為上述模糊圓之面積的 43〜5 7%的方式進行設定之條件。尤其，將用於測量之第工 光源與第1相機設置成夾持玻璃板<3而正對，藉此，可更有 效地進行缺陷檢查。 又，本發明之缺陷檢查系統包含：上述第2光源，其將 照明光照射至板狀體之面；以及上述第2相機，其將照射 後由板狀體之面反射而得之反射光聚光，拍攝亮視場反射 142304.doc 201009324 圖像，且其自板狀體觀察時係設置於與上述第2光源相同 側；上述處理裝置係自上述第2相機所拍攝之亮視場反射 圖像中，抽出由板狀體之面反射之圖像中之暗部之區域， 於該暗部之區域與上述亮部之區域接近並相向之情形時， 可根據缺陷之實像與缺陷之鏡像之位置偏差，求出位於板 狀體上之缺陷之厚度方向之位置資訊。 【實施方式】 以下，根據隨附圖式所示之較佳實施例，對本發明之缺 陷檢查系統及缺陷檢查方法加以詳細說明。 圖1所示之缺陷檢查系統1〇係包含第！缺陷檢查裝置12、 第2缺陷檢查裝置14以及處理裝置16而構成。 第1缺陷檢查裝置12與第2缺陷檢查裝置14係沿玻璃板G 之搬送路徑而自上游側依序設置。處理裝置丨6係對由第i 缺陷檢查裝置12與第2缺陷檢查裝置丨4獲得之圖像進行處 理而進行缺陷檢測之裝置。 玻璃板G係自熔融爐中取出且達到特定厚度之長條狀之 板材’其於設置在搬送路徑中之複數個驅動輥18上被搬 送。 第1缺陷檢查裝置12係位於缺陷檢查系統10之搬送侧最 上游，並檢查玻璃板G之缺陷之裝置。 具體而言，第1缺陷檢查裝置12包含：第1線狀光源2〇， 其自驅動輥 18之側（下側）朝玻璃板g之面投光；第1相機 22，其將通過玻璃板G之透射光聚光而拍攝亮視場圖像； 以及刀刀狀之光路遮蔽構件24，其設置於第1線狀光源2〇 142304.doc 201009324 之透射光之光路中之第1相機22前面的位置。 第1線狀光源20係射出大致平行光之LEDCLight Emitting Diode，發光二極體）光源，第！線狀光源2〇之出射口沿玻 璃板G之寬度方向（圖1中之與紙面垂直之方向）延伸成線 狀。第1線狀光源20之出射口設置於與玻璃板〇之面相距例 如100-900 mm之位置’光源之沿搬送方向之寬度L(參照圖 2)例如設定為1〜20 mm。再者’考慮到無需較高之位置精 度，較好的是將第1線狀光源20與玻璃板g之面隔開地設 置。LED光源中之光之種類並無特別限制，較佳使用白 色，但亦可為紅色、藍色、綠色等。具體而言，lED光源 包含發光之發光源（未圖示）、使所發出之光成為大致平行 光之菲！耳透鏡（未圖示）、使光強度大致均一之擴散板（未 圖示）以及使光之出射範圍變窄之狹縫板（未圖示）。藉此， 第1線狀光源20發出具有大致均一之光強度之大致平行 光。再者’即便如上所述般使用菲涅耳透鏡、擴散板以及 狹縫板’亦未必可使光強度均一且使光成為平行光，且光 強度具有指向特性，因而光會擴散。考慮此時之光強度之 指向特性’將表示指向特性之值設為α。 第1相機22係線感測器型相機，其設置於夾持玻璃板G而 與第1線狀光源20相對向之位置，且利用受光面直接讀取 透過玻璃板G之透射光。於與圖〗中之紙面垂直之方向上設 置有複數台第1相機22，其拍攝搬送方向之相同位置，而 且’複數台相機係以玻璃板G之寬度方向上之視野範圍彼 此部分性重疊之方式而設定，且以於玻璃板G之檢查部分 142304.doc •10· 201009324 中無非檢查區域之方式而配置。 第1相機22係設置為受光面位於第i相機22之成像透鏡 23(參照圖2)之焦點對準之位置，即與玻璃板G之面相距例 如200〜4〇〇 mm之位置。第1相機22令，包含具備成像透鏡 23之光學系統及未圖示之調整開口之光圈。利用第t相機 22所獲得之圖像資料每當被線狀地讀取時，被逐次發送至 處理裝置16。 光路遮蔽構件24係刀刃狀之構件，其於來自玻璃板G之 透射光之光路中之第1相機22前面的位置遮斷光路之一部 分。該光路遮蔽構件24之光路中之前端部分係以成為刃之 方式而經銳化。光路遮蔽構件24設置於第}相機22之光學 系統（成像透鏡23)前面之位置，例如與該光學系統相距 1 5 mm之位置。於保持光路遮蔽構件μ之部分中，設置 有可使光路遮蔽構件24以橫斷光路之方式而沿χ方向移動 之機構。此時，當於光路遮蔽構件24之透射光之通過位置 規定模糊圓時，該模糊圓表示自由與第i相機22中之成像 透鏡23之光軸正交之面所形成之受光面觀察的視野範圍， 藉由光路遮蔽構件24遮斷模糊圓之部分之面積相當於模糊 圓之面積的43〜57%，較好的是相當於大致5〇%。於小於 43〇/。之情形時，於亮視場圖像中難以產生後述之亮部，若 超過57°/。，則易成為暗視場圖像。 此種遮斷光路之範圍可藉由調整例如光遮蔽構件24與玻 璃板G之間之距離、及第i相機22之光圏值而實現。其原因 在於：藉由將遮斷光路之範圍設定為該範圍，如下所述， 142304.doc 11 201009324 與由存在於亮視場圖像内之氣泡等之缺陷所造成的暗部之 區域接近而高效地形成亮部之區域。 此時’較好的是當將表示第1線狀光源20之發光指向性 之值設為α時，相對於圖2所示之照明發光有效角Θ之視場 角沴之比率與值α相乘所得之值大於2。所謂照明發光有效 角β，係指自第1線狀光源20經由成像透鏡23而到達第1相 機22之受光元件之受光面的第1線狀光源2〇之透射光之光 束的擴散角之一半的角度。視場角彡係自第1相機22之受光 元件之受光面之位置經由成像透鏡23(使用透鏡之有效口 徑d)而到達第1線狀光源2〇之照射面的視野範圍之視角的 一半之角度。 此處，所謂表示第1線狀光源20之發光指向性之值α，如 圖3所示’係指當於橫袖上將與光源之照射面正交之方向 設為方位角0度來取方位角度，於縱軸上將最大光強度之 值設為1時的相對光強度之平均值。 進而，較好的是如圖2所示，以第1相機22之受光元件夾 持玻璃板G而與第1線狀光源20正對之方式而配置第1相機 22。再者，圖2中之d係成像透鏡23之有效口徑，且以f/F(f 為焦點距離’ F為F值（光圈值））而表示。如此，本發明中所 使用之照明發光有效角Θ及視場角0係如圖2所示，根據各 裝置之設定及配置而以幾何學之方式得以確定。 藉由此種第1缺陷檢查裝置12，可容易地檢測出存在於 玻璃板G中.之微小之氣泡等之缺陷。 第2缺陷檢查裝置14包含第2光源28與第2相機30。第2缺 142304.doc •12· 201009324 陷檢查裝置14係如下之裝置，其對於作為第！缺陷檢查裝 置之檢查對象而經檢查之玻璃板G，自一方之側對玻璃板 照明，此時，利用第2相機30接收玻璃板G之表面與背面所 反射之照明光，從而檢查缺陷。 將上述第1缺陷檢查裝置12之檢測結果與該第2缺陷檢查 裝置14之檢測結果加以組合而综合地評價，藉此，可更準 確地確定存在於玻璃板G中之缺陷。 第2光源2 8係朝玻璃板G之面射出大致平行光之[e D光 源’其使光自相對於玻璃板G之面而傾斜之方向入射。第2 光源28沿與圖1之紙面垂直之方向延伸。第2光源28中所使 用之LED光源中之光之種類並無特別限制，較佳使用白 色，但亦可為紅色、藍色、綠色等。具體而言，LED光源 包含發光之發光源（未圖示）、使所發出之光成為大致平行 光之菲 >圼耳透鏡（未圖示）、使光強度大致均一之擴散板（未 圖示）以及使光之出射範圍變窄之狹縫板（未圖示藉此， 第2線狀光源28發出具有大致均一之光強度之大致平行 光。 第2相機30係將自玻璃G之表面出射之反射光聚光，並拍 攝亮視場反射圖像之線感測器型相機。第2相機30自玻璃 板G觀察時係設置於與第2光源28相同側。 第2相機30所拍攝之圖像係經過第2光源28照明之後由玻 璃板G之背面反射之圖像，且係存在於玻璃板g上之缺陷 之區域成為暗部的圖像。於該圖像中包含：缺陷之實像， 其係自相對於玻璃板G之面而傾斜之方向入射至破璃板g 142304.doc 13 201009324 之表面，由玻璃板G之背面反射後’缺陷之區域經過該反 射光之光路而形成；以及缺陷之鏡像，其係自相對於玻璃 板G之表面而傾斜之方向入射至玻璃板〇之表面的入射光 經過玻璃板G内之存在於光路中的缺陷之區域後，由玻璃 板G之背面反射而形成。利用第2相機3 〇獲得之圖像資料每 當被線狀地讀取時’被逐次發送至處理裝置丨6。 處理裝置16亦為如下之裝置，其使用自第丨缺陷檢查裝 置12及第2缺陷檢查裝置14發送之圖像資料，檢測玻璃板g 之缺陷，識別其種類，並確定缺陷之於玻璃板G之厚度方 向上之位置。於處理裝置16上連接有顯示器32，於顯示器 32上，畫面顯示利用第1缺陷檢查裝置12及第2缺陷檢查裝 置14獲得之圖像、缺陷之檢測結果、識別結果、或者缺陷 之位置確定結果。 如上所述，利用第1缺陷檢查裝置12獲得之圖像為亮視 場圖像，玻璃板G之缺陷由於缺陷區域之漫反射而於圖像 中顯現為暗部。又’如上所述，於第1相機22之光學系統 之正前方設置有將光路之一部分遮斷之刀刃狀的光路遮蔽 構件24，藉此利用該光路遮蔽構件24，以與圖像中之暗部 之區域接近並相向或者接觸並相對之方式而形成亮部之區 域。該亮部係因缺陷之部分之折射異常而產生，其亮度比 亮視場圖像之背景部分高。 圖4 A係光路遮蔽構件24處於光路中時之缺陷圖像之一例 的模式圖。如圖4A所示，形成有與暗部之區域接近並相向 之亮部。與此相對，圖4B係光路遮蔽構件24未存在於光路 142304.doc -14 - 201009324 中（未將光遮斷）時之缺陷圖像之一例的模式圖。如圖4b所 示未开/成與暗部之區域接近並相向之亮部。一般認為於 儿視％圖像中’如圖4A所示，以與暗部之區域接近並相向 之方式形成亮部’其原因在於因存在於玻璃板以表面或 内部之氣泡或異物等，玻璃板G之表面成為凹凸形狀，玻 璃板G之折射異常稍微變化。實際上，並不會因於玻璃板 G之面上產生之劃痕或所附著之異物而形成如圖4A所示之 亮部。 ^ 因此’於處理裝置16中，在亮視場圖像中設定比亮視場 圖像中之背景部分之圖像資料之值高的閾值，將該閾值以 上之區域抽出作為亮部之區域。再者，自第1相機22發送 而來之圖像資料係經過檢查位置後利用線感測器型相機讀 取之一維之圖像資料，因此於處理裝置16中儲存複數行 (例如500行）之圖像資料而獲得一定之面區域之圖像之後， 開始抽出上述亮部之區域。當檢測存在於玻璃板G上之缺 • 陷區域時，該抽出之亮部之區域之位置資訊係以下述方式 使用。 利用第2缺陷檢查裝置14獲得後發送至處理裝置丨6之圖 像資料係亮視場反射圖像之資料，且係缺陷之部分成為暗 部之區域之圖像。如上所述，缺陷之實像與鏡像顯現為暗 部。根據玻璃板G之缺陷所存在之玻璃板g之厚度方向的 位置’缺陷之實像與鏡像產生位置偏差。例如於缺陷位於 玻璃板G之背面附近之情形時，實像與鏡像之位置偏差量 變小，於缺陷位於表面附近之情形時，實像與鏡像之位置 142304.doc -15- 201009324 偏差量變大。 因此’處理裝置16使用上述已抽出之亮部之區域之資 訊’確定缺陷之玻璃板G之寬度方向之位置。進而，使用 該寬度方向之位置，根據於利用第丨缺陷檢查裝置12獲得 之圖像與利用第2缺陷檢查裝置丨4獲得之圖像之間出現玻 璃板G之相同部位之圖像的時間偏差量，對利用第2相機3〇 獲得之缺陷之實像及鏡像之暗部之區域進行檢測。因第工 缺陷檢查裝置12之測量位置與第2缺陷檢查裝置14之測量 位置於搬送方向上之相隔距離及玻璃〇之搬送速度為已 知，故可根據該等資訊而求出上述時間偏差量。 為抽出暗部之區域，使用預先經設定之閾值而進行。 其次，求出上述實像之中心位置與鏡像之中心位置之位 置偏差量’根據該位置偏差量’算出缺陷之於玻璃板〇之 厚度方向上之位置。 ❹ 又，處理裝置16使用利用第2相機3〇獲得之缺陷之實像 而求出暗部之區域的大小，根據該區域之大小而推定缺陷 之大小。再者，處理襞置16係與基於上述亮部之區域之資 訊的處理相獨立，不使用上述亮部之區域之資訊，而使用 預先經设定之閾值來抽出暗部之區域。 處理裝置16使用已抽出之暗部之區域與亮部之區域之資 訊’判別亮部之區域與暗部之區域是否相接近並相向，且 使用所算出之存在於破璃板G中之缺陷之厚度方向之位 置、及根據已抽出之暗部之區域而求出之缺陷之大小及缺 陷之特徵量，確定缺陷之種類。較好的是，根據亮視場透 142304.doc -16- 201009324 射圖像中之亮部及暗部之區域與亮視場反射圖像中之暗部 之區域’識別缺陷之形狀，藉此，區分為例如由氣泡引起 之缺陷由異物引起之缺陷或者劃痕等而推定缺陷之種 類。 存在於玻璃板G之面上之異物或劃痕等之缺陷，於自第工 缺檢查裝置12獲得之亮視場圖像中不會形成亮部。 再者，於帛1缺陷檢查裝置12中，為有效地使亮視場圖 像中之與*陷之暗部之區域接近並相向的亮部出，見，較好 的是滿足以下條件。 亦即，於圖2所示之各部分之配置中，以滿足將值α與相 對於照明發光有效角9之視場角0之比率相乘所得之值大於 2的第1條件之方式，設定第1相機22與玻璃板G之間之距 離、玻璃板G與第1線狀光源2〇之間之距離、以及第t線狀 光源之照射寬度L。 又，以滿足第2條件之方式設定第i相機22之光圈值及第 1相機22與玻璃板(3之間之距離，上述第2條件係當規定光 路遮蔽構件24之配置位置之、由與成像透鏡23之光轴正交 之面所形成之規定自受光面觀察的視野範圍之邊緣的模糊 圓時’藉由光路遮蔽構件24遮斷模糊圓之部分之面積成為 模糊圓之面積之43〜57%。 進而’以第1相機22之受光元件夾持玻璃板G而位於第} 線狀光源20之最大光強度之方向的方式配置第1相機22。 以滿足該3個條件之方式配置各裝置，藉此，可使亮視 場圖像有效地產生上述亮部。 142304.doc 17 201009324 又藉由減小F值，拍攝中之被拍攝體焦點深度變淺， 生易發生圖像之影像模糊之不良。因此，為高效地抽出 亮部’較好的是F5.6〜FU，且較好的是於第1相機2〇之發 光部分34之寬度L為1〜20酿之範圍内，將值α與上述比率 相乘所得之值大於2。 於圖5Α及5Β所示之表中，表示根據第1相機20之發光部 分34之寬度L為L=1 mm、3職、4韻、5軸、7顏而規 定之照明發光有效角θ的數值、及根以值而規定之視場角 彡之值於各表之對應之欄中，表示卢/Θ乘以第〗光源2〇2α 之值所得之值。於以圖5Α&5Β所示之表中之粗框所包圍 之範圍内，g值α大致為1時，於圖4Α所示之α·#/θ之值大 於2之條件下，已確認有效地出現了亮部。藉此，當第t線 狀光源20之值α小於丨（考慮發光指向性）時，可以說於α·~θ 大於2之條件下會有效地出現亮部。 再者，將自玻璃板G至第1相機22之成像透鏡23之表面為 止之距離設為380 mm。自第！線狀光源2〇之照射面至玻璃 板G之測量位置為止之照明WD(w〇rking Distance，工作距 離）於圖5A中為200 mm，於圖5B_為4〇〇 mm。此時之藉由 光路遮蔽構件24遮斷模糊圓之部分之面積成為模糊圓之面 積的50%。 此種缺陷檢查系統1 〇於如下方面有效：可將平行光傾斜 地照射至玻璃板G，並高精度地抽出與紋影拍攝法類似之 先則之自動缺陷檢查中漏查的缺陷，而且可區分無法修復 之玻璃板G之缺陷。藉此，可避開無法修復之缺陷部分而 142304.doc -18- 201009324 將玻璃板G切割成特定之尺寸。 以上，對本發明之缺陷檢查系統及缺陷檢查方法進行了 詳細說明，但本發明並不限定於上述實施形態或實施例， 當然亦可於不脫離本發明之主旨之範圍内進行各種改良或 變更。 又’已參照特定之實施形態而對本發明進行了詳細說 明’但業者理應明白可不脫離本發明之精神與範圍而添加 各種變更或修正。本申請案係基於2〇〇8年8月25日申請之 •日本專利申請（日本專利特願2008-215091)者，其内容於此 作為參照而併入本申請案。 【圖式簡單說明】 圖1係對本發明之缺陷檢查系統及缺陷檢查方法之一實 施形態之概略構成進行說明的圖； 圖2係對本發明之缺陷檢查系統中之照明發光有效角Θ與 視場角4進行說明之圖； 圖3係對本發明之缺陷檢查系統中所使用之α進行說明之 is 園， 圖4A係模式性表示利用本發明之第1缺陷檢查裝置獲得 之圖像之一例的圖； 圖4B係模式性表示由通常之透射光所形成之亮視場圖像 獲得之圖像之一例的圖； 圖5 A係對用以於圖1所示之缺陷系統之第1缺陷檢查裝置 中產生亮部之較佳範圍進行說明的圖； 圖5B係對用以於圖1所示之缺陷系統之第1缺陷檢查裝置 142304.doc _ 19- 201009324 中產生亮部之較佳範圍進行說明的圖；及 圖6係對先前之缺陷檢查裝置之概略構成進行說明之 圖。 【主要元件符號說明】 10 缺陷檢查系統 12 第1缺陷檢查裝置 14 第2缺陷檢查裝置 16 處理裝置 18 驅動輥 20 第1線狀光源 22 第1相機 23 成像透鏡 24 光路遮蔽構件 28 第2光源 30 第2相機 32 顯示器 34 發光部分 142304.doc -20-Dispiay Panel), etc., such as those having a thickness of 丨mm or more, for example, a thickness of i 8 mm or a thickness of 2.8 claws, and further, can be used for manufacturing sheet glass for building materials and transparent articles in the form of a plate-like body. . [Effect of the Invention] In the defect inspection system and the defect inspection method of the present invention, a blade-shaped light path shielding member is disposed at a position in front of the first camera in the light path of the transmitted light of the second camera, and the light is captured from the first camera. In the field of view image, a signal value higher than a signal value of a background component of the bright field image is used as a threshold value, and a bright portion is extracted from the bright field image, and the extracted bright portion is used for detection. The defect region existing on the plate-shaped body can thereby efficiently discriminate and detect whether or not there is a defect on the plate-like body. In particular, the optical system of the i-th camera is set so as to satisfy the first condition and the second condition, whereby the bright field image can be effectively generated by the bright portion, and the first condition means that the first light source is indicated. The value of the illuminance directivity α (the value of 0 or more and 1 or less) and the value obtained by multiplying the ratio of the viewing angle to the effective angle of the illumination illuminance are greater than 2, and the second condition is to shield the blurring circle by the optical path shielding member. The condition is set such that the area of the portion is 43 to 5 7% of the area of the blurred circle. In particular, the first light source for measurement and the first camera are disposed so as to sandwich the glass sheet <3, whereby the defect inspection can be performed more effectively. Further, the defect inspection system of the present invention includes: the second light source that illuminates the surface of the plate-shaped body; and the second camera that reflects the reflected light from the surface of the plate-shaped body after the irradiation Light, photographing a bright field of view 142304.doc 201009324 image, which is disposed on the same side as the second light source when viewed from the plate; the processing device is a bright field reflection image taken from the second camera In the image, the region of the dark portion in the image reflected by the surface of the plate-like body is extracted, and when the region of the dark portion is close to and facing the region of the bright portion, the position of the image of the defect and the mirror of the defect may be deviated. Find the position information of the thickness direction of the defect on the plate-shaped body. [Embodiment] Hereinafter, a defect inspection system and a defect inspection method of the present invention will be described in detail based on preferred embodiments shown in the accompanying drawings. The defect inspection system 1 shown in Figure 1 contains the first! The defect inspection device 12, the second defect inspection device 14, and the processing device 16 are configured. The first defect inspection device 12 and the second defect inspection device 14 are sequentially disposed from the upstream side along the transport path of the glass sheet G. The processing device 6 is a device that performs processing for detecting defects by the images obtained by the i-th defect inspection device 12 and the second defect inspection device 丨4. The glass sheet G is a long sheet of material which has been taken out from the melting furnace and has reached a specific thickness, and is conveyed on a plurality of driving rolls 18 provided in the conveying path. The first defect inspection device 12 is a device that is located on the most upstream side of the conveyance side of the defect inspection system 10 and inspects the defects of the glass sheet G. Specifically, the first defect inspection device 12 includes a first linear light source 2 that projects light from the side (lower side) of the driving roller 18 toward the surface of the glass plate g; the first camera 22 passes through the glass plate. G transmitted light is collected to capture a bright field image; and a knife-shaped light path shielding member 24 is disposed in front of the first camera 22 in the transmitted light path of the first linear light source 2 142304.doc 201009324 s position. The first linear light source 20 is an LEDCLight Emitting Diode that emits substantially parallel light, a light-emitting diode) light source, the first! The exit opening of the linear light source 2 延伸 extends in a line shape along the width direction of the glass sheet G (the direction perpendicular to the paper surface in Fig. 1). The exit port of the first linear light source 20 is disposed at a distance of, for example, 100 to 900 mm from the surface of the glass plate. The width L (see Fig. 2) of the light source in the transport direction is set to, for example, 1 to 20 mm. Further, in view of the fact that it is not necessary to have a high positional accuracy, it is preferable to provide the first linear light source 20 at a distance from the surface of the glass sheet g. The type of light in the LED light source is not particularly limited, and white is preferably used, but may be red, blue, green, or the like. Specifically, the lED light source includes a light source (not shown) that emits light, and the emitted light is made into a substantially parallel light. An ear lens (not shown), a diffusing plate (not shown) that makes the light intensity substantially uniform, and a slit plate (not shown) that narrows the light emission range. Thereby, the first linear light source 20 emits substantially parallel light having a substantially uniform light intensity. Further, even if the Fresnel lens, the diffusion plate, and the slit plate are used as described above, the light intensity is uniform and the light is made parallel light, and the light intensity has a directivity characteristic, so that light is diffused. Considering the directivity characteristic of the light intensity at this time, the value indicating the directivity characteristic is set to α. The first camera 22 is a line sensor type camera which is disposed at a position facing the first linear light source 20 by sandwiching the glass sheet G, and directly reads the transmitted light transmitted through the glass sheet G by the light receiving surface. A plurality of first cameras 22 are disposed in a direction perpendicular to the paper surface in the drawing, which captures the same position in the transport direction, and the plurality of cameras partially overlap each other in the width direction of the glass plate G in the width direction. It is set as a mode, and it is arrange|positioned in the non-inspection area in the inspection part 142304.doc.10* 201009324 of the glass plate G. The first camera 22 is disposed such that the light receiving surface is located at the in-focus position of the imaging lens 23 (see Fig. 2) of the i-th camera 22, i.e., at a position spaced apart from the surface of the glass sheet G by, for example, 200 to 4 mm. The first camera 22 includes an optical system including an imaging lens 23 and an aperture of an adjustment opening (not shown). The image data obtained by the t-th camera 22 is sequentially transmitted to the processing device 16 every time it is read linearly. The optical path shielding member 24 is a blade-shaped member that blocks a part of the optical path from a position in front of the first camera 22 from the optical path of the glass plate G. The front end portion of the optical path of the optical path shielding member 24 is sharpened in such a manner as to become a blade. The light path shielding member 24 is disposed at a position in front of the optical system (imaging lens 23) of the camera 22, for example, at a position of 15 mm from the optical system. A mechanism for moving the optical path shielding member 24 in the x-direction so as to traverse the optical path is provided in a portion where the optical path shielding member μ is held. At this time, when the blurring circle is defined by the passing position of the transmitted light of the optical path shielding member 24, the blurred circle indicates the field of view observed by the light receiving surface formed by the surface orthogonal to the optical axis of the imaging lens 23 in the i-th camera 22. In the range, the area of the portion where the blurring circle is blocked by the light path shielding member 24 corresponds to 43 to 57% of the area of the blurred circle, and preferably corresponds to approximately 5%. Less than 43〇/. In the case of the bright field image, it is difficult to generate a bright portion to be described later, if it exceeds 57°/. , it is easy to become a dark field image. The range of such a blocking optical path can be achieved by adjusting, for example, the distance between the light shielding member 24 and the glass plate G, and the pupil value of the i-th camera 22. The reason for this is that by setting the range of the blocking optical path to the range, as described below, 142304.doc 11 201009324 is highly efficient in approaching the dark portion caused by defects such as bubbles existing in the bright field image. The area where the bright part is formed. At this time, it is preferable that when the value indicating the illuminance directivity of the first linear light source 20 is α, the ratio of the viewing angle 沴 of the effective illuminating angle 照明 shown in FIG. 2 is equal to the value α. The value obtained by multiplication is greater than 2. The illumination light-emission effective angle β is one of the diffusion angles of the light beam of the first linear light source 2 that reaches the light-receiving surface of the light-receiving element of the first camera 22 via the imaging lens 23 through the imaging lens 23. Angle. The angle of view is from the position of the light receiving surface of the light receiving element of the first camera 22 to the half of the field of view of the irradiation surface of the first linear light source 2 through the imaging lens 23 (using the effective aperture d of the lens). angle. Here, the value α indicating the illuminance directivity of the first linear light source 20 is as shown in FIG. 3, which means that the direction orthogonal to the irradiation surface of the light source is set to an azimuth angle of 0 degrees on the horizontal sleeve. The azimuth angle is the average of the relative light intensities when the value of the maximum light intensity is set to 1 on the vertical axis. Furthermore, as shown in Fig. 2, the first camera 22 is disposed such that the light receiving element of the first camera 22 sandwiches the glass sheet G and faces the first linear light source 20. Further, d in Fig. 2 is an effective aperture of the imaging lens 23, and is expressed by f/F (f is a focal length 'F is an F value (aperture value)). Thus, the illumination illumination effective angle Θ and the angle of view 0 used in the present invention are determined as shown in Fig. 2 in a geometric manner in accordance with the setting and arrangement of the respective devices. With such a first defect inspection device 12, it is possible to easily detect defects such as minute bubbles existing in the glass sheet G. The second defect inspection device 14 includes a second light source 28 and a second camera 30. The second absence 142304.doc •12· 201009324 The trap inspection device 14 is the following device, which is the first! The glass plate G to be inspected by the defect inspection device is illuminated from the side of the glass plate G. At this time, the second camera 30 receives the illumination light reflected from the front and back surfaces of the glass plate G to inspect the defect. The combination of the detection result of the first defect inspection device 12 and the detection result of the second defect inspection device 14 is comprehensively evaluated, whereby the defects existing in the glass sheet G can be more accurately determined. The second light source 28 emits an [e D light source] which emits substantially parallel light toward the surface of the glass sheet G, and causes light to enter from a direction inclined with respect to the surface of the glass sheet G. The second light source 28 extends in a direction perpendicular to the plane of the paper of Fig. 1. The type of light used in the LED light source used in the second light source 28 is not particularly limited, and white is preferably used, but may be red, blue, green or the like. Specifically, the LED light source includes a light-emitting source (not shown), a phenanthrene lens (not shown) that causes the emitted light to be substantially parallel light, and a diffusion plate that substantially uniformizes light intensity (not shown). And a slit plate for narrowing the light emission range (not shown, the second linear light source 28 emits substantially parallel light having substantially uniform light intensity. The second camera 30 is derived from the surface of the glass G A line sensor type camera that emits light from the reflected light and emits a bright field of view reflection image. The second camera 30 is disposed on the same side as the second light source 28 when viewed from the glass plate G. The second camera 30 is photographed. The image is an image reflected by the back surface of the glass sheet G after being illuminated by the second light source 28, and the region of the defect existing on the glass sheet g becomes an image of the dark portion. The image includes: a real image of the defect , which is incident on the surface of the glass plate g 142304.doc 13 201009324 from the direction inclined with respect to the surface of the glass sheet G, and is reflected by the back surface of the glass sheet G, and the region of the defect is formed by the optical path of the reflected light; And the mirror of the defect, which is relative to The incident light incident on the surface of the glass plate 倾斜 in the direction in which the glass plate G is inclined passes through the region of the defect existing in the optical path in the glass plate G, and is formed by being reflected by the back surface of the glass plate G. The second camera 3 is used. The image data obtained by 〇 is transmitted to the processing device 逐6 one by one every time it is read linearly. The processing device 16 is also a device that is used from the 丨 defect inspection device 12 and the second defect inspection device 14 The image data is sent, the defect of the glass plate g is detected, the type thereof is identified, and the position of the defect in the thickness direction of the glass plate G is determined. A display 32 is connected to the processing device 16, and the display is used on the display 32. The image obtained by the first defect inspection device 12 and the second defect inspection device 14, the detection result of the defect, the recognition result, or the position determination result of the defect. As described above, the image obtained by the first defect inspection device 12 is bright. In the field of view image, the defect of the glass plate G appears as a dark portion in the image due to the diffuse reflection of the defective region. Further, as described above, the optical system of the first camera 22 is disposed directly in front of the optical system. The blade-shaped optical path shielding member 24 is partially blocked by the optical path, whereby the optical path shielding member 24 is used to form a bright portion in such a manner as to approach or face or face the region of the dark portion in the image. The façade is caused by the refracting abnormality of the portion of the defect, and the brightness thereof is higher than the background portion of the bright field image. Fig. 4 is a schematic view showing an example of a defect image when the optical path shielding member 24 is in the optical path. As shown, a bright portion that is close to and opposite to the area of the dark portion is formed. In contrast, FIG. 4B is a defect diagram when the light path shielding member 24 is not present in the optical path 142304.doc -14 - 201009324 (the light is not blocked). A pattern diagram like an example. As shown in Fig. 4b, the portion that is not open/close to and close to the dark portion is bright. It is generally considered that in the child-view % image, as shown in FIG. 4A, the bright portion is formed in such a manner as to approach and face the region of the dark portion. The reason is that the glass plate is present due to bubbles or foreign matter present on the surface or inside of the glass plate. The surface of G has an uneven shape, and the refractive error of the glass sheet G slightly changes. Actually, a bright portion as shown in Fig. 4A is not formed due to scratches or foreign matter adhering to the surface of the glass sheet G. Therefore, in the processing device 16, a threshold value higher than the value of the image data of the background portion in the bright field image is set in the bright field image, and the region above the threshold is extracted as the bright portion. Furthermore, the image data transmitted from the first camera 22 is subjected to the inspection position, and the image data of one dimension is read by the line sensor type camera, so that the processing device 16 stores a plurality of lines (for example, 500 lines). After obtaining the image of a certain area, the image area of the above-mentioned bright part is extracted. When the defective area existing on the glass sheet G is detected, the positional information of the extracted bright portion is used in the following manner. The image data which is transmitted to the processing device 6 by the second defect inspection device 14 is the material of the bright field reflection image, and the portion of the defect becomes the image of the region of the dark portion. As described above, the real image and the image of the defect appear as dark portions. According to the position of the glass plate g in the thickness direction of the defect of the glass sheet G, the real image of the defect and the mirror image are displaced. For example, when the defect is located near the back surface of the glass sheet G, the deviation amount between the real image and the mirror image becomes small, and when the defect is located near the surface, the deviation between the real image and the mirror image position 142304.doc -15-201009324 becomes large. Therefore, the processing unit 16 determines the position of the defective glass sheet G in the width direction by using the information of the area of the extracted bright portion. Further, using the position in the width direction, the time deviation of the image of the same portion where the glass plate G appears between the image obtained by the second defect inspection device 12 and the image obtained by the second defect inspection device T4 is used. The amount is detected for the real image of the defect obtained by the second camera 3 and the dark portion of the mirror image. Since the distance between the measurement position of the work defect inspection device 12 and the measurement position of the second defect inspection device 14 in the transport direction and the transport speed of the glass crucible are known, the time deviation amount can be obtained from the information. . In order to extract the dark portion, the threshold is set in advance. Then, the position deviation amount 'from the center position of the real image and the center position of the mirror image is obtained, and the position of the defect in the thickness direction of the glass sheet 算出 is calculated based on the position deviation amount. Further, the processing device 16 obtains the size of the dark portion using the real image of the defect obtained by the second camera 3, and estimates the size of the defect based on the size of the region. Further, the processing unit 16 is independent of the processing of the information based on the area of the bright portion, and the area of the dark portion is extracted using the threshold value set in advance without using the information of the area of the bright portion. The processing device 16 uses the information of the region of the dark portion that has been extracted and the region of the bright portion to determine whether the region of the bright portion and the region of the dark portion are close to each other and face each other, and uses the calculated thickness direction of the defect existing in the glass plate G. The position and the size of the defect determined based on the extracted dark portion and the feature amount of the defect determine the type of the defect. Preferably, the shape of the defect is identified according to the area of the bright portion and the dark portion of the image and the dark portion of the bright field reflected image in the bright field of view 142304.doc -16 - 201009324, thereby distinguishing The type of the defect is estimated for, for example, a defect caused by a bubble or a scratch or the like caused by a foreign matter. A defect such as a foreign matter or a scratch existing on the surface of the glass sheet G does not form a bright portion in the bright field image obtained from the missing inspection device 12. Further, in the defect detecting device 12, it is preferable to satisfy the following conditions in order to effectively bring out the bright portion of the bright field image which is close to and opposite to the dark portion of the image. That is, in the arrangement of the respective portions shown in FIG. 2, the first condition that the value obtained by multiplying the value α by the ratio of the angle of view of the illumination light-emitting effective angle 9 is greater than 2 is set. The distance between the first camera 22 and the glass plate G, the distance between the glass plate G and the first linear light source 2A, and the irradiation width L of the t-th linear light source. Further, the aperture value of the i-th camera 22 and the distance between the first camera 22 and the glass plate (3) are set so as to satisfy the second condition, and the second condition is defined by the arrangement position of the optical path shielding member 24. When the blurring circle of the edge of the field of view defined by the light receiving surface is formed by the plane orthogonal to the optical axis of the imaging lens 23, the area of the portion of the blurred circle is blocked by the optical path shielding member 24 to become the area of the blurred circle. The first camera 22 is disposed so that the light receiving elements of the first camera 22 sandwich the glass sheet G and are positioned in the direction of the maximum light intensity of the linear light source 20. The three conditions are arranged so as to satisfy the three conditions. The device, whereby the bright field image can be effectively generated by the bright portion. 142304.doc 17 201009324 By reducing the F value, the focus of the subject in the shooting is shallow, and the image of the image is likely to occur. Therefore, in order to efficiently extract the bright portion, it is preferably F5.6 to FU, and it is preferable that the width L of the light-emitting portion 34 of the first camera 2 is in the range of 1 to 20, The value obtained by multiplying the value α by the above ratio is greater than 2. In the table shown in FIG. 5, the numerical value of the illumination light-emitting effective angle θ defined by the width L of the light-emitting portion 34 of the first camera 20 is L=1 mm, 3 positions, 4 rhymes, 5 axes, and 7 colors, and The value of the field of view angle 规定 specified by the value in the corresponding column of each table indicates the value obtained by multiplying the value of Lu/Θ by the value of the second light source 2〇2α. In the table shown in Figure 5Α&5Β In the range enclosed by the thick frame, when the g value α is approximately 1, under the condition that the value of α·#/θ shown in FIG. 4A is greater than 2, it has been confirmed that the bright portion is effectively present. When the value α of the t-line light source 20 is smaller than 丨 (considering the illuminance directivity), it can be said that the bright portion is effectively generated under the condition that α·~θ is larger than 2. Further, from the glass plate G to the first camera 22 The distance from the surface of the imaging lens 23 is set to 380 mm. The illumination WD (w〇rking Distance) from the irradiation surface of the linear light source 2〇 to the measurement position of the glass plate G is shown in Fig. 5A. 200 mm, which is 4 mm in Fig. 5B_. At this time, the area of the portion of the blurred circle by the optical path shielding member 24 becomes 50% of the area of the blurred circle. The defect inspection system 1 is effective in that the parallel light can be obliquely irradiated to the glass plate G, and the defects which are missed in the automatic defect inspection similar to the schlieren method can be extracted with high precision, and the difference can not be repaired. The defect of the glass plate G. Thereby, the defective portion which cannot be repaired can be avoided, and the glass plate G is cut into a specific size by 142304.doc -18-201009324. The defect inspection system and the defect inspection method of the present invention have been carried out. The present invention is not limited to the above-described embodiments or examples, and various modifications and changes can be made without departing from the spirit and scope of the invention. The present invention has been described in detail with reference to the particular embodiments of the present invention. It is understood that various changes and modifications may be added without departing from the spirit and scope of the invention. The present application is based on Japanese Patent Application No. 2008-215091, filed on Jan. 25, 2008. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a view for explaining a schematic configuration of an embodiment of a defect inspection system and a defect inspection method according to the present invention; FIG. 2 is a view showing an effective illumination angle and field of illumination in the defect inspection system of the present invention. Fig. 3 is a view for explaining α used in the defect inspection system of the present invention, and Fig. 4A is a view schematically showing an example of an image obtained by the first defect inspection device of the present invention. FIG. 4B is a view schematically showing an example of an image obtained by a bright field image formed by normal transmitted light; FIG. 5A is a first defect inspection device for the defect system shown in FIG. 1. FIG. 5B is a view illustrating a preferred range for generating a bright portion in the first defect inspection device 142304.doc _ 19- 201009324 of the defect system shown in FIG. 1; FIG. 6 is a diagram for explaining a schematic configuration of a conventional defect inspection device. [Description of main component symbols] 10 Defect inspection system 12 First defect inspection device 14 Second defect inspection device 16 Processing device 18 Driving roller 20 First linear light source 22 First camera 23 Imaging lens 24 Optical path shielding member 28 Second light source 30 2nd camera 32 display 34 light emitting part 142304.doc -20-