JPH09287920A - Method for evaluating shape of object to be measured, shape of bubble inside glass and degree of defect of glass - Google Patents
Method for evaluating shape of object to be measured, shape of bubble inside glass and degree of defect of glassInfo
- Publication number
- JPH09287920A JPH09287920A JP8101606A JP10160696A JPH09287920A JP H09287920 A JPH09287920 A JP H09287920A JP 8101606 A JP8101606 A JP 8101606A JP 10160696 A JP10160696 A JP 10160696A JP H09287920 A JPH09287920 A JP H09287920A
- Authority
- JP
- Japan
- Prior art keywords
- measured
- glass
- light
- histogram
- evaluating
- 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.)
- Pending
Links
Landscapes
- Length Measuring Devices By Optical Means (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
Description
【０００１】[0001]
【発明の属する技術分野】本発明は、透明体に存在する
被測定物の形状を評価する方法に関し、特にガラスの内
部に発生する泡の大きさを測定して評価し、さらにこの
泡によるガラスの欠陥度合を評価する方法に関する。BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for evaluating the shape of an object to be measured which is present in a transparent body, and in particular, the size of bubbles generated inside the glass is measured and evaluated. The present invention relates to a method of evaluating the degree of defect of.
【０００２】[0002]
【従来の技術】従来、ガラスの泡の径の測定は、人によ
る官能検査、ルーペなどによって行われてきた。しか
し、従来の官能検査では検査員による個人差や疲労具合
等によるばらつきが大きいため、測定の信頼性に劣るも
のであった。またルーペによる測定は労力と時間がかか
るうえ、個人差や読みとり誤差などの問題点もあった。2. Description of the Related Art Heretofore, the diameter of glass bubbles has been measured by a sensory test by a person, a magnifying glass, or the like. However, in the conventional sensory test, the reliability of the measurement was poor because of large variations due to individual differences among the inspectors and the degree of fatigue. In addition, measurement with a loupe takes time and labor, and there are problems such as individual differences and reading errors.
【０００３】そこで、人の個人的な能力に頼る方法では
なく、画像処理を用いた二値化による方法を用いて泡の
大きさを測定する方法が提案されてきた。例えば、特開
昭６１−１２３９８５号や特開平７−２７０１３６号に
は、固定したしきい値による方法、判別分析、モード法
などと呼ばれる統計手段によりしきい値を決めて二値化
する方法が開示されており、それらを改良して照明の変
化をモニターすることを特徴とした二値化の方法も提案
されていた。Therefore, there has been proposed a method of measuring the size of a bubble by using a binarization method using image processing, instead of relying on a person's personal ability. For example, in Japanese Patent Laid-Open No. 61-123985 and Japanese Patent Laid-Open No. 7-270136, there is a method of binarizing a threshold value by a fixed threshold value method, a discriminant analysis method, or a statistical method called a modal method. It has been disclosed, and a binarization method characterized by improving them and monitoring a change in illumination has also been proposed.
【０００４】[0004]
【発明が解決しようとする課題】しかし、これらの従来
法は照明光の変動などの対策はなされているが、画像全
面に対して被測定物が小さく、背景部分に対する被測定
物の濃淡のコントラストが低い被測定物を測定しようと
する場合には、十分な測定精度が得られなかった。すな
わち、上記のように被測定物が小さい場合には、被測定
物と背景との境界がはっきりしないため、被測定物の形
状の認識がうまくいかず、被測定物のコントラストのば
らつきによって測定のばらつきが大きくなってしまう。However, although these conventional methods have taken measures against fluctuations in illumination light, etc., the object to be measured is small over the entire surface of the image, and the contrast of the density of the object to be measured with respect to the background portion is small. When attempting to measure an object to be measured with a low value, sufficient measurement accuracy could not be obtained. That is, when the object to be measured is small as described above, the boundary between the object to be measured and the background is not clear, so that the shape of the object to be measured cannot be recognized well, and the difference in the contrast of the object to be measured causes the measurement to fail. The variation becomes large.
【０００５】このような欠点を解消する手段として、解
像度を上げたりカメラの感度や安定性を上げて従来技術
が適用できるようにすることが考えられる。しかしなが
ら、時間や演算部分のパワーをより必要とすることにな
り、あまり実用的ではない。As a means for solving such a drawback, it is conceivable to increase the resolution and the sensitivity and stability of the camera so that the prior art can be applied. However, it requires more time and power of the calculation part, which is not practical.
【０００６】また、別の従来から知られている方法とし
て、ｐ−タイル法と呼ばれるヒストグラムの最大値、あ
るいは最小値より特定の面積に相当する画素数となると
ころでしきい値を決める方法がある。この方法は、例え
ば特開昭６３−４３７３号等に開示されており、二値化
しきい値を変えながら特定の特徴量の変化の具合により
しきい値を決める二値化の方法である。As another conventionally known method, there is a method called a p-tile method for determining a threshold value at the number of pixels corresponding to a specific area from the maximum value or the minimum value of a histogram. . This method is disclosed in, for example, Japanese Patent Laid-Open No. 63-4373, and is a binarization method in which the threshold value is determined according to the degree of change in a specific feature value while changing the binarization threshold value.
【０００７】しかし、これらの方法は、パターンの検査
などの被測定物の形状があらかじめわかっている場合は
有効である。しかし、測定する被測定物の大きさや形や
濃淡のコントラストが被測定物に応じてまちまちで予想
できない場合には、ある程度のしきい値を予め設定する
ことが困難であるため、ガラス内の泡の大きさ測定など
には適用が困難であった。However, these methods are effective when the shape of the object to be measured such as pattern inspection is known in advance. However, if the size and shape of the measured object to be measured and the contrast of shades cannot be predicted depending on the measured object, it is difficult to set a certain threshold value in advance, and therefore bubbles in the glass It was difficult to apply it to the measurement of the size.
【０００８】本発明の目的は、従来技術が有していた前
述の欠点を解消することにあり、従来知られていなかっ
た透明体にある被測定物の形状、ガラス内部にある泡の
形状およびガラスの欠陥の度合の評価方法を新規に提供
することにある。The object of the present invention is to eliminate the above-mentioned drawbacks of the prior art. The shape of the object to be measured in a transparent body, the shape of bubbles inside the glass, and It is to provide a new method for evaluating the degree of glass defects.
【０００９】[0009]
【課題を解決するための手段】本発明は、上記課題に鑑
みてなされたものであり、透明体に光を照射し、透明体
を介した光を撮像して透明体にある被測定物の形状を評
価する方法において、透明体を撮像した領域のうち被測
定物を含む被測定物の近傍を測定有効領域に定め、該有
効領域における透明体を介した光の濃度のヒストグラム
を作製し、該ヒストグラムのデータに基づいて被測定物
の評価のしきい値を決定することを特徴とする透明体に
ある被測定物の形状の評価方法を提供するものである。SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and irradiates a transparent body with light, images the light passing through the transparent body, and measures an object to be measured on the transparent body. In the method of evaluating the shape, the vicinity of the object to be measured including the object to be measured is defined as the measurement effective area in the area where the transparent body is imaged, and a histogram of the density of light through the transparent body in the effective area is created, The present invention provides a method for evaluating the shape of an object to be measured on a transparent body, which comprises determining a threshold value for evaluating the object to be measured based on the data of the histogram.
【００１０】また、本発明は、ガラスに光を照射し、ガ
ラスを介した光を撮像してガラス内部にある泡の形状を
評価する方法において、ガラスを撮像した領域のうち泡
を含む泡の近傍を測定有効領域に定め、該有効領域にお
けるガラスを介した光の濃度のヒストグラムを作製し、
該ヒストグラムのデータに基づいてしきい値を決定して
撮像した光の情報をデジタル化し、該デジタル化した情
報から泡の大きさを求めることを特徴とするガラス内部
にある泡の形状の評価方法を提供するものである。Further, according to the present invention, in a method of irradiating a glass with light and imaging the light passing through the glass to evaluate the shape of the foam inside the glass, the foam containing the foam in the area where the glass is imaged is evaluated. Determine the vicinity to the measurement effective area, create a histogram of the density of light through the glass in the effective area,
A method for evaluating the shape of a bubble inside glass, characterized in that a threshold value is determined based on the data of the histogram, information of the imaged light is digitized, and the size of the bubble is obtained from the digitized information. Is provided.
【００１１】さらにまた、本発明は、ガラスに光を照射
し、ガラスを介した光を撮像してガラスの欠陥の度合を
評価する方法において、ガラスを撮像した領域のうちガ
ラス内部にある泡を含む泡の近傍を測定有効領域に定
め、該有効領域におけるガラスを介した光の濃度のヒス
トグラムを作製し、該ヒストグラムのデータに基づいて
しきい値を決定して撮像した光の情報をデジタル化し、
該デジタル化した情報から泡の大きさを求め、ヒストグ
ラムより求めた泡部分の濃度に相当する値とを組み合わ
せてあるいはそれらを変数とする数式により泡の欠点と
しての強さを評価することを特徴とするするガラスの欠
陥の度合の評価方法を提供するものである。Furthermore, the present invention is a method of irradiating a glass with light and imaging the light passing through the glass to evaluate the degree of defects in the glass. The vicinity of the containing bubble is set as the measurement effective area, a histogram of the density of light through the glass in the effective area is created, the threshold value is determined based on the data of the histogram, and the imaged light information is digitized. ,
The size of a bubble is obtained from the digitized information, and the strength as a defect of the bubble is evaluated in combination with a value corresponding to the concentration of the bubble portion obtained from the histogram or by a mathematical expression using these as variables. The present invention provides a method for evaluating the degree of glass defects.
【００１２】[0012]
【発明の実施の形態】以下に図面に基づいて本発明を詳
細に説明する。図１は本発明を用いた被測定物の評価方
法の一例として、ガラス内部の泡の大きさを測定する方
法を説明する概略斜視図である。１は撮像装置であるＣ
ＣＤカメラとレンズ、５はその信号を演算、処理する装
置、２はガラス、３は被測定物である泡である。４は白
色拡散光源であり、泡は明視野で撮像される。DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below in detail with reference to the drawings. FIG. 1 is a schematic perspective view illustrating a method for measuring the size of bubbles inside glass as an example of the method for evaluating an object to be measured using the present invention. 1 is an image pickup device C
The CD camera and lens, 5 are devices for calculating and processing the signals, 2 is glass, and 3 is a bubble which is an object to be measured. 4 is a white diffuse light source, and the bubbles are imaged in a bright field.
【００１３】この条件では、泡の輪郭部分（周縁部分）
は背景よりも黒く、低い濃度で撮像されるが、画像の濃
度値を反転させて泡を高い濃度の部分として認識、測定
することももちろん可能である。Under this condition, the contour portion (peripheral portion) of the bubble
Is blacker than the background and is imaged at a low density, but it is of course possible to invert the density value of the image and recognize and measure bubbles as a high density portion.
【００１４】本例で用いた光学的諸条件は、レンズは焦
点距離１５ｍｍ、絞りはＦ１６であり、撮像面から被測
定物までの距離は９５ｍｍ、測定面においては１画素の
大きさが約３５μｍである。また、画像は５１２×４８
０画素あり、被測定物の泡は０．１〜１ｍｍ程度の大き
さである。The optical conditions used in this example are that the lens has a focal length of 15 mm, the diaphragm is F16, the distance from the imaging surface to the object to be measured is 95 mm, and the size of one pixel on the measuring surface is about 35 μm. Is. Also, the image is 512 x 48
There are 0 pixels, and the bubble of the measured object has a size of about 0.1 to 1 mm.
【００１５】この条件でガラス内部の泡を撮像した場合
に得られる画像を模式的に示したのが図２、図３であ
る。泡には様々な形状のものがあり、球に近い形状のも
のは図２のようにしっかりした輪郭を持つが、肉厚方向
に薄く、偏平した泡は図３のように輪郭における濃度差
の小さい、濃淡のコントラストの低い画像が得られる。
この２例は代表的なものを模式的に示したものである
が、実際の被測定物である泡は、形状、濃淡のコントラ
ストとも様々なものが存在している。FIGS. 2 and 3 schematically show images obtained when the bubbles inside the glass are imaged under these conditions. There are various shapes of bubbles, and those with a shape close to a sphere have a firm contour as shown in Fig. 2, but bubbles that are thin in the wall thickness direction and flat are of the density difference in the contour as shown in Fig. 3. A small image with low contrast of light and shade is obtained.
These two examples schematically show typical ones, but there are various bubbles, which are actually measured objects, in terms of shape and contrast of light and shade.
【００１６】図４は得られる画像全体の模式図である。
撮像領域（画像）全体のうち、６はヒストグラムを求め
るための測定有効領域（以下ウインドウという）を示し
ている。本例では、ウインドウの大きさを８０×８０画
素の大きさとしている。本発明における１つの例として
は、画像中のウインドウの位置を固定し、被測定物であ
る泡の像がウインドウ内に入るようにカメラを移動させ
る。ウインドウの大きさは現在のカメラの位置合わせ手
段で、被測定物の泡が完全にウインドウ内に入りきる最
小の大きさとする。別の例としては、画像中の最小値の
位置の近傍を欠点であるとして、最小値を中心とした８
０×８０画素の範囲をウインドウとすることがあげられ
る。FIG. 4 is a schematic diagram of the entire image obtained.
Of the entire imaging area (image), 6 indicates a measurement effective area (hereinafter referred to as a window) for obtaining a histogram. In this example, the size of the window is 80 × 80 pixels. As one example of the present invention, the position of the window in the image is fixed, and the camera is moved so that the image of the bubble, which is the object to be measured, enters the window. The size of the window is the minimum size of the current camera alignment means that allows the bubbles of the object to be measured to completely enter the window. As another example, assuming that the neighborhood of the position of the minimum value in the image is a defect, the
It is possible to use a range of 0 × 80 pixels as a window.
【００１７】次に、上記例に関する本発明の方法を、図
５に示す流れ図を用いて説明する。まず、泡を含む部分
を透過した白色拡散光源４からの光を、撮像装置１で撮
像する。こうして撮像された光の画像を画像処理装置に
入力し、撮像装置１で撮像された撮像領域のうち、泡を
含む泡の近傍をウインドウに定める（５１）。次に、泡
の画像データをコンピュータ装置等の画像処理装置に入
力し、泡を含むウィンドウ内の光の濃度ヒストグラムを
作製する（５２）。The method of the present invention for the above example will now be described with reference to the flow chart shown in FIG. First, the image pickup device 1 takes an image of the light from the white diffused light source 4 that has passed through a portion including bubbles. The image of the light thus captured is input to the image processing device, and the vicinity of bubbles including bubbles in the imaging region imaged by the imaging device 1 is set as a window (51). Next, the image data of the bubbles is input to an image processing device such as a computer, and a density histogram of light in a window containing bubbles is created (52).
【００１８】こうして得られたヒストグラムのデータよ
り、最大頻度を持つ濃度値Ｍ1 、最小濃度値Ｍ2 、判別
分析によるしきい値Ｔを決定する（５３）。図２のよう
な輪郭のはっきりした泡からは図６のようなヒストグラ
ムが得られ、図３のような輪郭の薄い泡からは図７のよ
うなヒストグラムが得られる。図６、７において、横軸
は濃度を表す値で、図６および図７の７、８、９は各Ｍ
2 、Ｔ、Ｍ1 の値を示している。From the thus obtained histogram data, the density value M 1 having the maximum frequency, the minimum density value M 2 , and the threshold value T by discriminant analysis are determined (53). A bubble as shown in FIG. 6 is obtained from a bubble having a clear outline as shown in FIG. 2, and a histogram as shown in FIG. 7 is obtained as a bubble with a thin outline as shown in FIG. In FIGS. 6 and 7, the horizontal axis is a value representing the density, and 7, 8 and 9 in FIGS.
The values of 2 , T and M 1 are shown.
【００１９】上記の判別分析の手法によるしきい値Ｔ
は、以下のようにして得られる。ヒストグラムをあるし
きい値で２つのクラスに分ける場合、２つのクラスの平
均値の分散と各クラスの分散の比が最大となるようにし
きい値を求める。The threshold value T obtained by the above discriminant analysis method
Is obtained as follows. When the histogram is divided into two classes with a certain threshold value, the threshold value is obtained so that the ratio of the variance of the average value of the two classes and the variance of each class is maximized.
【００２０】クラス１の画素数をｗ1 、濃度の平均値を
Ｍa 、分散をσ1 、クラス２の画素数をｗ2 、濃度の平
均値をＭb 、分散をσ2 、全平均値をＭt とすると、ク
ラス内分散は、（σw ）2 ＝ｗ1 ×（σ1 ）2 ＋ｗ2 ×
（σ2 ）2 であり、クラス間分散は、（σb ）2 ＝ｗ1
×（Ｍa −Ｍt ）2 ＋ｗ2 ×（Ｍb −Ｍt ）2 となる。
しきい値を少しずつ変化させ、そのつどσb ／σw を求
め、σb ／σw が最大となるときのしきい値を判別分析
の手法によるしきい値とする。The number of pixels of class 1 is w 1 , the average value of density is M a , the variance is σ 1 , the number of pixels of class 2 is w 2 , the average value of density is M b , the variance is σ 2 , and the total average value is Is M t , the within-class variance is (σ w ) 2 = w 1 × (σ 1 ) 2 + w 2 ×
(Σ 2 ) 2 and the interclass variance is (σ b ) 2 = w 1
× a (M a -M t) 2 + w 2 × (M b -M t) 2.
The gradually changed threshold, determined in each case σ b / σ w, σ b / σ w is the threshold according to the threshold of the discrimination analysis method when the maximum.
【００２１】ところで、ヒストグラムの最大頻度を持つ
濃度値は背景の濃度を代表しており、一般に判別分析の
手法により求めた値は被測定物の濃度により変化する値
である。そのため、両者の関係は背景に対する被測定物
の濃淡のコントラストとして作用する。すなわち、被測
定物と背景との濃淡のコントラストが高い場合は、判別
分析などの手法によるしきい値を用いた被測定物の形状
の認識はうまくいく。By the way, the density value having the maximum frequency in the histogram represents the density of the background, and generally, the value obtained by the discriminant analysis method is a value which changes depending on the density of the object to be measured. Therefore, the relationship between the two acts as a contrast of the contrast of the object to be measured with respect to the background. That is, when the contrast between the object and the background is high, the recognition of the shape of the object using the threshold value by a method such as discriminant analysis is successful.
【００２２】しかし、濃淡のコントラストが低い場合
は、両者の値は一致するかあるいはきわめて近くなり、
判別分析などの手法による値は安定感を失う。そこで、
ヒストグラムの最大濃度値または最小濃度値と、最大頻
度を持つ濃度値の差を変数とする数式を用いて相対的に
しきい値を決定し、被測定物の形状の認識を行った方
が、安定な解が得られる。However, when the contrast of light and shade is low, the values of both are the same or very close to each other,
Values obtained by methods such as discriminant analysis lose the sense of stability. Therefore,
It is more stable if the shape of the DUT is recognized by relatively determining the threshold value using a mathematical expression that uses the difference between the maximum or minimum density value in the histogram and the density value with the maximum frequency as a variable. Can be obtained.
【００２３】そこで、上記例のように、画像全面に対し
て被測定物が小さく、背景部分に対する被測定物の濃淡
のコントラストが低く、ばらつきがあり、被測定物の大
きさや形が予想できない場合には、上記のヒストグラム
の最大濃度値（Ｍ1 ）または最小濃度値（Ｍ2 ）と最大
頻度を持つ濃度値（Ｔ）の差を変数とする数式を用い、
被測定物に応じて相対的にしきい値を決定する手法を用
いることによって、安定した被測定物の形状の認識が可
能となり、その画像を用いた測定も高精度に行いうる。Therefore, as in the above example, when the object to be measured is small over the entire image, the contrast of the density of the object to be measured with respect to the background is low, and there are variations, and the size and shape of the object to be measured cannot be predicted. Is used as a variable, the mathematical expression in which the difference between the maximum density value (M 1 ) or the minimum density value (M 2 ) of the histogram and the density value (T) having the maximum frequency is used as a variable,
By using the method of relatively determining the threshold value according to the object to be measured, the shape of the object to be measured can be stably recognized, and the measurement using the image can be performed with high accuracy.
【００２４】以下に、Ｍ1 −Ｔ（＝ａとする）の値によ
り場合分けする例を示す。例えばＭ1 −Ｔの値が大きい
場合（図２のような輪郭のはっきりした泡：図６に相
当）しきい値はＴよりわずかに小さい値とし、Ｍ1 −Ｔ
の値が小さい場合（図３のような輪郭の薄い泡：図７に
相当）しきい値はＴよりかなり小さい値とし、Ｍ1 −Ｔ
の値が中間の場合（図６と図７の中間状態に相当）しき
い値は上記の２つのケースの中間程度の値とする。具体
的には、本例で測定しようとしている被測定物である泡
の測定に適当な値として、以下の表１のようにしきい値
を定めるが、被測定物の母集団に応じて、しきい値の決
め方は適宜選択される。The following is an example of case classification according to the value of M 1 -T (= a). For example, when the value of M 1 -T is large (a bubble having a clear contour as shown in FIG. 2: equivalent to FIG. 6), the threshold value is set to a value slightly smaller than T, and M 1 -T
When the value of is small (a bubble with a thin contour as shown in FIG. 3: corresponding to FIG. 7), the threshold value is set to be much smaller than T, and M 1 −T
When the value of is intermediate (corresponding to the intermediate state in FIGS. 6 and 7), the threshold value is set to an intermediate value between the above two cases. Specifically, a threshold value is set as shown in Table 1 below as a value suitable for measurement of bubbles, which are the objects to be measured in this example, but depending on the population of the objects to be measured, The method of determining the threshold value is appropriately selected.
【００２５】[0025]
【表１】 [Table 1]
【００２６】このようにして決定したしきい値を用い、
しきい値以上の出力の画素部分を「１」（泡の周縁部分
でない背景部分や泡の中心領域等に相当）とし、しきい
値未満の画素部分を「０」（泡の周縁部分に相当）とし
て二値化処理をする（５５）。その後、測定対象以下の
小さな点欠陥を除去してノイズやほこりの影響を除去
し、さらに泡全体の情報にするために周縁部分にあわせ
て泡の中心部分を埋める（出力情報を１→０とする）こ
とにより泡の形状を塗りつぶし、泡部分に相当する画素
の数を数えることにより最大径（Ｌ1 ）および面積
（Ｓ）を求める。Using the threshold value thus determined,
The pixel portion of the output that is equal to or more than the threshold value is set to "1" (corresponding to the background portion which is not the peripheral portion of the bubble, the central region of the bubble, etc.), and the pixel portion below the threshold value is "0" (corresponding to the peripheral portion of the bubble () Is binarized (55). After that, small point defects below the object to be measured are removed to remove the influence of noise and dust, and the central portion of the bubble is filled in accordance with the peripheral portion in order to obtain information on the entire bubble (output information is 1 → 0. By applying the above), the shape of the bubble is filled, and the maximum diameter (L 1 ) and the area (S) are obtained by counting the number of pixels corresponding to the bubble portion.
【００２７】Ｌ1 と垂直方向の径（最小径Ｌ2 ）は泡を
楕円として近似し、Ｌ2 ＝Ｓ／Ｌ1×４／πで求める。
泡径を（Ｌ1 ＋Ｌ2 ）／２で求めることもできる。な
お、図８のような細長い泡については図中の１０に相当
する両突端部分が量子化により測定されないため、その
補正として、縦横比に応じた補正を行う。この補正は、
例えばＬ1 ／Ｌ2 が２未満のものは細長い泡でないと判
断し、Ｌ1 、Ｌ2 に補正は加えずそのまま用い、Ｌ1 ／
Ｌ2 が２以上のものを細長い泡と判断し、Ｌ1 （補正
値）＝Ｌ1 ＋Ｌ1 ／Ｌ2 とする等が例示できる。この場
合、Ｌ1 ／Ｌ2 は泡の両先端の1 画素分の量子化誤差の
補正分に相当する。こうして、泡の大きさが測定される
（５６）。The diameter in the direction perpendicular to L 1 (minimum diameter L 2 ) is approximated by using a bubble as an ellipse, and is calculated by L 2 = S / L 1 × 4 / π.
The bubble diameter can also be determined by (L 1 + L 2 ) / 2. It should be noted that, in the case of a long and narrow bubble as shown in FIG. 8, both tip portions corresponding to 10 in the figure are not measured by quantization, so correction is made according to the aspect ratio. This correction is
For example, if L 1 / L 2 is less than 2, it is determined that the bubbles are not elongated bubbles, and L 1 and L 2 are used as they are without correction and L 1 / L 2
One having L 2 of 2 or more is determined to be an elongated bubble, and L 1 (correction value) = L 1 + L 1 / L 2 can be exemplified. In this case, L 1 / L 2 corresponds to the correction amount of the quantization error for one pixel at both ends of the bubble. Thus, the bubble size is measured (56).
【００２８】こうして得られた値に関し、最大径、縦横
の平均径について顕微鏡による実測定との比較を行っ
た。両者の差は最大径０．８ｍｍ以下の泡５０個に対し
て、標準偏差で最大径３７μｍ、平均径２６μｍであ
り、本発明による方法は、顕微鏡により時間をかけて行
った実測定に十分に見合うものであった。With respect to the values thus obtained, the maximum diameter and the average length and width were compared with actual measurement by a microscope. The difference between the two is 50 μm with a maximum diameter of 0.8 mm or less with a standard deviation of a maximum diameter of 37 μm and an average diameter of 26 μm, and the method according to the present invention is sufficient for actual measurement performed with a microscope over time. It was worth it.
【００２９】また、泡の評価の例として、二値化に用い
たしきい値と測定した面積の積を用いて、泡の認識度合
を評価した（５７）。これは、しきい値が泡の存在を認
識できる濃度レベルを示す値であり、面積は泡の存在を
認識できる大きさを示す値であることから、あるレベル
で泡が認識されるものであるかそうでないかの指標にな
る。他に、Ｍ1 −Ｍ2 と面積の積によって評価する等、
多くの組み合わせにより評価指標を決めることができ、
評価の目的に応じて最適な評価指標を用いることが可能
である。As an example of bubble evaluation, the product of the threshold value used for binarization and the measured area was used to evaluate the degree of bubble recognition (57). This is because the threshold value is a value indicating the concentration level at which the presence of bubbles can be recognized, and the area is a value indicating the size at which the presence of bubbles can be recognized, so that bubbles are recognized at a certain level. It will be an indicator of whether or not. In addition, evaluation by the product of M 1 -M 2 and the area,
You can decide the evaluation index by many combinations,
It is possible to use the optimum evaluation index according to the purpose of evaluation.
【００３０】ちなみに、従来用いられている手段である
Ｔをしきい値とした場合、テストした泡の２５％以上の
形状の認識ができず、測定不能であった。By the way, when T, which is a conventionally used means, was used as a threshold value, the shape of 25% or more of the bubbles tested could not be recognized and measurement was impossible.
【００３１】本発明において透明体としては、建築用や
車両用の窓に用いられるガラス板、ブラウン管用ガラ
ス、液晶基板用のガラス板等のガラス物品、合成樹脂製
の透明樹脂板等があげられる。これら、透明体に存在す
る被測定物としては、上記例にある泡のほか、透明体の
表面にある傷、しみや汚れ等が例示できる。In the present invention, examples of the transparent body include glass plates used for windows for buildings and vehicles, glass for cathode ray tubes, glass articles such as glass plates for liquid crystal substrates, and transparent resin plates made of synthetic resin. . Examples of these objects to be measured that are present in the transparent body include the bubbles in the above examples, as well as scratches, stains and stains on the surface of the transparent body.
【００３２】上記の被測定物の評価項目としては、上記
例における被測定物の大きさのほか、被測定物の存在の
有無、被測定物の存在が許容範囲であるかどうか等が例
示できる。As the evaluation item of the above-mentioned object to be measured, in addition to the size of the object to be measured in the above example, the presence or absence of the object to be measured, whether the existence of the object to be measured is within an allowable range, etc. can be exemplified. .
【００３３】またしきい値の決定方法も、変数を最大濃
度値あるいは最小濃度値と、最大頻度を持つ値に限るも
のではなく、評価の目的に応じてヒストグラムから求め
られる他のパラメータを適宜用いることもできる。The method of determining the threshold value is not limited to the variable having the maximum density value or the minimum density value and the value having the maximum frequency, and other parameters obtained from the histogram according to the purpose of evaluation are appropriately used. You can also
【００３４】[0034]
【発明の効果】本発明によれば、被測定物を含む被測定
物の近傍に適当な大きさのウインドウをもうけ、そのウ
インドウ内の濃度ヒストグラムを作製しているので、濃
度ヒストグラムの中の背景のばらつきによる影響を極力
小さくすることができ、そのヒストグラムデータの特徴
よりその後の処理のしきい値を決定している。そのた
め、画像全面に対して被測定物が小さい場合、背景部分
に対する被測定物の濃淡のコントラストが小さくい場
合、さらには被測定物の１つ１つでそれらの特徴がまち
まちで、そのばらつきが大きい場合や、これらの組み合
わさった場合等、被測定物の大きさや形が予想できない
場合においても、安定した被測定物の形状の認識が可能
となり、その画像を用いた測定も高精度に行うことがで
きる。According to the present invention, a window of an appropriate size is provided in the vicinity of an object to be measured including the object to be measured, and a density histogram in the window is created. It is possible to minimize the influence of the variation of the above, and the threshold value of the subsequent processing is determined based on the characteristics of the histogram data. Therefore, when the object to be measured is small with respect to the entire surface of the image, the contrast of the light and shade of the object to be measured with respect to the background portion is small, and further, the characteristics of the respective objects to be measured are different, and the variations are different. Even if the size or shape of the DUT cannot be predicted, such as when the size is large or a combination of these, it is possible to stably recognize the shape of the DUT and perform measurement using that image with high accuracy. be able to.
【００３５】特に、本発明によれば、例えばガラス内の
泡径の正確な測定が可能となり、測定の自動化による大
幅な省力化が可能になるとともに、製品の安定した品質
評価が可能となる。Particularly, according to the present invention, for example, the bubble diameter in the glass can be accurately measured, the labor of the measurement can be greatly saved by the automation of the measurement, and the stable quality evaluation of the product can be performed.
【図１】本発明における被測定物の測定に用いる装置の
配置構成を示す概略斜視図FIG. 1 is a schematic perspective view showing an arrangement configuration of an apparatus used for measuring an object to be measured in the present invention.
【図２】輪郭のはっきりした被測定物が撮像された画像
例を示す模式図FIG. 2 is a schematic diagram showing an example of an image in which an object to be measured having a clear contour is captured.
【図３】輪郭の薄い被測定物が撮像された画像例を示す
模式図FIG. 3 is a schematic diagram showing an example of an image in which an object to be measured having a thin contour is captured.
【図４】本発明における画像とウインドウと被測定物と
の関係の一例を示す模式図FIG. 4 is a schematic diagram showing an example of a relationship between an image, a window, and an object to be measured in the present invention.
【図５】本発明における被測定物の評価の方法の流れの
一例を示すフロー図FIG. 5 is a flow chart showing an example of the flow of a method for evaluating an object to be measured in the present invention.
【図６】輪郭のはっきりした被測定物を測定した場合に
得られるヒストグラムの一例を示すグラフFIG. 6 is a graph showing an example of a histogram obtained when an object to be measured having a clear contour is measured.
【図７】輪郭の薄い被測定物を測定した場合に得られる
ヒストグラムの一例を示すグラフFIG. 7 is a graph showing an example of a histogram obtained when measuring an object having a thin contour.
【図８】長い泡の補正を説明する概念図FIG. 8 is a conceptual diagram illustrating correction of long bubbles.
１：撮像装置 ２：ガラス ３：被測定物である泡 ４：白色拡散光源 ５：信号を処理、演算する装置 ６：ウインドウ ７：最小値 ８：判別分析によるしきい値 ９：最大頻度を持つ値 １０：細長い泡の突端部分 1: Imaging device 2: Glass 3: Bubble which is an object to be measured 4: White diffused light source 5: Device for processing and calculating a signal 6: Window 7: Minimum value 8: Threshold by discriminant analysis 9: Having a maximum frequency Value 10: Tip of elongated foam
Claims (9)
撮像して透明体にある被測定物の形状を評価する方法に
おいて、透明体を撮像した領域のうち被測定物を含む被
測定物の近傍を測定有効領域に定め、該有効領域におけ
る透明体を介した光の濃度のヒストグラムを作製し、該
ヒストグラムのデータに基づいてしきい値を決定して撮
像した光の情報をデジタル化し、該デジタル化した情報
から被測定物の評価をすることを特徴とする透明体にあ
る被測定物の形状の評価方法。1. A method of evaluating the shape of an object to be measured in a transparent body by irradiating the transparent body with light and imaging the light passing through the transparent body, wherein The vicinity of the object to be measured including is determined as the measurement effective area, a histogram of the density of light through the transparent body in the effective area is created, and the threshold value is determined based on the data of the histogram, and the information of the imaged light is obtained. A method for evaluating the shape of an object to be measured on a transparent body, comprising: digitizing the object, and evaluating the object to be measured from the digitized information.
周辺にある背景に相当する部分との境界における濃淡の
コントラストによりしきい値の決定手法を変えることを
特徴とする請求項１の透明体にある被測定物の形状の評
価方法。2. The threshold value determining method is changed according to the object to be measured, by the contrast of light and shade at the boundary between the object to be measured and a portion corresponding to the background around the object to be measured. Item 1. A method for evaluating the shape of an object to be measured which is a transparent body.
別分析の手法により求めた濃度値との差によってしきい
値を決定することを特徴とする請求項２の被測定物の評
価方法。3. The method for evaluating an object to be measured according to claim 2, wherein the threshold value is determined by the difference between the density value having the maximum frequency of the histogram and the density value obtained by the discriminant analysis method.
別分析の手法により求めた濃度値との差が大きい場合
に、判別分析の手法により求めた濃度値あるいはそれに
近い値をしきい値とすることを特徴とする請求項３の被
測定物の評価方法。4. When the difference between the density value having the maximum frequency in the histogram and the density value obtained by the discriminant analysis method is large, the density value obtained by the discriminant analysis method or a value close thereto is used as the threshold value. The method for evaluating an object to be measured according to claim 3, wherein.
別分析の手法により求めた濃度値との差が小さい場合
に、ヒストグラムにおける最大濃度値あるいは最小濃度
値と、最大頻度を持つ濃度値の差を用いてしきい値を決
めることを特徴とする請求項３の被測定物の評価方法。5. A difference between the maximum density value or the minimum density value in the histogram and the density value having the maximum frequency when the difference between the density value having the maximum frequency in the histogram and the density value obtained by the discriminant analysis method is small. The method for evaluating an object to be measured according to claim 3, wherein the threshold value is determined by using.
の二値化を行い、二値化されて１になった画素の数およ
び二値化されて０になった画素の数のうちの少なくとも
一方から被測定物を評価することを特徴とする請求項１
〜５のいずれかの被測定物の評価方法。6. The number of pixels binarized to 1 and the number of pixels binarized to 1 and the number of pixels binarized to 0 by binarizing each pixel to which light is incident using the threshold value. The object to be measured is evaluated from at least one of the two.
The evaluation method of the to-be-measured object in any one of-5.
濃度値と、最大頻度を持つ濃度値の差を変数とする数式
を用いて相対的にしきい値を決定することを特徴とする
請求項１〜６のいずれかの被測定物の評価方法。7. The threshold value is relatively determined by using a mathematical expression in which a difference between a maximum density value or a minimum density value of the histogram and a density value having a maximum frequency is used as a variable. 6. The method for evaluating an object to be measured according to any of 6.
撮像してガラス内部にある泡の形状を評価する方法にお
いて、ガラスを撮像した領域のうち泡を含む泡の近傍を
測定有効領域に定め、該有効領域におけるガラスを介し
た光の濃度のヒストグラムを作製し、該ヒストグラムの
データに基づいてしきい値を決定して撮像した光の情報
をデジタル化し、該デジタル化した情報から泡の大きさ
を求めることを特徴とするガラス内部にある泡の形状の
評価方法。8. A method for evaluating the shape of bubbles inside a glass by irradiating the glass with light and imaging the light passing through the glass, and measuring the vicinity of bubbles containing bubbles in the region where the glass is imaged. Area, and create a histogram of the density of light through the glass in the effective area, determine the threshold value based on the data of the histogram to digitize the information of the imaged light, from the digitized information A method for evaluating the shape of bubbles inside glass, which comprises determining the size of bubbles.
撮像してガラスの欠陥の度合を評価する方法において、
ガラスを撮像した領域のうちガラス内部にある泡を含む
泡の近傍を測定有効領域に定め、該有効領域におけるガ
ラスを介した光の濃度のヒストグラムを作製し、該ヒス
トグラムのデータに基づいてしきい値を決定して撮像し
た光の情報をデジタル化し、該デジタル化した情報から
泡の大きさを求め、ヒストグラムより求めた泡部分の濃
度に相当する値とを組み合わせてあるいはそれらを変数
とする数式により泡の欠点としての強さを評価すること
を特徴とするガラスの欠陥の度合の評価方法。9. A method for evaluating the degree of glass defects by irradiating the glass with light and imaging the light transmitted through the glass,
The vicinity of bubbles containing bubbles inside the glass in the imaged area of the glass is set as the measurement effective area, a histogram of the density of light passing through the glass in the effective area is created, and the threshold is determined based on the data of the histogram. A value is determined, the information of the imaged light is digitized, the size of the bubble is obtained from the digitized information, and the value corresponding to the concentration of the bubble portion obtained from the histogram is combined or a mathematical expression using them as variables. A method for evaluating the degree of glass defect, which is characterized by evaluating the strength of the bubble as a defect.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP8101606A JPH09287920A (en) | 1996-04-23 | 1996-04-23 | Method for evaluating shape of object to be measured, shape of bubble inside glass and degree of defect of glass |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP8101606A JPH09287920A (en) | 1996-04-23 | 1996-04-23 | Method for evaluating shape of object to be measured, shape of bubble inside glass and degree of defect of glass |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| JPH09287920A true JPH09287920A (en) | 1997-11-04 |
Family
ID=14305061
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP8101606A Pending JPH09287920A (en) | 1996-04-23 | 1996-04-23 | Method for evaluating shape of object to be measured, shape of bubble inside glass and degree of defect of glass |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JPH09287920A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002286655A (en) * | 2001-03-26 | 2002-10-03 | Shibuya Kogyo Co Ltd | Article inspection method |
| JP2010008125A (en) * | 2008-06-25 | 2010-01-14 | Toppan Printing Co Ltd | Bubble sorting method in glass substrate |
| JP2011117951A (en) * | 2009-11-05 | 2011-06-16 | Ohara Inc | Inspection method of internal state of optical material, and method for manufacturing optical element |
| JP2013257219A (en) * | 2012-06-13 | 2013-12-26 | Nisshin Seifun Group Inc | Surface shape measuring device and method |
| CN104568990A (en) * | 2015-01-10 | 2015-04-29 | 浙江大学 | Method for detecting bubble defect inside glass based on Mie scattering |
| CN104614386A (en) * | 2015-02-12 | 2015-05-13 | 江苏宇迪光学股份有限公司 | Lens defects type identification method |
| JP2020085775A (en) * | 2018-11-29 | 2020-06-04 | 住友金属鉱山株式会社 | Floss bubble diameter measurement device, flotation machine using the same, and floss bubble diameter measurement method |
-
1996
- 1996-04-23 JP JP8101606A patent/JPH09287920A/en active Pending
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002286655A (en) * | 2001-03-26 | 2002-10-03 | Shibuya Kogyo Co Ltd | Article inspection method |
| JP2010008125A (en) * | 2008-06-25 | 2010-01-14 | Toppan Printing Co Ltd | Bubble sorting method in glass substrate |
| JP2011117951A (en) * | 2009-11-05 | 2011-06-16 | Ohara Inc | Inspection method of internal state of optical material, and method for manufacturing optical element |
| JP2013257219A (en) * | 2012-06-13 | 2013-12-26 | Nisshin Seifun Group Inc | Surface shape measuring device and method |
| CN104568990A (en) * | 2015-01-10 | 2015-04-29 | 浙江大学 | Method for detecting bubble defect inside glass based on Mie scattering |
| CN104614386A (en) * | 2015-02-12 | 2015-05-13 | 江苏宇迪光学股份有限公司 | Lens defects type identification method |
| JP2020085775A (en) * | 2018-11-29 | 2020-06-04 | 住友金属鉱山株式会社 | Floss bubble diameter measurement device, flotation machine using the same, and floss bubble diameter measurement method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| TW229266B (en) | ||
| JPH0650290B2 (en) | How to match detected particles | |
| KR20010067345A (en) | Nondestructive inspection method and an apparatus thereof | |
| JP3095958B2 (en) | Method and apparatus for automatically characterizing, optimizing and inspecting an analysis method by penetrant inspection | |
| CN109596530A (en) | The dark field polarization imaging device and method of optical surface point and dust defect classification | |
| JPH09287920A (en) | Method for evaluating shape of object to be measured, shape of bubble inside glass and degree of defect of glass | |
| JPH0674865A (en) | Method and device for inspecting air bubble and foreign object in optical fiber preform | |
| JP2009075034A (en) | Method and apparatus for inspection of surface defect | |
| JP3789696B2 (en) | Battery electrolyte leak inspection apparatus and battery electrolyte leak inspection method | |
| JP2005189113A (en) | Surface inspecting device and surface inspection method | |
| US5606410A (en) | Method for controlling the surface state of one face of a solid and the associated device | |
| JP3417494B2 (en) | Method and apparatus for inspecting surface undulation of glass substrate | |
| CN1721841B (en) | Optical measure method and device | |
| US7679737B2 (en) | Method, system and apparatus of inspection | |
| JPH0629705B2 (en) | Plate inspection method | |
| JPH08178855A (en) | Method for inspecting light-transmissive object or specular object | |
| KR20220154345A (en) | Contour detection device using gaussian-weighted least squares and contour detection method thereof | |
| JP2715897B2 (en) | IC foreign matter inspection apparatus and method | |
| JP2982473B2 (en) | Painted surface inspection equipment | |
| JP2021148678A (en) | Defect detection device and defect detection method | |
| JPH03108638A (en) | Method of identifying particle aggregation pattern | |
| JP2005043222A (en) | Object surface inspection device and object surface inspection method | |
| KR20200044172A (en) | Deep learning method for analyzing defectiveness | |
| JPH09113463A (en) | Surface defect inspection apparatus | |
| JPH09229870A (en) | Method and apparatus for inspecting glass article for cathode ray tube |