JP3332208B2 - Defect detection method and apparatus for netted glass - Google Patents

Defect detection method and apparatus for netted glass

Info

Publication number
JP3332208B2
JP3332208B2 JP16826697A JP16826697A JP3332208B2 JP 3332208 B2 JP3332208 B2 JP 3332208B2 JP 16826697 A JP16826697 A JP 16826697A JP 16826697 A JP16826697 A JP 16826697A JP 3332208 B2 JP3332208 B2 JP 3332208B2
Authority
JP
Japan
Prior art keywords
glass
net
signal
defect
line
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
JP16826697A
Other languages
Japanese (ja)
Other versions
JPH1114334A (en
Inventor
真一 岡村
耕三 若野
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Central Glass Co Ltd
Original Assignee
Central Glass Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Central Glass Co Ltd filed Critical Central Glass Co Ltd
Priority to JP16826697A priority Critical patent/JP3332208B2/en
Publication of JPH1114334A publication Critical patent/JPH1114334A/en
Application granted granted Critical
Publication of JP3332208B2 publication Critical patent/JP3332208B2/en
Anticipated expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Description

【発明の詳細な説明】DETAILED DESCRIPTION OF THE INVENTION

【0001】[0001]

【発明の属する技術分野】本発明は網または線入りガラ
スの欠陥検出方法、および装置に関し、画像処理技術に
より網または線の形状に無関係に、画像信号より網部分
を除去し欠陥を検出する。
BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and an apparatus for detecting a defect in a glass having a net or a line, and detects a defect by removing a net portion from an image signal irrespective of the shape of a net or a line by an image processing technique.

【0002】[0002]

【従来の技術】従来より網入りガラスの欠陥検出方法や
装置は数多く知られている。例えば、特開平6−347
410号公報には、検査すべき網入り板ガラスの画像を
取り込むステップと、前記取り込んだ画像を網形状に対
応した所定量だけ移動させるステップと、移動前の画像
と移動後の画像とを重ね合わせるステップと、重ね合わ
せた画像のうち重複した部分を取り除くステップと、前
記重複した部分を取り除いた画像から欠陥部を判別する
ステップとからなる網入りガラスの欠陥検出方法が開示
されている。
2. Description of the Related Art Many methods and apparatuses for detecting defects in screened glass have been known. For example, JP-A-6-347
No. 410 discloses a step of capturing an image of a screened glass sheet to be inspected, a step of moving the captured image by a predetermined amount corresponding to a mesh shape, and a step of superimposing an image before movement and an image after movement. There is disclosed a method of detecting a defect in a glass-screened glass, comprising: a step of removing an overlapped portion from an overlapped image; and a step of determining a defective portion from the image from which the overlapped portion has been removed.

【0003】また、特開平8−304295号公報に
は、一方向に搬送される、内部または表面に高反射率の
物体を含む板状透明物体の表面に存在する欠陥を検出す
る方法において、前記板状透明物体の一方の面の側にそ
の長さ方向が前記一方向と直交する方向に対して角度θ
をなすように設けられた照明装置により、前記一方の面
を照明し、前記一方の面の側に設けられたラインセンサ
により、前記一方向と直交する方向に対し前記角度θを
なす方向を走査方向として、前記板状透明物体の一方の
面を撮像する表面欠陥検出方法、および装置が開示され
ている。
Japanese Patent Application Laid-Open No. Hei 8-304295 discloses a method for detecting a defect present on the surface of a plate-shaped transparent object that is transported in one direction and includes an object having a high reflectance inside or on the surface. On one side of the plate-shaped transparent object, the length direction has an angle θ with respect to a direction orthogonal to the one direction.
The one surface is illuminated by an illuminating device provided so as to form a line, and the line sensor provided on the one surface side scans a direction making the angle θ with respect to a direction orthogonal to the one direction. As a direction, a surface defect detection method and an apparatus for imaging one surface of the plate-shaped transparent object are disclosed.

【0004】さらに、本出願人による特開平6−148
100号公報にて、連続して搬送されるリボン状の網入
りガラスに光源からの光を投光し、網入りガラスとは反
対側に設けた複数台の2次元カメラにより網入りガラス
を撮像し、得られた濃淡の信号により網入りガラスの検
査をする方法において、所定の範囲に縦、横あるいは斜
め方向に連続した濃信号があるときに、その濃信号の幅
方向の中心から等間隔で網以上の幅にその濃信号を除去
し、残った信号の大きさを求めることにより欠点を検査
するようにした網入りガラスの検査方法を開示してい
る。
Further, Japanese Patent Laid-Open No. 6-148 filed by the present applicant has been disclosed.
In Japanese Patent No. 100, light from a light source is projected onto a ribbon-shaped netted glass that is continuously conveyed, and the two-dimensional cameras provided on the opposite side of the netted glass image the netted glass. Then, in the method of inspecting the netted glass based on the obtained density signal, when there is a density signal that is continuous in a vertical, horizontal or diagonal direction in a predetermined range, the density signal is equally spaced from the center in the width direction. Discloses a method for inspecting a screened glass in which the dark signal is removed to a width larger than the screen, and the defect is inspected by determining the magnitude of the remaining signal.

【0005】[0005]

【発明が解決しようとする課題】前記の特開平6−34
7410号公報に記載のものは、網入りガラスの成型時
に網が引き延ばされる等により僅か変形するため、ガラ
スリボンの幅方向のパターンが僅か異なり同一形状でな
いため、取り込んだ画像を所定量だけ移動させて移動前
の画像と移動後の画像とを重ね合わせ、重ね合わせた画
像のうち重複した部分を取り除く処理において、重なら
ないケースが発生する。また、網の上に欠陥が存在する
場合は網を消去した後に欠陥が分割されて検出され、欠
陥サイズを正しく計測できないという懸念がある。
The above-mentioned Japanese Patent Application Laid-Open No. Hei 6-34 is disclosed.
The glass described in Japanese Patent No. 7410 is slightly deformed due to elongation of the net when forming the netted glass, so the pattern in the width direction of the glass ribbon is slightly different and not the same shape, so that the captured image is moved by a predetermined amount. Then, in the process of superimposing the image before the movement and the image after the movement and removing the overlapping portion of the superimposed images, a case where the images do not overlap occurs. Further, when a defect exists on the net, there is a concern that the defect is divided and detected after erasing the net, and the defect size cannot be measured correctly.

【0006】また、特開平8−304295号公報に記
載のものは、反射を利用して欠陥検査をする方法のた
め、網入りガラスの表面欠陥しか検出することができ
ず、また本公報に記載のものは表面が高反射率の物体の
表面欠陥検査に関するもので、研磨前や研磨途中の網入
りガラスの表面が平滑でなく、表面に微細な凹凸がある
場合は等は光源からの光が散乱するために欠陥の検出が
できないという問題点があった。
The method disclosed in Japanese Patent Application Laid-Open No. Hei 8-304295 is a method for inspecting defects by utilizing reflection, so that it can detect only surface defects of screened glass. Is related to surface defect inspection of objects with high reflectivity.If the surface of the meshed glass before or during polishing is not smooth and there are fine irregularities on the surface, etc. There is a problem that the defect cannot be detected due to scattering.

【0007】さらに、特開平6−148100号公報に
記載のものは、濃信号が連続している場合に網とみなす
方法であり、網を確実に撮像し、画像データの網部分が
途切れることの無く、連続するように1つの画素幅を小
さくしなけらばならず、また網を確実に除去しようとす
る場合、マスキング幅を大きめにとらざるを得ないた
め、カメラの視野を大きくとることができす、カメラの
台数が増加し、コストアップとなるという問題点があ
る。また網の種類が平行な金属線、菱形(菱網)、正方
形(角網)等によって検出アルゴリズムが異なり、複雑
になるという問題点もあった。
Further, the method described in Japanese Patent Application Laid-Open No. 6-148100 is a method in which a dark signal is regarded as a net when it is continuous. It is necessary to reduce the width of one pixel so that it is continuous, and in order to remove the net without fail, the masking width must be set to be large, so that the camera's field of view may be large. However, there is a problem that the number of cameras can be increased and the cost is increased. There is also a problem that the detection algorithm differs depending on the type of the mesh, such as a parallel metal wire, a rhombus (rhombus), a square (square mesh), and the like, which is complicated.

【0008】[0008]

【課題を解決するための手段】本発明は、上記問題点、
すなわち網入りガラスの網の種類に係わらず同一アルゴ
リズムで確実に網部分の画像のみを消去して欠陥部と網
部とを判別し、欠陥の検出を行うことを目的とするもの
である。 すなわち、本発明は、連続して搬送されるリボ
ン状の網または線入りガラスに光源からの光を投光し、
網または線入りガラスを挟んで光源とは反対側に設けた
複数台のエリアカメラにより網または線入りガラスを撮
像し、得られた濃淡の信号により網または線入りガラス
の検査をする方法において、撮像した画像データの二値
化処理後、濃信号部分の全周囲を一画素分消去する収縮
処理を所定回数行うことにより網または線に相当する濃
信号のみを消去させ、残った濃信号を欠陥とし、画像デ
ータに基づいて、その大きさ、位置等を判別することを
特徴とする網入りガラスの欠陥検出方法である。あるい
は、本発明は、前記エリアカメラに代えて、搬送される
網または線入りガラスの幅方向に走査するラインカメラ
を複数台配置し、該ラインカメラで網または線入りガラ
スの搬送速度に同期させて取り込んだ信号を記憶保持さ
せて2次元の画像データとし、濃信号部分の全周囲を一
画素分消去する収縮処理を所定回数行うことにより網ま
たは線に相当する濃信号のみを消去させ、残った濃信号
を欠陥とし、画像データに基づいて、その大きさ、位置
等を判別することを特徴とする網入りガラスの欠陥検出
方法である。あるいはまた、本発明は、淡信号部分を収
縮させる画像処理装置を用いる場合に、上述の網入りガ
ラスの欠陥検出方法の画像信号を二値化処理した後、網
線の部分と欠陥の部分を濃信号として収縮処理する方法
に代えて、収縮処理する前にネガポジ反転して網線の部
分と欠陥の部分を淡信号とし、ガラスの正常部分を濃信
号とした後に、収縮処理を行って網線の部分と欠陥の部
分の淡信号部分を収縮させる方法とすることを特徴とす
る網入りガラスの欠陥検出方法である。 あるいはまた、
本発明は、連続して搬送されるリボン状の網または線入
りガラスに光を投光する光源と、網または線入りガラス
を挟んで光源とは反対側に設け、網または線入りガラス
を撮像する複数台のエリアカメラと、該エリアカメラに
より得られた画像データを処理する画像処理手段と、該
画像データにより欠陥部を判別する検出制御手段からな
る網または線入りガラスの欠陥検出装置において、前記
検出制御手段は画像データを二値化処理後、濃信号部分
の全周囲を一画素分消去する収縮処理を所定回数行うこ
とにより網または線に相当する濃信号のみを消去させる
機能、および残った濃信号を欠陥とし、画像データに基
づいてその大きさ、位置等を判別する機能を有するよう
構成したことを特徴とする網入りガラスの欠陥検出装
置であるあるいはまた、本発明は、前記エリアカメラ
に代えて、搬送される網または線入りガラスの幅方向に
走査するラインカメラを複数台配置し、該ラインカメラ
で網または線入りガラスの搬送速度に同期させて取り込
んだ信号を記憶保持させて2次元の画像データとする記
憶装置とからなることを特徴とする上述の網入りガラス
の欠陥検出装置である。
SUMMARY OF THE INVENTION The present invention provides the above-mentioned problems,
In other words, the purpose of the present invention is to reliably delete only the image of the net portion by the same algorithm regardless of the type of net of the netted glass, discriminate the defective portion from the net portion, and detect the defect.
It is. That is, the present invention projects light from a light source onto a ribbon-shaped net or a lined glass that is continuously conveyed,
In a method of inspecting the net or the lined glass by a signal of the obtained density, the net or the lined glass is imaged by a plurality of area cameras provided on the opposite side to the light source with the net or the lined glass interposed therebetween, After the binarization processing of the captured image data, by performing a predetermined number of times of contraction processing for erasing one pixel around the entire dark signal portion, only the dark signal corresponding to the mesh or line is deleted, and the remaining dark signal is defective. And determining the size, position, etc., based on the image data.
This is a method for detecting defects in a screened glass. Alternatively, according to the present invention , instead of the area camera, a plurality of line cameras that scan in the width direction of the net or lined glass to be conveyed are arranged, and the line camera synchronizes with the conveying speed of the net or lined glass. The captured signal is stored and retained to form two-dimensional image data, and a predetermined number of times of contraction processing for erasing the entire periphery of the dark signal portion by one pixel eliminates only the dark signal corresponding to a net or a line, thereby removing the remaining signal. Defect detection using a dark signal as a defect and determining its size, position, etc. based on image data
Is the way. Alternatively, the present invention collects the light signal portion.
When using an image processing device for shrinking,
After binarizing the image signal of the Las defect detection method,
A method of shrinking a line part and a defect part as a dark signal
Instead of the negative / positive reversal before shrinking,
Minute and defective parts are light signals, and normal parts of glass are thick signals
After the shrinkage process, shrinkage processing is performed to
The light signal portion of the minute
This is a method for detecting defects in screened glass. Alternatively,
The present invention provides a light source that projects light on a ribbon-shaped net or lined glass that is continuously conveyed, and is provided on the opposite side of the light source with the net or lined glass interposed, and images the net or lined glass. A plurality of area cameras, image processing means for processing image data obtained by the area camera, and a net or lined glass defect detection device comprising detection control means for determining a defective portion based on the image data, The detection control means performs a predetermined number of times of contraction processing for erasing the entire periphery of the dark signal portion by one pixel after binarizing the image data, thereby erasing only the dark signal corresponding to a net or a line. Characterized in that it has a function of determining the size, position, etc. of the dark signal as a defect based on the image data.
It is a place . Alternatively, according to the present invention , instead of the area camera, a plurality of line cameras that scan in the width direction of the net or lined glass to be conveyed are arranged, and the line camera synchronizes with the conveyance speed of the net or lined glass. The above-described defect detection apparatus for a glass-screened glass, comprising: a storage device that stores and holds a signal fetched and makes two-dimensional image data.

【0009】[0009]

【発明の実施の形態】本発明は、連続して搬送されるリ
ボン状の網入りガラス中に含まれる網線等の金属線以外
のガラス中に含まれる欠陥、例えば泡や、失透、タレ、
砂利異物等の未溶解による欠陥を検出するものである。
BEST MODE FOR CARRYING OUT THE INVENTION The present invention relates to a method for manufacturing a glass sheet having a ribbon-like mesh, which is continuously conveyed, and which contains defects such as bubbles, devitrification, sagging, etc. ,
This is to detect defects due to undissolved material such as gravel.

【0010】構成は前記した通りであるが、搬送される
網入りガラスの下方に面状の光源を設け、該光源の上方
で網入りガラスを挟んで前記光源とは反対側には複数台
のエリアカメラ(二次元カメラ)を一列に配置し、該光
源から網入りガラスに光を投光させ、前記複数台のエリ
アカメラにより網入りガラスを撮像し、網線や欠陥等の
光を透過させない黒色部分、光を透過させる白色部分、
光を一部だけ透過させる部分を灰色部分等、例えば25
6階調の濃淡レベルの画像信号を得、画像処理装置に取
り込む。
The structure is as described above, but a planar light source is provided below the netted glass to be conveyed, and a plurality of light sources are provided above the light source on the opposite side of the light source with the netted glass therebetween. Area cameras (two-dimensional cameras) are arranged in a line, light is emitted from the light source to the netted glass, the netted glass is imaged by the plurality of area cameras, and light such as mesh lines and defects are not transmitted. Black part, white part that transmits light,
A part that transmits light only partially is a gray part, for example, 25.
An image signal of six gradation levels is obtained and taken into the image processing device.

【0011】画像処理装置は撮像した画像データを所定
の輝度レベル(スライスレベル)を境界とし、スライス
レベルよりも暗い画素部分を濃レベル(1)とし、スラ
イスレベルよりも明るい画素部分を淡レベル(0)とす
る二値化処理を行う。
The image processing apparatus sets captured image data at a predetermined luminance level (slice level) as a boundary, sets a pixel portion darker than the slice level to a dark level (1), and a pixel portion brighter than the slice level to a light level (1). 0) is performed.

【0012】二値化処理後、濃信号部分の全周囲を一画
素分消去する収縮処理を所定回数行い、網に相当する濃
信号のみを消去させた段階で、網線の示す画素数より多
い画素数でなる欠陥部分は消去されないで残る画素分が
あるので、残った濃信号を欠陥とみなし、収縮回数から
欠陥の大きさを逆算推定し、欠陥位置、大きさ等をパソ
コン等に出力記憶させ、後工程において欠陥部分を含む
箇所のガラスを切断除去させればよい。
After the binarization processing, a predetermined number of times of contraction processing for erasing the entire periphery of the dark signal portion by one pixel is performed, and only the dark signal corresponding to the halftone is erased. Since the defective portion consisting of the number of pixels remains for some pixels without being erased, the remaining dark signal is regarded as a defect, the size of the defect is estimated backward from the number of contractions, and the defect position and size are output and stored in a personal computer etc. Then, in a later step, the glass at the portion including the defective portion may be cut and removed.

【0013】尚、収縮処理とは画像処理装置において周
知の技術であって、濃信号部分または淡信号部分の全周
囲を一画素分消去するものである。また、検出できる欠
陥の大きさは網線の線径より大であり、カメラの視野に
よる画素の大きさ、および収縮処理回数により左右され
るが、収縮処理する回数は画素の大きさ、網線の線径等
により決定される。
The shrinking process is a well-known technique in an image processing apparatus, and is to erase the entire periphery of a dark signal portion or a light signal portion by one pixel. The size of the defect that can be detected is larger than the wire diameter of the mesh line, and depends on the size of the pixel in the field of view of the camera and the number of times of the shrinking process. Is determined by the wire diameter or the like.

【0014】本発明のように、画像処理装置の収縮処理
により網線を除去すれば、網模様のパターンを記憶し処
理する等の複雑な処理が不要となり、網線の種類に係わ
らず網線を表す濃信号画像を除去でき、同一ロジックと
なるので、演算処理速度も高速化できる。
As in the present invention, if a halftone line is removed by the contraction process of the image processing apparatus, complicated processing such as storing and processing a halftone pattern becomes unnecessary, and the halftone line is independent of the type of halftone line. Can be removed and the same logic can be used, so that the arithmetic processing speed can be increased.

【0015】[0015]

【実施例】以下、本発明に係わる網入りガラスの欠陥検
出方法、および装置について図面に基づき詳細に説明す
る。
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A method and an apparatus for detecting a defect of netted glass according to the present invention will be described below in detail with reference to the drawings.

【0016】図1に示すように、矢印方向にリボン状に
連続的に搬送される網入りガラス1の下方に欠陥3を検
出するためのボックスに収納された高周波蛍光灯などの
光源4を配設し、網入りガラス1の上方には網入りガラ
ス1の全幅を撮像するために検査幅を網入りガラスの全
幅よりやや大きめとし、この検査幅を分割して複数台の
CCDカメラなどのエリアカメラ5、5、・・を一列に
配設した。
As shown in FIG. 1, a light source 4 such as a high-frequency fluorescent lamp housed in a box for detecting a defect 3 is disposed below a netted glass 1 continuously conveyed in a ribbon shape in the direction of an arrow. The inspection width is set slightly larger than the entire width of the netted glass in order to image the entire width of the netted glass 1 above the netted glass 1, and the inspection width is divided to divide the area of a plurality of CCD cameras and the like. The cameras 5, 5, ... are arranged in a line.

【0017】各エリアカメラ5、5、・・から出力され
る画像信号は、画像処理装置6に入力され、画像処理装
置6で網線2を表す画像信号を除去して、網入りガラス
1の欠陥3を表す画像信号のみとし、欠陥位置、大きさ
等の信号をパソコン7などに出力する。
An image signal output from each of the area cameras 5, 5,... Is input to an image processing device 6, and the image processing device 6 removes the image signal representing the reticulated line 2, and outputs Only the image signal representing the defect 3 is output, and signals such as the defect position and size are output to the personal computer 7 or the like.

【0018】前記の画像処理装置6による網線2部分の
除去、及び欠陥3の検出手順について、さらに詳細に説
明する。網入りガラス1の下方に設けた光源4によって
光を照射された網入りガラス1を複数台のエリアカメラ
5、5、・・で撮像し、画像信号を画像処理装置6に出
力する。画像処理装置6に入力された256階調の濃淡
信号は画像処理装置6によって2値化処理され、濃信号
(1)か淡信号(0)のいずれかになる。2値化処理に
ついては各画素毎にその輝度の信号レベルが所定の値
(スライスレベル)より大きいかどうかが比較され、大
きい場合は淡信号、小さい場合は濃信号として表示さ
れ、図2に示すように網と欠陥部分は黒く塗りつぶした
濃信号として表示される。
The procedure for removing the portion of the reticulated line 2 and detecting the defect 3 by the image processing apparatus 6 will be described in further detail. A plurality of area cameras 5, 5,... Capture images of the screened glass 1 illuminated by the light source 4 provided below the screened glass 1, and output image signals to the image processing device 6. The grayscale signal of 256 gradations input to the image processing device 6 is binarized by the image processing device 6 to be either a dark signal (1) or a light signal (0). In the binarization process, it is compared whether the signal level of the luminance is greater than a predetermined value (slice level) for each pixel, and when it is larger, it is displayed as a light signal, and when it is smaller, it is displayed as a dark signal, as shown in FIG. As described above, the net and the defective portion are displayed as dark signals filled in black.

【0019】いま、カメラ5の分解能により格差がある
が、1画素の大きさを0.5mmとし、ガラス中に挿入
する網の線径が0.42〜0.64mmφとすれば、網
線2の部分が1〜2画素となり、網線2と網線2の交点
部分は4画素程度となり、検出できる欠陥部分は縦方
向、横方向ともに5画素程度以上となる。
Although there is a difference depending on the resolution of the camera 5, if the size of one pixel is 0.5 mm and the wire diameter of the mesh inserted into the glass is 0.42 to 0.64 mmφ, the mesh 2 Is 1-2 pixels, the intersection of the mesh lines 2 is about 4 pixels, and the detectable defect is about 5 pixels or more in both the vertical and horizontal directions.

【0020】画像処理装置6による収縮処理について
は、図5に示すように斜線部分と黒く塗った部分を併せ
てあらわした欠陥3である濃信号の画素に対し、収縮処
理を1回行うと、濃信号の画素と淡信号の画素の境界部
の濃信号の画素、つまり全周囲の左下がりの斜線部の画
素が淡信号化して消え、黒く塗った画素部分、及び右下
がりの斜線部のみが残る。
As for the contraction processing by the image processing device 6, as shown in FIG. 5, once the contraction processing is performed on the pixel of the dark signal which is the defect 3 which represents the shaded portion and the black portion, The dark signal pixels at the boundary between the dark signal pixels and the light signal pixels, that is, the pixels at the lower left diagonal lines around the entire area are converted to light signals and disappear, and only the pixel portions painted black and the lower right diagonal lines are removed. Remains.

【0021】いま、画像処理装置6によって、図2に示
すような画像データに対して収縮処理を1回行うと、図
3のようになり、網線2を表す部分と欠陥3を表す部分
がそれぞれ細くあるいは小さくなる。1回の収縮処理で
は網線2の交点部分に濃信号が残ってしまう場合が起き
るため、収縮処理を2回おこなって確実に網線部分を除
去する必要がある。
Now, when the image processing apparatus 6 performs the contraction processing once on the image data as shown in FIG. 2, the result becomes as shown in FIG. Each becomes thinner or smaller. In one contraction process, a dark signal may remain at the intersection of the halftone line 2, so it is necessary to perform the contraction process twice to reliably remove the halftone portion.

【0022】図4は2回目の収縮処理を行った図であ
り、網線2の部分が除去され、欠陥3の部分のみが残っ
ている状態である。本実施例で収縮処理の回数を2回と
したのは、網線2のクロス部に濃信号の画素が残るケー
スがあるためであり、カメラの絞りを開放させ明るくさ
せれば、光の屈折により網線2の部分の濃信号部分が細
くなり、1回の収縮処理でも網線2の部分が除去され
る。このため収縮処理回数は画素の大きさ、線径、検出
したい欠陥3の大きさ、カメラの絞り等により適宜選択
すればよい。
FIG. 4 is a view in which the second shrinking process has been performed, and shows a state in which the portion of the mesh line 2 has been removed and only the portion of the defect 3 remains. The reason why the number of times of the contraction processing is set to two in this embodiment is that there is a case where a dark signal pixel remains in the cross portion of the mesh line 2. As a result, the dark signal portion of the portion of the mesh line 2 becomes thin, and the portion of the mesh line 2 is removed even by one contraction process. Therefore, the number of times of contraction processing may be appropriately selected depending on the size of the pixel, the diameter of the line, the size of the defect 3 to be detected, the aperture of the camera, and the like.

【0023】以上のようにして、網入りガラス1や線入
りガラスの網線2の部分を除去しガラス中の欠陥を検出
したがこれに限るものではない。カメラはエリアカメラ
(二次元カメラ)5を用いたが、ラインカメラを用い網
入りガラス1の幅方向に走査させ、網入りガラス1の搬
送に合わせて取り込んだ信号をメモり内に記憶保持さ
せ、2次元の画像データとすれば、前記エリアカメラ5
の処理と同様に処理させることができる。
As described above, the netted glass 1 and the portion of the netted wire 2 of the lined glass are removed to detect defects in the glass, but the present invention is not limited to this. An area camera (two-dimensional camera) 5 was used as the camera, but a line camera was used to scan in the width direction of the netted glass 1 and a signal taken in accordance with the conveyance of the netted glass 1 was stored and stored in a memory. If the image data is two-dimensional, the area camera 5
Can be performed in the same manner as the above processing.

【0024】尚、ラインカメラを使用した場合は同一分
解能で走査幅が広くとれるのでカメラの台数を減らすこ
とができ、また光源4についてもラインカメラの走査方
向である網入りガラス1の幅方向に直線上に設ければ良
いので、簡略化できる。
When a line camera is used, the scanning width can be widened with the same resolution, so that the number of cameras can be reduced. Also, the light source 4 can be moved in the scanning direction of the line camera in the width direction of the netted glass 1. Since it may be provided on a straight line, it can be simplified.

【0025】エリアカメラ5、5、・・で撮像した画像
信号を二値化処理した後、網線2の部分と欠陥3の部分
を濃信号として収縮処理したが、収縮処理する前にネガ
ポジ反転して網線2の部分と欠陥3の部分を淡信号とし
ガラスの正常部分を濃信号とした後に、収縮処理を行っ
て網線2の部分と欠陥3の部分の淡信号部分を収縮させ
る方法でも良い。これは画像処理装置6の種類によって
は収縮処理命令が淡信号部分を収縮させるようにしてい
る場合もあり、適宜必要に応じて行えばよい。
After binarizing the image signals picked up by the area cameras 5, 5,..., The portion of the reticulated line 2 and the portion of the defect 3 are contracted as dark signals. A method of shrinking the light signal portions of the portion of the mesh line 2 and the defect 3 by performing shrinkage processing after making the portion of the mesh line 2 and the portion of the defect 3 a light signal and making the normal portion of the glass a dark signal. But it is good. Depending on the type of the image processing device 6, the contraction processing command may cause the light signal portion to contract, and may be appropriately performed as needed.

【0026】前記説明で濃信号部あるいは淡信号部の収
縮処理について説明したが、濃信号部と淡信号部を逆に
して膨張処理と表現する場合もある。以上のようにして
検出された欠陥3は、更に通常ラベリングと呼ばれる演
算処理を行い、求められた欠陥3の位置と検出画素数等
の情報をパソコン7等に出力記憶させ、収縮回数と画素
の大きさから、実欠陥の大きさを推定し、後工程におい
てパソコン7等の指示制御に基づいて欠陥3と判定され
た部分を含む箇所のガラスを切断除去し、あるいは各種
管理情報に加工することもできる。
In the above description, the contraction process of the dark signal portion or the light signal portion has been described. However, the dark signal portion and the light signal portion may be reversed and expressed as expansion processing. The defect 3 detected as described above is further subjected to a calculation process usually called labeling, and information such as the position of the defect 3 and the number of detected pixels is output and stored in the personal computer 7 or the like. Estimating the size of the actual defect from the size, and cutting and removing the glass including the portion determined as the defect 3 based on the instruction control of the personal computer 7 or the like in a later process, or processing the glass into various management information. Can also.

【0027】[0027]

【発明の効果】本発明は、網入りガラスや線入りガラス
の網の模様種類に関係なく、撮像した網入りガラスや線
入りガラスの画像を収縮処理することにより網線部分を
同一アルゴリズムで除去でき、ガラス中の欠陥を確実に
検出できる。
According to the present invention, the reticulated portion is removed by the same algorithm by shrinking the image of the reticulated glass or the lined glass regardless of the pattern type of the reticulated glass or the lined glass. It is possible to reliably detect defects in the glass.

【図面の簡単な説明】[Brief description of the drawings]

【図1】本発明の検査装置の要部概略図FIG. 1 is a schematic view of a main part of an inspection apparatus according to the present invention.

【図2】網入りガラスの一部を撮像し、2値化した画像
信号を示す図。
FIG. 2 is a diagram showing a binarized image signal obtained by imaging a part of netted glass;

【図3】図2の画像信号に収縮処理を行った状態を示す
図。
FIG. 3 is a diagram showing a state in which a contraction process has been performed on the image signal of FIG. 2;

【図4】図3の画像信号にさらに収縮処理を行った状態
を示す図。
FIG. 4 is a diagram showing a state in which a contraction process is further performed on the image signal of FIG. 3;

【図5】収縮処理を説明する図。FIG. 5 is a diagram illustrating a contraction process.

【符号の説明】[Explanation of symbols]

1 網入りガラス 2 網線 3 欠陥 4 光源 5 エリアカメラ 6 画像処理装置 7 パソコン DESCRIPTION OF SYMBOLS 1 Netted glass 2 Netting 3 Defect 4 Light source 5 Area camera 6 Image processing device 7 Personal computer

───────────────────────────────────────────────────── フロントページの続き (56)参考文献 特開 平8−114432(JP,A) 特開 平6−347410(JP,A) 特開 平6−148100(JP,A) 特開 平6−18432(JP,A) (58)調査した分野(Int.Cl.7,DB名) G01B 11/00 - 11/30 102 G01N 21/84 - 21/958 C03B 13/12 ────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A-8-114432 (JP, A) JP-A-6-347410 (JP, A) JP-A-6-148100 (JP, A) JP-A-6-148100 18432 (JP, A) (58) Field surveyed (Int. Cl. 7 , DB name) G01B 11/00-11/30 102 G01N 21/84-21/958 C03B 13/12

Claims (5)

(57)【特許請求の範囲】(57) [Claims] 【請求項1】連続して搬送されるリボン状の網または線
入りガラスに光源からの光を投光し、網または線入りガ
ラスを挟んで光源とは反対側に設けた複数台のエリアカ
メラにより網または線入りガラスを撮像し、得られた濃
淡の信号により網または線入りガラスの検査をする方法
において、撮像した画像データの二値化処理後、濃信号
部分の全周囲を一画素分消去する収縮処理を所定回数行
うことにより網または線に相当する濃信号のみを消去さ
せ、残った濃信号を欠陥とし、画像データに基づいて、
その大きさ、位置等を判別することを特徴とする網入り
ガラスの欠陥検出方法。
1. A plurality of area cameras provided by projecting light from a light source onto a ribbon-shaped net or lined glass which is continuously conveyed, and provided on the opposite side of the light source with the net or lined glass interposed therebetween. In the method of imaging the net or the lined glass by using the obtained grayscale signal and inspecting the net or the lined glass, after binarizing the captured image data, the entire periphery of the dark signal portion is equivalent to one pixel. By performing a predetermined number of contraction processes for erasing, only the dark signal corresponding to the screen or line is erased, and the remaining dark signal is regarded as a defect, and based on the image data,
A method for detecting a defect in a netted glass, comprising determining its size, position, and the like.
【請求項2】前記エリアカメラに代えて、搬送される網
または線入りガラスの幅方向に走査するラインカメラを
複数台配置し、該ラインカメラで網または線入りガラス
の搬送速度に同期させて取り込んだ信号を記憶保持させ
て2次元の画像データとし、濃信号部分の全周囲を一画
素分消去する収縮処理を所定回数行うことにより網また
は線に相当する濃信号のみを消去させ、残った濃信号を
欠陥とし、画像データに基づいて、その大きさ、位置等
を判別することを特徴とする網入りガラスの欠陥検出方
法。
2. In place of the area camera, a plurality of line cameras for scanning in the width direction of the net or lined glass to be conveyed are arranged, and the line cameras are synchronized with the conveying speed of the net or lined glass. The captured signal is stored and retained to form two-dimensional image data, and a predetermined number of times of contraction processing for erasing the entire periphery of the dark signal portion by one pixel eliminates only the dark signal corresponding to a mesh or a line, and the remaining signal remains. A method for detecting a defect in a screened glass, wherein a size, a position, and the like of the dark signal are determined based on image data.
【請求項3】淡信号部分を収縮させる画像処理装置を用
いる場合に、請求項1または2記載の網入りガラスの欠
陥検出方法の画像信号を二値化処理した後、網線の部分
と欠陥の部分を濃信号として収縮処理する方法に代え
て、収縮処理する前にネガポジ反転して網線の部分と欠
陥の部分を淡信号とし、ガラスの正常部分を濃信号とし
た後に、収縮処理を行って網線の部分と欠陥の部分の淡
信号部分を収縮させることを特徴とする網入りガラスの
欠陥検出方法。
3. An image processing apparatus for contracting a light signal portion is used.
If there is, the lack of the netted glass according to claim 1 or 2
After binarizing the image signal of the defect detection method,
Instead of shrinking the defective part as a dark signal
Before shrinking, reverse the negative / positive
The light spot signal is used for the inflected part and the dark signal is used for the normal part of the glass.
After the shrinkage process, the shaded area
Shrink the signal part
Defect detection method.
【請求項4】連続して搬送されるリボン状の網または線
入りガラスに光を投光する光源と、網または線入りガラ
スを挟んで光源とは反対側に設け、網または線入りガラ
スを撮像する複数台のエリアカメラと、該エリアカメラ
により得られた画像データを処理する画像処理手段と、
該画像データにより欠陥部を判別する検出制御手段から
なる網または線入りガラスの欠陥検出装置において、前
記検出制御手段は画像データを二値化処理後、濃信号部
分の全周囲を一画素分消去する収縮処理を所定回数行う
ことにより網または線に相当する濃信号のみを消去させ
る機能、および残った濃信号を欠陥とし、画像データに
基づいてその大きさ、位置等を判別する機能を有するよ
うに構成したことを特徴とする網入りガラスの欠陥検出
装置。
4. A light source for projecting light onto a ribbon-shaped net or lined glass that is continuously conveyed, and a light source that is provided on the opposite side of the light source with the net or lined glass interposed therebetween. A plurality of area cameras for imaging, and image processing means for processing image data obtained by the area camera,
In a net or line defect detecting apparatus for detecting a defect in a net or a lined glass, the detecting control means erases the entire periphery of the dark signal portion by one pixel after binarizing the image data. A function of deleting only a dark signal corresponding to a screen or a line by performing a predetermined number of times of contraction processing, and a function of determining a size, a position, and the like of the remaining dark signal based on image data as a defect. A defect detection apparatus for a glass-screened glass, comprising:
【請求項5】前記エリアカメラに代えて、搬送される網
または線入りガラスの幅方向に走査するラインカメラを
複数台配置し、該ラインカメラで網または線入りガラス
の搬送速度に同期させて取り込んだ信号を記憶保持させ
て2次元の画像データとする記憶装置とからなることを
特徴とする請求項4記載の網入りガラスの欠陥検出装
置。
5. In place of the area camera, a plurality of line cameras for scanning in the width direction of the net or lined glass to be conveyed are arranged, and the line cameras are synchronized with the conveying speed of the net or lined glass. 5. The apparatus according to claim 4, further comprising a storage device that stores and retains the captured signal to generate two-dimensional image data.
JP16826697A 1997-06-25 1997-06-25 Defect detection method and apparatus for netted glass Expired - Fee Related JP3332208B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP16826697A JP3332208B2 (en) 1997-06-25 1997-06-25 Defect detection method and apparatus for netted glass

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP16826697A JP3332208B2 (en) 1997-06-25 1997-06-25 Defect detection method and apparatus for netted glass

Publications (2)

Publication Number Publication Date
JPH1114334A JPH1114334A (en) 1999-01-22
JP3332208B2 true JP3332208B2 (en) 2002-10-07

Family

ID=15864835

Family Applications (1)

Application Number Title Priority Date Filing Date
JP16826697A Expired - Fee Related JP3332208B2 (en) 1997-06-25 1997-06-25 Defect detection method and apparatus for netted glass

Country Status (1)

Country Link
JP (1) JP3332208B2 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2020163054A1 (en) * 2019-02-06 2020-08-13 Corning Incorporated Methods of processing a viscous ribbon

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4779790B2 (en) * 2006-04-26 2011-09-28 セントラル硝子株式会社 Glass plate surface identification method and apparatus
JP5207456B2 (en) * 2008-06-20 2013-06-12 富士フイルム株式会社 Defect detection apparatus and method
JP5136277B2 (en) * 2008-08-18 2013-02-06 セントラル硝子株式会社 Defect detection method for netted or lined glass
CN105319219B (en) * 2014-07-23 2019-05-14 塞米西斯科株式会社 Bad inspection system and method
JP6776815B2 (en) * 2016-10-31 2020-10-28 大日本印刷株式会社 Laminate inspection equipment, methods and programs
CN108195848A (en) * 2018-01-16 2018-06-22 深圳精创视觉科技有限公司 The full-scale warpage flatness detecting device of glass and detection method
CN112834522A (en) * 2020-12-30 2021-05-25 深圳南玻应用技术有限公司 Visual detection feedback control system
CN115356355B (en) * 2022-10-21 2022-12-27 沃卡姆(山东)真空玻璃科技有限公司 Automatic detection blanking conveying line and detection method for vacuum laminated glass

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2020163054A1 (en) * 2019-02-06 2020-08-13 Corning Incorporated Methods of processing a viscous ribbon
CN113614047A (en) * 2019-02-06 2021-11-05 康宁公司 Method of treating adhesive tape
CN113614047B (en) * 2019-02-06 2023-09-19 康宁公司 Method for treating adhesive tape
US11878928B2 (en) 2019-02-06 2024-01-23 Corning Incorporated Methods of processing a viscous ribbon

Also Published As

Publication number Publication date
JPH1114334A (en) 1999-01-22

Similar Documents

Publication Publication Date Title
TWI484161B (en) Defect inspection system and filming device for defect inspection, image processing device for defect inspection, image processing program for defect inspection, recording media, and image processing method for defect inspection used therein
JPH04220551A (en) Method and apparatus for inspecting flaw of transparent body
JP3332208B2 (en) Defect detection method and apparatus for netted glass
JPWO2007132925A1 (en) Surface inspection device
JPH11211442A (en) Method and device for detecting defect of object surface
JPH08101130A (en) Surface flaw inspecting device
JP3695120B2 (en) Defect inspection method
JP5136277B2 (en) Defect detection method for netted or lined glass
JPH0283438A (en) Apparatus for evaluating deposit or the like on surface of transparent plate
JP3283356B2 (en) Surface inspection method
JP2710527B2 (en) Inspection equipment for periodic patterns
JP2005083906A (en) Defect detector
JP4956077B2 (en) Defect inspection apparatus and defect inspection method
JPH0711414B2 (en) Glass bottle uneven thickness inspection method
JP2003329428A (en) Device and method for surface inspection
JP2002074332A (en) Method and device for detecting sealing defect of packing material and packing body
JPH0763699A (en) Flaw inspection apparatus
JP3202330B2 (en) Defect inspection equipment
JP3025562B2 (en) Surface condition inspection method using bright and dark illumination
JPH11132966A (en) Meandering following apparatus for flaw inspection apparatus
JPS61126455A (en) Image processing method
JP2941107B2 (en) Work surface inspection method
JP2007003389A (en) Inspection device and inspection method
JP3284463B2 (en) Bottle body defect detection method
JPH07301609A (en) Defect inspection method

Legal Events

Date Code Title Description
FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20080726

Year of fee payment: 6

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20080726

Year of fee payment: 6

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20090726

Year of fee payment: 7

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20090726

Year of fee payment: 7

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20090726

Year of fee payment: 7

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20100726

Year of fee payment: 8

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20100726

Year of fee payment: 8

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20110726

Year of fee payment: 9

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20110726

Year of fee payment: 9

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20120726

Year of fee payment: 10

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20120726

Year of fee payment: 10

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20130726

Year of fee payment: 11

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20130726

Year of fee payment: 11

LAPS Cancellation because of no payment of annual fees