JP2019158857A - Method and apparatus for inspecting linear pattern defect in width direction of metal strip surface - Google Patents
Method and apparatus for inspecting linear pattern defect in width direction of metal strip surface Download PDFInfo
- Publication number
- JP2019158857A JP2019158857A JP2018121415A JP2018121415A JP2019158857A JP 2019158857 A JP2019158857 A JP 2019158857A JP 2018121415 A JP2018121415 A JP 2018121415A JP 2018121415 A JP2018121415 A JP 2018121415A JP 2019158857 A JP2019158857 A JP 2019158857A
- Authority
- JP
- Japan
- Prior art keywords
- difference
- output
- linear
- rectangular
- value
- 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.)
- Granted
Links
- 229910052751 metal Inorganic materials 0.000 title claims abstract description 115
- 239000002184 metal Substances 0.000 title claims abstract description 115
- 230000007547 defect Effects 0.000 title claims abstract description 82
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000007689 inspection Methods 0.000 claims abstract description 29
- 238000012935 Averaging Methods 0.000 claims abstract description 19
- 238000003384 imaging method Methods 0.000 claims abstract description 5
- 238000012545 processing Methods 0.000 claims description 15
- 238000000605 extraction Methods 0.000 claims description 5
- 239000000284 extract Substances 0.000 claims description 2
- 238000012546 transfer Methods 0.000 description 23
- 238000004519 manufacturing process Methods 0.000 description 5
- 230000005856 abnormality Effects 0.000 description 4
- 230000000052 comparative effect Effects 0.000 description 3
- 238000012937 correction Methods 0.000 description 3
- 229910000831 Steel Inorganic materials 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 239000010959 steel Substances 0.000 description 2
- 238000009966 trimming Methods 0.000 description 2
- RYGMFSIKBFXOCR-UHFFFAOYSA-N Copper Chemical compound [Cu] RYGMFSIKBFXOCR-UHFFFAOYSA-N 0.000 description 1
- 239000000956 alloy Substances 0.000 description 1
- 229910045601 alloy Inorganic materials 0.000 description 1
- 229910052782 aluminium Inorganic materials 0.000 description 1
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 description 1
- 229910052802 copper Inorganic materials 0.000 description 1
- 239000010949 copper Substances 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- -1 etc. Substances 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000005304 joining Methods 0.000 description 1
- 150000002739 metals Chemical class 0.000 description 1
- 238000003466 welding Methods 0.000 description 1
- 230000037303 wrinkles Effects 0.000 description 1
Abstract
Description
本発明は、金属帯表面の幅方向線状模様欠陥の検査方法及び検査装置に関し、特に、金属帯の幅方向（単に「幅方向」と称することもある。）に沿って発生する線状の軽微な模様欠陥であっても精度よく検出することのできる、金属帯表面の幅方向線状模様欠陥の検査方法及び検査装置に関する。 The present invention relates to an inspection method and inspection apparatus for a widthwise linear pattern defect on a surface of a metal strip, and in particular, a linear shape generated along the width direction of the metal strip (sometimes simply referred to as “width direction”). The present invention relates to a width direction linear pattern defect inspection method and inspection apparatus capable of accurately detecting even a slight pattern defect.
金属帯の製造ラインにおいては、金属帯の品質不良を防止するために、金属帯の表面検査を行う必要がある。従来、金属帯の表面検査をする際には、金属帯の表面に光を照射し、その反射光を画像取得手段において受光、撮像し、得られた表面画像を用いて異常の有無が判定される。 In a metal band production line, it is necessary to inspect the surface of the metal band in order to prevent the quality defect of the metal band. Conventionally, when performing surface inspection of a metal strip, the surface of the metal strip is irradiated with light, the reflected light is received and imaged by an image acquisition means, and the presence or absence of an abnormality is determined using the obtained surface image. The
金属帯の表面に発生する欠陥は、表面画像中において輝度及び明度等の出力信号値が、異常に高い或いは低い値を示す。よって、表面画像において出力信号値の閾値を設け、出力信号値の絶対値が当該閾値を超える箇所を抽出することにより、金属帯の表面欠陥を検出するという方法が一般的に行われている。 Defects that occur on the surface of the metal strip show abnormally high or low values of output signal values such as brightness and brightness in the surface image. Therefore, a method of detecting a surface defect of a metal strip is generally performed by providing a threshold value of the output signal value in the surface image and extracting a portion where the absolute value of the output signal value exceeds the threshold value.
具体的な表面検査装置を開示した文献として、特許文献１が挙げられる。特許文献１では、一つの光源によって金属帯の表面を幅方向に光走査し、反射光のうち正反射成分を受光する光電変換器と、乱反射成分を受光する光電変換器とを用いて、２つの光電変換器より出力された信号を差し引くことにより、乱反射光の影響を排して欠陥信号を強調する方法が開示されている。 Patent document 1 is mentioned as a literature which disclosed the specific surface inspection apparatus. In Patent Document 1, a surface of a metal strip is optically scanned in the width direction by one light source, and a photoelectric converter that receives a regular reflection component of reflected light and a photoelectric converter that receives a diffuse reflection component are used. A method is disclosed in which a defect signal is enhanced by subtracting the signals output from two photoelectric converters to eliminate the influence of irregularly reflected light.
金属帯の品質検査を行う場合、特に問題となる欠陥として、金属帯の幅方向に沿って発生する線状模様欠陥の一種である、溶接部の転写痕が挙げられる。当該転写痕の発生するメカニズムについて以下に説明する。 When performing a quality inspection of a metal strip, a particularly problematic defect is a transfer mark of a weld, which is a kind of linear pattern defect that occurs along the width direction of the metal strip. A mechanism for generating the transfer mark will be described below.
金属帯の製造設備では、複数の金属帯の長手方向（単に「長手方向」と称することもある。）の端面同士を幅方向に溶接して、金属帯を連続的に通板することが行われる。通板の際には、金属帯の表裏面と接触する各種の通板ロールが用いられる。金属帯の幅方向溶接部が当該通板ロールの表面に接触すると、溶接部の僅かな凹凸によって当該通板ロール表面に線状模様が転写される場合がある。次いで、当該通板ロール表面に転写された線状模様が溶接部以外の金属帯と接触すると、金属帯表面に僅かな線状模様が転写されて溶接部の転写痕が発生する。このように、金属帯表面へ転写される溶接部の転写痕は、その反射光の減衰度合いもわずかであり、視認も困難である。しかし、当該転写痕による金属帯の表面品質低下は問題であり、製造の段階で確実に当該転写痕を検出することが望まれる。 In a metal strip manufacturing facility, the end surfaces of a plurality of metal strips in the longitudinal direction (also simply referred to as “longitudinal direction”) are welded in the width direction, and the metal strips are continuously passed through. Is called. When passing plates, various passing plate rolls that come into contact with the front and back surfaces of the metal strip are used. When the widthwise welded portion of the metal band comes into contact with the surface of the sheet passing roll, a linear pattern may be transferred to the surface of the sheet passing roll due to slight unevenness of the welded portion. Subsequently, when the linear pattern transferred to the surface of the sheet passing roll comes into contact with the metal band other than the welded portion, a slight linear pattern is transferred to the surface of the metal band, and a transfer mark of the welded portion is generated. Thus, the transfer mark of the welded portion transferred to the surface of the metal band has a slight degree of attenuation of the reflected light and is difficult to visually recognize. However, the deterioration of the surface quality of the metal band due to the transfer mark is a problem, and it is desired to detect the transfer mark reliably at the manufacturing stage.
しかし、従来の表面検査方法では、前述した転写痕を精度よく検知することが難しいという問題があった。その理由として、転写痕が凹凸のほとんどない模様状の欠陥であって、金属帯表面に不可避的に発生するノイズ成分と十分に分離できない、つまり十分なＳ／Ｎ比を確保できないということが挙げられる。より具体的には、金属帯表面には性状の不均一さや表面全体にわたる模様等が不可避的に発生し、これらのノイズ成分も撮影画像において出力信号の異常値として現れることがある。上述したように、出力信号値の強度を基準に前述した転写痕を検出しようとすると、不可避的に発生するノイズ成分による異常と前述した転写痕に基づく異常との区別がつかず、前述した転写痕の検出が困難である。 However, the conventional surface inspection method has a problem that it is difficult to accurately detect the transfer marks described above. The reason is that the transfer mark is a pattern-like defect with almost no unevenness and cannot be sufficiently separated from noise components inevitably generated on the surface of the metal band, that is, a sufficient S / N ratio cannot be secured. It is done. More specifically, unevenness of properties or a pattern over the entire surface inevitably occurs on the surface of the metal band, and these noise components may also appear as abnormal values of the output signal in the photographed image. As described above, when trying to detect the transfer mark described above based on the intensity of the output signal value, it is impossible to distinguish between the abnormality caused by the unavoidable noise component and the abnormality based on the transfer mark described above. It is difficult to detect a mark.
本発明は、上述の問題点を鑑み、金属帯の幅方向に延在する溶接部由来の転写痕を精度よく検出することのできる、金属帯表面の幅方向線状模様欠陥の検査方法及び検査装置を提供することを課題とする。 In view of the above-described problems, the present invention can accurately detect a transfer mark derived from a weld extending in the width direction of a metal strip, and a method for inspecting the width direction linear pattern defect on the surface of the metal strip and the inspection. It is an object to provide an apparatus.
本発明の手段は、次のとおりである。
［１］金属帯表面を撮像して表面画像を取得し、画素数及び基準画素を設定した矩形フィルタを用いて、前記表面画像における基準画素の出力信号値を取得し、前記矩形フィルタ内の平均化処理した出力信号値を取得し、さらに前記基準画素の出力信号値と前記矩形フィルタ内の平均化処理した出力信号値との差分を矩形出力差分として取得し、前記矩形フィルタに比べて前記金属帯の長手方向の画素数を少なくした線形フィルタを用いて、前記表面画像における基準画素の出力信号値を取得し、前記線形フィルタ内の平均化処理した出力信号値を取得し、さらに前記基準画素の出力信号値と前記線形フィルタ内の平均化処理した出力信号値との差分を線形出力差分として取得し、その後、同一の基準画素について、前記矩形出力差分と前記線形出力差分との差の絶対値を取得するとともに、所定の閾値Ａを設定し、出力差分の差の絶対値が前記所定の閾値Ａ以上又は閾値Ａを超える画素を抽出し、当該画素に対応する金属帯表面部位を幅方向線状模様欠陥として判定することを特徴とする金属帯表面の幅方向線状模様欠陥の検査方法。
［２］前記矩形出力差分を所定の閾値Ｂを用いて２値化し、前記線形出力差分を所定の閾値Ｂを用いて２値化し、２値化した矩形出力差分と２値化した線形出力差分との差の絶対値を２値化出力差分差として取得して、当該絶対値が所定値Ｃ以上である画素に対応する金属帯表面部分を幅方向線状模様欠陥として判定することを特徴とする［１］に記載の金属帯表面の幅方向線状模様欠陥の検査方法。
［３］前記矩形フィルタ内の基準画素の出力信号値を所定の閾値Ｄを用いて２値化し、前記矩形フィルタ内の平均化処理した出力信号値を前記所定の閾値Ｄを用いて２値化し、前記矩形フィルタ内の前記基準画素の出力信号２値化値と前記平均化処理した出力信号２値化値との差を矩形出力２値化差分として取得し、前記線形フィルタ内の基準画素の出力信号値を前記所定の閾値Ｄを用いて２値化し、前記線形フィルタ内の平均化処理した出力信号値を前記所定の閾値Ｄを用いて２値化し、前記線形フィルタ内の前記基準画素の出力信号２値化値と前記平均化処理した出力信号２値化値との差を線形出力２値化差分として取得し、前記矩形出力２値化差分と前記線形出力２値化差分との差の絶対値を取得して、当該絶対値が所定値Ｅ以上である画素に対応する金属帯表面部位を幅方向線状模様欠陥として判定することを特徴とする［１］に記載の金属帯表面の幅方向線状模様欠陥の検査方法。
［４］前記線形フィルタの幅方向の画素数を前記矩形フィルタの幅方向の画素数に一致させることを特徴とする［１］〜［３］のいずれかに記載の金属帯表面の幅方向線状模様欠陥の検査方法。
［５］金属帯表面を撮像して表面画像を取得する画像取得手段と、画素数および基準画素を設定した矩形フィルタを用いて、前記表面画像における基準画素の出力信号値を取得する矩形基準画素出力値取得手段と、前記矩形フィルタ内の平均化処理した出力信号値を取得する矩形平均化出力値取得手段と、前記矩形基準画素出力値取得手段で取得した基準画素の出力信号値と前記矩形平均化出力値取得手段で取得した矩形フィルタ内の平均化処理した出力信号値との差分を矩形出力差分として取得する矩形出力差分取得手段と、前記矩形フィルタに比べて前記金属帯の長手方向の画素数を少なくした線形フィルタを用いて、前記表面画像における基準画素の出力信号値を取得する線形基準画素出力値取得手段と、前記線形フィルタ内の平均化処理した出力信号値を取得する線形平均化出力値取得手段と、前記線形基準画素出力値取得手段で取得した基準画素の出力信号値と前記線形平均化出力値取得手段で取得した線形フィルタ内の平均化処理した出力信号値との差分を線形出力差分として取得する線形出力差分取得手段と、同一の基準画素について、前記矩形出力差分と前記線形出力差分との差の絶対値を取得する出力差分差取得手段と、所定の閾値Ａを設定するとともに、前記出力差分差取得手段により取得した出力差分の差の絶対値が、前記所定の閾値Ａを超える画素を抽出する出力差分差画素抽出手段と、当該出力差分差画素抽出手段で抽出した画素に対応する金属帯表面部位を幅方向線状模様欠陥として判定する判定手段とを備えることを特徴とする金属帯表面の幅方向線状模様欠陥の検査装置。
［６］前記矩形出力差分取得手段で取得した矩形出力差分を所定の閾値Ｂを用いて２値化する矩形出力差分２値化手段と、前記線形出力差分取得手段で取得した線形出力差分を所定の閾値Ｂを用いて２値化する線形出力差分２値化手段と、２値化した矩形出力差分と２値化した線形出力差分との差の絶対値を取得する２値化出力差分差取得手段と、当該絶対値が所定値Ｃ以上である画素に対応する金属帯表面部分を幅方向線状模様欠陥として判定する判定手段と、を備えることを特徴とする［５］に記載の金属帯表面の幅方向線状模様欠陥の検査装置。
［７］前記矩形基準画素出力値取得手段で取得した基準画素の出力信号値を所定の閾値Ｄを用いて２値化する矩形基準画素出力２値化手段と、前記矩形平均化出力値取得手段で取得した矩形フィルタ内の平均化処理した出力信号値を前記所定の閾値Ｄを用いて２値化する線形基準画素２値化手段と、前記矩形フィルタ内の平均化処理した出力信号値を２値化した値と前記基準画素の出力信号値を２値化した値との差を矩形出力２値化差分として取得する矩形出力２値化差分取得手段と、前記線形フィルタ内の平均化処理した出力信号値を２値化した値と前記基準画素の出力信号値を２値化した値との差を線形出力２値化差分として取得する線形出力２値化差分取得手段と、前記矩形出力２値化差分と前記線形出力２値化差分との差の絶対値を取得する出力２値化差分差取得手段と、当該絶対値が所定値Ｅ以上である画素に対応する金属帯表面部位を幅方向線状模様欠陥として判定する判定手段と、を備えることを特徴とする［５］に記載の金属帯表面の幅方向線状模様欠陥の検査装置。
［８］前記線形フィルタの幅方向の画素数と前記矩形フィルタの幅方向の画素数とが一致することを特徴とする［５］〜［７］のいずれかに記載の金属帯表面の幅方向線状模様欠陥の検査装置。
Means of the present invention are as follows.
[1] Obtain a surface image by imaging the surface of the metal band, obtain an output signal value of a reference pixel in the surface image using a rectangular filter in which the number of pixels and a reference pixel are set, and calculate the average in the rectangular filter The output signal value obtained by the averaging process is acquired, and the difference between the output signal value of the reference pixel and the averaged output signal value in the rectangular filter is acquired as a rectangular output difference, and the metal signal is compared with the rectangular filter. Using a linear filter in which the number of pixels in the longitudinal direction of the band is reduced, an output signal value of a reference pixel in the surface image is acquired, an output signal value averaged in the linear filter is acquired, and the reference pixel The difference between the output signal value of the output signal value and the output signal value averaged in the linear filter is obtained as a linear output difference, and then, for the same reference pixel, the rectangular output difference and the The absolute value of the difference from the output difference is acquired, and a predetermined threshold A is set, and a pixel whose absolute value of the difference in output difference is equal to or greater than the predetermined threshold A or exceeds the threshold A, and corresponds to the pixel A method for inspecting a widthwise linear pattern defect on a surface of a metal band, wherein the metal band surface portion to be determined is determined as a widthwise linear pattern defect.
[2] The rectangular output difference is binarized using a predetermined threshold B, the linear output difference is binarized using a predetermined threshold B, and the binarized rectangular output difference and binarized linear output difference The absolute value of the difference between the two is obtained as a binarized output difference difference, and the metal band surface portion corresponding to a pixel having the absolute value equal to or greater than a predetermined value C is determined as a widthwise linear pattern defect. The inspection method of the width direction linear pattern defect of the metal strip surface as described in [1].
[3] The output signal value of the reference pixel in the rectangular filter is binarized using a predetermined threshold D, and the averaged output signal value in the rectangular filter is binarized using the predetermined threshold D. The difference between the output signal binarized value of the reference pixel in the rectangular filter and the averaged output signal binarized value is obtained as a rectangular output binarized difference, and the reference pixel in the linear filter The output signal value is binarized using the predetermined threshold D, the averaged output signal value in the linear filter is binarized using the predetermined threshold D, and the reference pixel in the linear filter is binned. The difference between the output signal binarized value and the averaged output signal binarized value is acquired as a linear output binarized difference, and the difference between the rectangular output binarized difference and the linear output binarized difference The absolute value is acquired and the absolute value is greater than or equal to the predetermined value E Method of inspecting the width direction linear pattern defects metal strip surface according to [1], wherein the determining the metal strip surface portion corresponding to the pixel as the width direction linear pattern defect.
[4] The width direction line on the surface of the metal band according to any one of [1] to [3], wherein the number of pixels in the width direction of the linear filter is made to coincide with the number of pixels in the width direction of the rectangular filter. Inspection method for pattern defects.
[5] A rectangular reference pixel that acquires an output signal value of a reference pixel in the surface image using an image acquisition unit that captures the surface of the metal band and acquires a surface image, and a rectangular filter in which the number of pixels and the reference pixel are set Output value acquisition means, rectangular average output value acquisition means for acquiring the averaged output signal value in the rectangular filter, reference pixel output signal value acquired by the rectangular reference pixel output value acquisition means and the rectangle Rectangle output difference acquisition means for acquiring a difference from the averaged output signal value in the rectangular filter acquired by the averaged output value acquisition means as a rectangular output difference, and in the longitudinal direction of the metal band compared to the rectangular filter Linear reference pixel output value acquisition means for acquiring an output signal value of a reference pixel in the surface image using a linear filter with a reduced number of pixels, and averaging in the linear filter A linear averaged output value acquisition means for acquiring the processed output signal value; an output signal value of the reference pixel acquired by the linear reference pixel output value acquisition means; and a linear filter acquired by the linear averaged output value acquisition means Linear output difference acquisition means for acquiring a difference from the averaged output signal value as a linear output difference, and an output difference for acquiring the absolute value of the difference between the rectangular output difference and the linear output difference for the same reference pixel A difference acquisition unit; and an output difference difference pixel extraction unit that sets a predetermined threshold A and extracts pixels whose absolute value of the difference between the output differences acquired by the output difference difference acquisition unit exceeds the predetermined threshold A. And a determination means for determining a metal band surface portion corresponding to the pixel extracted by the output difference difference pixel extraction means as a width direction linear pattern defect. Inspection apparatus of Jo pattern defect.
[6] A rectangular output difference binarizing unit that binarizes the rectangular output difference acquired by the rectangular output difference acquiring unit using a predetermined threshold B, and a linear output difference acquired by the linear output difference acquiring unit is predetermined. Binary output difference binarization means for binarization using the threshold B of the image, and binarized output difference difference acquisition for acquiring the absolute value of the difference between the binarized rectangular output difference and the binarized linear output difference The metal strip according to [5], further comprising: a means for determining a surface portion of the metal strip corresponding to a pixel whose absolute value is equal to or greater than the predetermined value C as a widthwise linear pattern defect. Inspection device for widthwise pattern defects on the surface.
[7] A rectangular reference pixel output binarizing unit that binarizes the output signal value of the reference pixel acquired by the rectangular reference pixel output value acquiring unit using a predetermined threshold D, and the rectangular averaged output value acquiring unit Linear reference pixel binarization means for binarizing the averaged output signal value in the rectangular filter obtained in step 2 using the predetermined threshold D, and the averaged output signal value in the rectangular filter is 2 A rectangular output binarized difference acquisition means for acquiring a difference between a binarized value and a value obtained by binarizing the output signal value of the reference pixel as a rectangular output binarized difference, and averaging processing in the linear filter Linear output binarized difference acquisition means for acquiring a difference between a binarized value of the output signal value and a binarized value of the output signal value of the reference pixel as a linear output binarized difference; and the rectangular output 2 The absolute value of the difference between the binarized difference and the linear output binarized difference And an output binarized difference difference obtaining means to be obtained, and a judging means for judging a metal band surface portion corresponding to a pixel whose absolute value is equal to or greater than a predetermined value E as a widthwise linear pattern defect. [5] The inspection apparatus for the widthwise linear pattern defect on the surface of the metal strip according to [5].
[8] The width direction of the surface of the metal strip according to any one of [5] to [7], wherein the number of pixels in the width direction of the linear filter and the number of pixels in the width direction of the rectangular filter match. Inspection device for linear pattern defects.
本発明によれば、金属帯の幅方向に延在する金属帯溶接部由来の転写痕を、金属帯表面性状の不均一さや表面の模様等の他のノイズと分離して精度よく検出することができる。 According to the present invention, the transfer mark derived from the metal band weld extending in the width direction of the metal band can be accurately detected by separating it from other noise such as non-uniformity of the metal band surface property and surface pattern. Can do.
以下、本発明を具体的に説明する。 Hereinafter, the present invention will be specifically described.
金属帯の製造設備においては、図１に示すように、金属帯１は通板ロール２と接触しつつ、連続的に通板される。金属帯１の表面には光照射手段３から光線が照射される。金属帯１の表面で反射した光線は、画像取得手段４にて受光され、金属帯１の表面画像が取得される。表面画像のデータは、演算装置５に送られて画像処理（詳細については後述する。）を経た後に、金属帯１表面の幅方向線状模様欠陥の有無の判定が行われる。 In the metal strip manufacturing facility, as shown in FIG. 1, the metal strip 1 is continuously passed while being in contact with the passing plate roll 2. The surface of the metal strip 1 is irradiated with light from the light irradiation means 3. The light beam reflected by the surface of the metal band 1 is received by the image acquisition means 4 and the surface image of the metal band 1 is acquired. The surface image data is sent to the arithmetic unit 5 and subjected to image processing (details will be described later), and then the presence / absence of a width-direction linear pattern defect on the surface of the metal strip 1 is determined.
画像取得手段４による撮像を行う通板ラインの位置は、金属帯１が通板ロール２と接触している表面側が望ましい。当該箇所では、金属帯１の上下動が抑えられ、ピントずれ等の少ない安定した表面画像を得ることができる。 The position of the sheet passing line for imaging by the image acquisition means 4 is preferably on the surface side where the metal strip 1 is in contact with the sheet passing roll 2. In this place, the vertical movement of the metal strip 1 is suppressed, and a stable surface image with little focus deviation can be obtained.
金属帯の製造過程では、複数の金属帯の長手方向の端面同士を幅方向に溶接して繋ぎ合わせることで、金属帯１の連続通板が可能となる。金属帯１の溶接部における凹凸形状は、通板ロール２と接触する際に通板ロール２表面に線状模様を形成させる。次いで、通板ロール２の当該線状模様が金属帯１の表面に再度転写することで、金属帯１の表面には僅かな線状模様の転写痕が形成される。当該転写痕は、金属帯表面に形成される幅方向線状模様欠陥の典型例である。当該転写痕は、目視では困難な程度にごく微小な凹凸形状であり、金属帯表面性状の不均一さや表面全体にわたる模様等のその他のノイズ成分との分離が難しい。しかし本発明に係る検査方法によれば、当該転写痕を精度よく検出することができる。 In the manufacturing process of the metal strip, the continuous strips of the metal strip 1 can be achieved by welding and joining the end faces in the longitudinal direction of the plurality of metal strips in the width direction. The uneven shape in the welded portion of the metal band 1 forms a linear pattern on the surface of the sheet passing roll 2 when contacting the sheet passing roll 2. Next, the linear pattern of the sheet passing roll 2 is transferred again to the surface of the metal band 1, whereby a slight linear pattern transfer mark is formed on the surface of the metal band 1. The transfer mark is a typical example of the widthwise linear pattern defect formed on the surface of the metal band. The transfer mark has a concavo-convex shape that is extremely minute to the extent that it is difficult to visually observe, and is difficult to separate from other noise components such as unevenness of the surface property of the metal band and a pattern over the entire surface. However, according to the inspection method of the present invention, the transfer mark can be detected with high accuracy.
ここで、本発明の主な特徴である演算装置５において行われる画像処理について、詳細に説明する。 Here, the image processing performed in the arithmetic unit 5 which is the main feature of the present invention will be described in detail.
画像取得手段によって得られた表面画像は、二次元に配列された多数の画素からなり、各画素において出力信号値が記録される。出力信号は、光の強度等を示すパラメータであればよく、例えば輝度及び明度が挙げられる。金属帯の表面に疵等の異常が存在すると、当該箇所の出力信号値は周囲より高い、又は低いといった挙動を示す。 The surface image obtained by the image acquisition means is composed of a number of pixels arranged two-dimensionally, and an output signal value is recorded at each pixel. The output signal only needs to be a parameter indicating the intensity of light, and examples thereof include brightness and brightness. When an abnormality such as a wrinkle is present on the surface of the metal strip, the output signal value at the corresponding portion exhibits a behavior such as higher or lower than the surroundings.
本発明では、まず表面画像について、矩形フィルタを用いて平均化処理を行う。 In the present invention, the surface image is first averaged using a rectangular filter.
具体的には、図２を用いて説明する。まず、矩形フィルタを表面画像の一部分に当てはめる。当該矩形フィルタは、金属帯の幅方向及び長手方向にそれぞれ複数の画素数が設定されたフィルタであればよい。図２の例における矩形フィルタは、金属帯の幅方向の画素数が５であり、長手方向の画素数が５である。矩形フィルタ内の画素のうち、所定の位置を基準画素（入力画素）として設定する。次に、矩形フィルタ内の画素の出力信号値から、フィルタ内に存在する全画素の出力信号値の平均値を求める。その後、前記基準画素の出力信号値と前記出力信号値の平均値との差分を取得する。さらに所定の値を中央値として加算して補正を行ってもよい。当該中央値は、出力信号値の計算値（平均値を減算した後の値）が０に極めて近い値となると計算上不便であることから、計算値の操作性を向上させるために便宜的に設定してもよい。図２の例では、輝度値の最大値（２５６）と最小値（０）との略中間に位置する中間値（１２７）を中央値として、補正を行った。 Specifically, this will be described with reference to FIG. First, a rectangular filter is applied to a part of the surface image. The rectangular filter may be a filter in which a plurality of pixels are set in the width direction and the longitudinal direction of the metal strip. The rectangular filter in the example of FIG. 2 has 5 pixels in the width direction of the metal band and 5 pixels in the longitudinal direction. Among the pixels in the rectangular filter, a predetermined position is set as a reference pixel (input pixel). Next, an average value of the output signal values of all the pixels existing in the filter is obtained from the output signal values of the pixels in the rectangular filter. Thereafter, the difference between the output signal value of the reference pixel and the average value of the output signal values is obtained. Furthermore, correction may be performed by adding a predetermined value as a median value. This median value is inconvenient in calculation when the calculated value of the output signal value (value after subtracting the average value) is very close to 0. Therefore, for the sake of convenience in order to improve the operability of the calculated value. It may be set. In the example of FIG. 2, correction is performed with an intermediate value (127) positioned approximately in the middle between the maximum value (256) and the minimum value (0) of the luminance value as a median value.
上記の処理を、元の表面画像の幅方向及び長手方向の各画素について行う。それぞれの画素について、計算によって求めた出力信号値の差分を表示した画像を作成してもよい。これにより、表面画像と、該表面画像について矩形フィルタを用いて平均化処理した画像との差分を取った矩形出力差分画像が生成される。 The above processing is performed for each pixel in the width direction and the longitudinal direction of the original surface image. You may create the image which displayed the difference of the output signal value calculated | required by calculation about each pixel. As a result, a rectangular output difference image is generated by taking the difference between the surface image and an image obtained by averaging the surface image using a rectangular filter.
次に、図３を用いて線形フィルタを用いた平均化処理について説明する。線形フィルタは、矩形フィルタに比べて、金属帯長手方向の画素数が少ない。また、線形フィルタは、金属帯の幅方向の画素数が長手方向の画素数よりも多いことが好ましい。図３の例では長手方向の画素数を１とし、幅方向の画素数を５としている。線形フィルタにおいても同様に基準画素を設定し、次に線形フィルタ内の画素の出力信号値から、線形フィルタ内に存在する全画素の出力信号値の平均値を求める。その後、前記基準画素の出力信号値と前記出力信号値の平均値との差分を取得する。さらに所定の値を中央値として加算して補正を行ってもよい。尚、当該中央値は、前述した矩形状のフィルタにおける平均化処理で用いた値（１２７）と同様の値とする。上述の計算を、元の表面画像の幅方向及び長手方向の各画素を基準画素として行う。それぞれの画素について、計算によって得られた出力信号値の差分を表示した画像を作成してもよい。以上により、線形出力差分画像を取得する。 Next, the averaging process using a linear filter will be described with reference to FIG. The linear filter has fewer pixels in the metal band longitudinal direction than the rectangular filter. In the linear filter, the number of pixels in the width direction of the metal strip is preferably larger than the number of pixels in the longitudinal direction. In the example of FIG. 3, the number of pixels in the longitudinal direction is 1, and the number of pixels in the width direction is 5. Similarly, in the linear filter, the reference pixel is set, and the average value of the output signal values of all the pixels existing in the linear filter is obtained from the output signal value of the pixel in the linear filter. Thereafter, the difference between the output signal value of the reference pixel and the average value of the output signal values is obtained. Furthermore, correction may be performed by adding a predetermined value as a median value. The median value is the same value as the value (127) used in the averaging process in the rectangular filter described above. The above calculation is performed using the pixels in the width direction and the longitudinal direction of the original surface image as reference pixels. You may create the image which displayed the difference of the output signal value obtained by calculation about each pixel. Thus, a linear output difference image is acquired.
尚、矩形フィルタ内の出力信号値の平均値及び線形フィルタ内の出力信号値の平均値を求める手段は、算出平均、相乗平均、調和平均などいずれの平均化処理を行ってもよい。 The means for obtaining the average value of the output signal values in the rectangular filter and the average value of the output signal values in the linear filter may perform any averaging process such as calculation average, geometric average, and harmonic average.
尚、前述の処理は、金属帯表面を撮像した表面画像中の全ての画素について行うことが好ましいが、一部の画素のみに限って行ってもよい。例えば、幅方向端部をトリミングして製造する金属帯では、端部のトリミング部分を除いた幅方向中央部の画素のみに限って、前述の処理を行ってもよい。 The above-described processing is preferably performed for all the pixels in the surface image obtained by imaging the surface of the metal band, but may be performed only for some pixels. For example, in the case of a metal band manufactured by trimming the width direction end portion, the above-described processing may be performed only for the pixel in the width direction center portion excluding the trimming portion at the end portion.
線形フィルタは、長手方向の画素数を前述した矩形状のフィルタの長手方向の画素数よりも少なくする。本発明で測定対象とする幅方向の線状模様欠陥をより精度よく抽出するという観点からは、線形フィルタの長手方向における画素数は１〜３とするのが好ましく、１とするのが最も好ましい。 The linear filter makes the number of pixels in the longitudinal direction smaller than the number of pixels in the longitudinal direction of the rectangular filter described above. From the viewpoint of more accurately extracting the linear pattern defect in the width direction to be measured in the present invention, the number of pixels in the longitudinal direction of the linear filter is preferably 1 to 3, and most preferably 1. .
また、線形フィルタの幅方向の画素数と、矩形フィルタの幅方向の画素数とを一致させることが好ましい。これにより、フィルタ内の出力信号の平均化処理において、金属帯の幅方向における矩形フィルタの出力信号の平均値と線形フィルタの出力信号の平均値との重み付けを同じにできるので、これらの差を信頼度高く取得できるという効果を奏する。 In addition, it is preferable that the number of pixels in the width direction of the linear filter is equal to the number of pixels in the width direction of the rectangular filter. Thereby, in the averaging process of the output signal in the filter, the weight of the average value of the output signal of the rectangular filter and the average value of the output signal of the linear filter in the width direction of the metal band can be made the same. There is an effect that it can be obtained with high reliability.
上述のような平均化処理を行うことによって、矩形出力差分画像では矩形フィルタを用いた出力信号値の差分が各画素について取得され、線形出力差分画像では線形フィルタを用いた出力信号値の差分が各画素について取得される。矩形出力差分画像と線形出力差分画像とは同じ表面画像を元に生成されるので、矩形出力差分画像と線形出力差分画像との同じ位置の画素は、それぞれ金属帯表面の同一箇所に対応する。 By performing the averaging process as described above, the difference between the output signal values using the rectangular filter is obtained for each pixel in the rectangular output difference image, and the difference between the output signal values using the linear filter is obtained in the linear output difference image. Obtained for each pixel. Since the rectangular output difference image and the linear output difference image are generated based on the same surface image, pixels at the same position in the rectangular output difference image and the linear output difference image respectively correspond to the same location on the surface of the metal strip.
次に、矩形出力差分画像と線形出力差分画像との差を取る。これにより、各画素において矩形出力差分画像の出力信号値の差分から、線形出力差分画像の出力信号値の差分を減算した出力信号値の差分の差を取得する。尚、取得した出力信号値の差分の差を、基準画素に対応させて画像として表示してもよい。 Next, the difference between the rectangular output difference image and the linear output difference image is taken. As a result, the difference between the output signal values obtained by subtracting the difference between the output signal values of the linear output difference image from the difference between the output signal values of the rectangular output difference image at each pixel is acquired. Note that the difference in the difference between the acquired output signal values may be displayed as an image corresponding to the reference pixel.
これにより、幅方向の出力信号値の差分の差を除く出力信号値の差分の差を０に近づけることができ、その結果幅方向のみ出力信号値の差分の差を強調することができる。さらに、２値化を行うことで各画素の出力信号値の差分の差の絶対値が閾値を超える画素と、閾値以下となる画素とに分けられ、出力信号値の差分の差の絶対値が大きい画素に対応する金属帯表面部位を幅方向線状模様欠陥として特定することができる。 As a result, the difference in output signal value difference excluding the difference in output signal value in the width direction can be brought close to 0, and as a result, the difference in output signal value difference can be emphasized only in the width direction. Furthermore, by performing binarization, the absolute value of the difference between the output signal values of each pixel is divided into a pixel that exceeds the threshold and a pixel that is less than or equal to the threshold, and the absolute value of the difference between the output signal values is A metal band surface portion corresponding to a large pixel can be specified as a widthwise linear pattern defect.
図４に、上述で説明した画像処理の工程のフロー図（その１）を示す。
まず、金属帯表面を撮像した画像を取得する（Ｓ０）。
次に、矩形フィルタの画素数を設定し（Ｓ１）、当該矩形フィルタ画素内の基準画素を設定し（Ｓ２）、矩形基準画素出力値取得手段により当該基準画素の出力信号値を取得する（Ｓ３）。
さらに、矩形平均化出力値取得手段により矩形フィルタ内の画素の出力信号値の算術平均値を算出する（Ｓ４）。
矩形出力差分取得手段により、前記Ｓ３で取得した基準画素の出力信号値と、前記Ｓ４で取得した矩形フィルタ内の出力信号値の算術平均値との差分（矩形出力差分）を算出し取得する（Ｓ５）。
なお、前記矩形出力差分には、中央値を設定して加算してもよく（Ｓ６）、矩形出力差分の画像（矩形出力差分画像）を表示してもよい（Ｓ７）。
また、前記矩形フィルタに比べて金属帯長手方向の画素数を少なくした線形フィルタの画素数を設定し（Ｓ１１）、当該線形フィルタ画素内の基準画素を設定し（Ｓ１２）、線形基準画素出力値取得手段により当該基準画素の出力信号値を取得する（Ｓ１３）。線形フィルタの幅方向画素数は、前記矩形フィルタの幅方向画素数に一致させるとよい。
次に、線形平均化出力値取得手段により線形フィルタ内の画素の出力信号値の算術平均値を算出する（Ｓ１４）。
線形出力差分取得手段により、前記Ｓ１３で取得した基準画素の出力信号値と、前記Ｓ１４で取得した線形フィルタ内の出力信号値の算術平均値との差分（線形出力差分）を算出し取得する（Ｓ１５）。
なお、前記線形出力差分には、中央値を設定して加算してしてもよく（Ｓ１６）、線形出力差分の画像（線形出力差分画像）を表示してもよい（Ｓ１７）。
その後、出力差分差取得手段により、同一の基準画素について、前記Ｓ５で取得した矩形出力差分と前記Ｓ１５で取得した線形出力差分との差を算出し、その絶対値を取得する（Ｓ２１）。尚、中央値の加算を行った場合は、加算後の矩形出力差分と加算後の線形出力差分との差を算出する。
出力差分差画素抽出手段により、閾値を設定して、当該差分の差の絶対値を２値化する（Ｓ２２）。
最後に、判定手段により当該２値化して残存した画素（前記絶対値が閾値を超える画素、又は、前記絶対値が閾値以上の画素）を幅方向線状模様欠陥として判定する（Ｓ２３）。
FIG. 4 is a flowchart (part 1) of the image processing steps described above.
First, the image which imaged the metal belt surface is acquired (S0).
Next, the number of pixels of the rectangular filter is set (S1), the reference pixel in the rectangular filter pixel is set (S2), and the output signal value of the reference pixel is acquired by the rectangular reference pixel output value acquisition unit (S3). ).
Further, the arithmetic average value of the output signal values of the pixels in the rectangular filter is calculated by the rectangular average output value acquisition means (S4).
The rectangular output difference acquisition means calculates and acquires a difference (rectangular output difference) between the output signal value of the reference pixel acquired in S3 and the arithmetic average value of the output signal value in the rectangular filter acquired in S4 ( S5).
Note that a median value may be set and added to the rectangular output difference (S6), or an image of the rectangular output difference (rectangular output difference image) may be displayed (S7).
Further, the number of pixels of the linear filter in which the number of pixels in the metal band longitudinal direction is reduced compared to the rectangular filter is set (S11), the reference pixel in the linear filter pixel is set (S12), and the linear reference pixel output value is set. The output signal value of the reference pixel is acquired by the acquisition means (S13). The number of pixels in the width direction of the linear filter may match the number of pixels in the width direction of the rectangular filter.
Next, the arithmetic average value of the output signal values of the pixels in the linear filter is calculated by the linear averaged output value acquisition means (S14).
The linear output difference acquisition means calculates and acquires the difference (linear output difference) between the output signal value of the reference pixel acquired in S13 and the arithmetic average value of the output signal value in the linear filter acquired in S14 ( S15).
The linear output difference may be set by adding a median value (S16), or an image of the linear output difference (linear output difference image) may be displayed (S17).
Thereafter, for the same reference pixel, the difference between the rectangular output difference acquired in S5 and the linear output difference acquired in S15 is calculated by the output difference difference acquisition means, and the absolute value is acquired (S21). When the median is added, the difference between the rectangular output difference after addition and the linear output difference after addition is calculated.
The threshold value is set by the output difference difference pixel extraction means, and the absolute value of the difference difference is binarized (S22).
Finally, a pixel remaining after the binarization by the determination unit (the pixel whose absolute value exceeds the threshold value or the pixel whose absolute value exceeds the threshold value) is determined as a widthwise linear pattern defect (S23).
尚、上述した矩形基準画素出力値取得手段、矩形平均化出力値取得手段、矩形出力差分取得手段、線形基準画素出力値取得手段、線形平均化出力値取得手段、線形出力差分取得手段、出力差分差取得手段、出力差分差画素抽出手段、及び判定手段は、全て図１に示す同一の演算装置５内に構成されるようにすればよい。 The above-described rectangular reference pixel output value acquisition means, rectangular average output value acquisition means, rectangular output difference acquisition means, linear reference pixel output value acquisition means, linear averaged output value acquisition means, linear output difference acquisition means, output difference The difference acquisition unit, the output difference difference pixel extraction unit, and the determination unit may be configured in the same arithmetic device 5 shown in FIG.
次いで、本発明の実施形態について説明する。 Next, embodiments of the present invention will be described.
矩形出力差分画像を取得する過程では、基準画素の出力信号値から、矩形フィルタ内の出力信号値の平均値を差し引く（図４のＳ５）。具体例として、溶接部の転写痕（線状模様欠陥部）を含む金属帯表面の画像について、矩形出力差分画像を取得する（図７（ａ））。取得した矩形出力差分画像の出力信号値の大小を強調するために当該矩形出力差分画像について２値化処理を行った画像を図８（ａ）に示す。図８（ａ）中において、出力信号値の絶対値が閾値を超える部分を黒色、出力信号値の絶対値が閾値以内である部分を白色で示す。図８（ａ）では、幅方向（図面の左右方向）に延在する線状模様欠陥以外にも、金属帯表面全体にわたる模様等に由来するノイズ部分が多く、線状模様欠陥のみを精度よく検知することができない。 In the process of acquiring the rectangular output difference image, the average value of the output signal values in the rectangular filter is subtracted from the output signal value of the reference pixel (S5 in FIG. 4). As a specific example, a rectangular output difference image is acquired for the image of the surface of the metal band including the transfer mark (linear pattern defect portion) of the welded portion (FIG. 7A). FIG. 8A shows an image obtained by binarizing the rectangular output difference image in order to emphasize the magnitude of the output signal value of the acquired rectangular output difference image. In FIG. 8A, the portion where the absolute value of the output signal value exceeds the threshold is shown in black, and the portion where the absolute value of the output signal value is within the threshold is shown in white. In FIG. 8A, in addition to the linear pattern defect extending in the width direction (left-right direction of the drawing), there are many noise portions derived from the pattern over the entire surface of the metal band, and only the linear pattern defect is accurately detected. It cannot be detected.
線形出力差分画像を取得する過程では、基準画素の出力信号値から、線形フィルタの出力信号値の平均値を差し引く（図４のＳ１５）。取得した線形出力差分画像の出力信号値の大小を強調するために２値化処理した画像を、図８（ｂ）に示す。線状模様欠陥部のように幅方向に沿って出力信号値が変動しない箇所は、差分を取る過程で出力信号値同士が相殺されることになり、線形出力差分画像において強調されない。つまり、図８（ａ）に示す矩形出力差分画像の２値化処理画像と図８（ｂ）に示す線形出力差分画像の２値化処理画像とを比較するとわかるとおり、線形出力差分画像では、幅方向の線状模様欠陥以外のノイズ部分のみを強調することができる。 In the process of acquiring the linear output difference image, the average value of the output signal values of the linear filter is subtracted from the output signal value of the reference pixel (S15 in FIG. 4). FIG. 8B shows an image that has been binarized in order to emphasize the magnitude of the output signal value of the acquired linear output difference image. A portion where the output signal value does not vary along the width direction, such as a linear pattern defect portion, cancels out the output signal values in the process of taking the difference, and is not emphasized in the linear output difference image. That is, as can be seen by comparing the binarized image of the rectangular output difference image shown in FIG. 8A and the binarized image of the linear output difference image shown in FIG. Only noise portions other than the widthwise linear pattern defects can be emphasized.
すなわち、ノイズ部分と線状模様欠陥部との両方が強調された矩形出力差分画像と、ノイズ部分のみが強調された線形出力差分画像との差をとることによって得られる判定画像においては、線状模様欠陥部のみを強調することが可能となる。上記で取得した矩形出力差分画像（図７（ａ））と線形出力差分画像（図７（ｂ））との差を取得した画像を図９（ａ）に、さらに２値化した画像を図９（ｂ）に示す。このように、幅方向に延在する欠陥以外の部分は、差をとる過程で出力信号値の差分を相殺して殆ど０にすることができるので、幅方向に延在する欠陥については上述の線状模様のように僅かな凹凸であっても精度よく検出することができる。 That is, in the determination image obtained by taking the difference between the rectangular output difference image in which both the noise portion and the linear pattern defect portion are enhanced and the linear output difference image in which only the noise portion is enhanced, Only the pattern defect portion can be emphasized. FIG. 9A shows an image obtained by acquiring the difference between the rectangular output difference image (FIG. 7A) acquired above and the linear output difference image (FIG. 7B), and FIG. It is shown in 9 (b). As described above, since the difference other than the defect extending in the width direction can be made almost zero by offsetting the difference in the output signal value in the process of taking the difference, the defect extending in the width direction is described above. Even slight irregularities such as a linear pattern can be detected with high accuracy.
本発明に係る検査方法及び検査装置は、連続通板中の金属帯に特に好適に適用することができる。連続通板中の金属帯は、連続的又は断続的に画像取得手段によって金属帯の表面画像が取得されており、当該表面画像に対して本発明で示した画像処理を行うことで、幅方向線状模様欠陥のみを精度よく検出することができる。幅方向線状模様欠陥が検出された部分については、出荷を控える、或いは切断して除去するといった処置をとることができる。 The inspection method and inspection apparatus according to the present invention can be particularly suitably applied to a metal strip in a continuous sheet. As for the metal band in the continuous plate, the surface image of the metal band is acquired continuously or intermittently by the image acquisition means, and by performing the image processing shown in the present invention on the surface image, the width direction Only linear pattern defects can be detected with high accuracy. About the part where the width direction linear pattern defect was detected, measures, such as refraining from shipment or cutting and removing, can be taken.
なお、本発明は、図５にフローを示すとおり、図４のフローにおけるＳ２１の代わりに矩形出力差分と線形出力差分とをそれぞれ２値化し（矩形出力差分２値化手段：Ｓ３１、線形出力差分２値化手段：Ｓ４１）、Ｓ２２の代わりに２値化した矩形出力差分と２値化した線形出力差分との差を取得してもよい（２値化出力差分差取得手段：Ｓ５１）。
矩形出力差分取得手段では、矩形基準画素の出力信号値から、矩形フィルタ内の出力信号値の平均値を差し引く（図５のＳ５）。具体例として、溶接部の転写痕（線状模様欠陥部）を含む金属帯表面の画像について、矩形出力差分画像を取得する（図１１（ａ））。その後に、取得した矩形出力差分画像の出力信号値の２値化処理（図５のＳ３１）を行い、その画像を図１１（ｂ）に示す。図１１（ｂ）中において、出力信号値の絶対値が閾値を超える部分を黒色、出力信号値の絶対値が閾値以内である部分を白色で示す。
次に、線形出力差分取得手段では、線形基準画素の出力信号値から、線形フィルタの出力信号値の平均値を差し引く（図５のＳ１５）。具体例として、溶接部の転写痕（線状模様欠陥部）を含む金属帯表面の画像について、線形出力差分画像を取得する（図１２（ａ））。その後に、取得した線形出力差分画像の出力信号値の２値化処理（図５のＳ４１）を行い、その画像を図１２（ｂ）に示す。図１２（ｂ）中において、出力信号値の絶対値が閾値を超える部分を黒色、出力信号値の絶対値が閾値以内である部分を白色で示す。
さらに、ノイズ部分と線状模様欠陥部との両方が強調された矩形出力差分2値化画像と、ノイズ部分のみが強調された線形出力差分２値化画像との差をとる（図５のＳ５１）。これによって得られる判定画像を図１３に示す。図１３においては、線状模様欠陥部のみを強調することが可能となる。このように、幅方向に延在する欠陥以外の部分は、差をとる過程で出力信号値の差分を相殺して殆ど０にすることができるので、幅方向に延在する欠陥については上述の線状模様のように僅かな凹凸であっても精度よく検出することができる。
As shown in the flow of FIG. 5, the present invention binarizes the rectangular output difference and the linear output difference instead of S21 in the flow of FIG. 4 (rectangular output difference binarization means: S31, linear output difference). Binarization means: S41) Instead of S22, the difference between the binarized rectangular output difference and the binarized linear output difference may be acquired (binarization output difference difference acquisition means: S51).
In the rectangular output difference acquisition means, the average value of the output signal values in the rectangular filter is subtracted from the output signal value of the rectangular reference pixel (S5 in FIG. 5). As a specific example, a rectangular output difference image is acquired for the image of the surface of the metal band including the transfer mark (linear pattern defect portion) of the welded portion (FIG. 11A). Thereafter, binarization processing (S31 in FIG. 5) of the output signal value of the acquired rectangular output difference image is performed, and the image is shown in FIG. In FIG. 11B, the portion where the absolute value of the output signal value exceeds the threshold is shown in black, and the portion where the absolute value of the output signal value is within the threshold is shown in white.
Next, the linear output difference acquisition means subtracts the average value of the output signal value of the linear filter from the output signal value of the linear reference pixel (S15 in FIG. 5). As a specific example, a linear output difference image is acquired for the image of the surface of the metal band including the transfer mark (linear pattern defect portion) of the welded portion (FIG. 12A). Thereafter, the binarization process (S41 in FIG. 5) of the output signal value of the acquired linear output difference image is performed, and the image is shown in FIG. In FIG. 12B, the portion where the absolute value of the output signal value exceeds the threshold is shown in black, and the portion where the absolute value of the output signal value is within the threshold is shown in white.
Further, the difference between the rectangular output difference binarized image in which both the noise portion and the linear pattern defect portion are emphasized and the linear output difference binarized image in which only the noise portion is emphasized are taken (S51 in FIG. 5). ). FIG. 13 shows a determination image obtained as a result. In FIG. 13, only the linear pattern defect portion can be emphasized. As described above, since the difference other than the defect extending in the width direction can be made almost zero by offsetting the difference in the output signal value in the process of taking the difference, the defect extending in the width direction is described above. Even slight irregularities such as a linear pattern can be detected with high accuracy.
また、本発明は、図６にフローを示すとおり、図４のフローにおけるＳ３及びＳ１３の代わりに基準画素の出力信号値をそれぞれ２値化して（矩形基準画素出力値２値化手段：Ｓ６１、線形基準画素出力値２値化手段：Ｓ７１）、Ｓ４及びＳ１４の代わりに出力信号値の平均値をそれぞれ２値化して（矩形平均化出力値２値化手段：Ｓ６２、線形平均化出力値２値化手段：Ｓ７２）、Ｓ５及びＳ１５の代わりに前記２値化した基準画素の出力信号値と前記２値化した出力信号値の平均値との差分をそれぞれ取得し（矩形出力２値化差分取得手段：Ｓ６３、線形出力２値化差分取得手段：Ｓ７３）、矩形フィルタを用いて得られた２値化した差分と線形フィルタを用いて得られた２値化した差分との差を取得してもよい（出力２値化差分差取得手段：Ｓ８１）。
矩形基準画素出力値２値化手段では、矩形基準画素の出力信号値を２値化する（図６のＳ６１）。具体例として、溶接部の転写痕（線状模様欠陥部）を含む金属帯表面の画像について、矩形基準画素出力値２値化画像を取得する（図１４（ａ））。また、矩形平均化出力値２値化手段では、矩形フィルタ内の出力信号値の平均値を２値化する（図６のＳ６２）。具体例として、溶接部の転写痕（線状模様欠陥部）を含む金属帯表面の画像について、矩形平均化出力値２値化画像を取得する（図１４（ｂ））。次いで、前記２値化した基準画素の出力信号値と前記２値化した出力信号値の平均値との差分をそれぞれ取得する（矩形出力２値化差分取得手段：Ｓ６３）。矩形出力信号値を２値化した差分の具体例を図１５に示す。
線形基準画素出力値２値化手段では、線形基準画素の出力信号値を２値化する（図６のＳ７１）。具体例として、溶接部の転写痕（線状模様欠陥部）を含む金属帯表面の画像について、線形基準画素出力値２値化画像を取得する（図１６（ａ））。また、線形平均化出力値２値化手段では、線形フィルタ内の出力信号値の平均値を２値化する（図６のＳ７２）。具体例として、溶接部の転写痕（線状模様欠陥部）を含む金属帯表面の画像について、線形平均化出力値２値化画像を取得する（図１６（ｂ））。次いで、前記２値化した基準画素の出力信号値と前記２値化した出力信号値の平均値との差分をそれぞれ取得する（線形出力２値化差分取得手段：Ｓ７３）。線形出力信号値を２値化した差分の具体例を図１７に示す。
その後に、ノイズ部分と線状模様欠陥部との両方が強調された矩形出力２値化差分画像と、ノイズ部分のみが強調された線形出力２値化差分画像との差をとる（図６のＳ８１）。これによって得られる判定画像を図１８に示す。図１８においては、線状模様欠陥部のみを強調することが可能となる。このように、幅方向に延在する欠陥以外の部分は、差をとる過程で出力信号値の差分を相殺して殆ど０にすることができるので、幅方向に延在する欠陥については上述の線状模様のように僅かな凹凸であっても精度よく検出することができる。
Further, as shown in the flow of FIG. 6, the present invention binarizes the output signal value of the reference pixel instead of S3 and S13 in the flow of FIG. 4 (rectangular reference pixel output value binarization means: S61, Linear reference pixel output value binarizing means: S71), instead of S4 and S14, the average value of the output signal value is binarized (rectangular averaged output value binarizing means: S62, linear averaged output value 2) Binarization means: S72), instead of S5 and S15, obtain the difference between the binarized reference pixel output signal value and the average value of the binarized output signal value (rectangular output binarization difference) Acquisition means: S63, linear output binarized difference acquisition means: S73), obtains the difference between the binarized difference obtained using the rectangular filter and the binarized difference obtained using the linear filter (Acquire output binary difference difference) Stage: S81).
The rectangular reference pixel output value binarization means binarizes the output signal value of the rectangular reference pixel (S61 in FIG. 6). As a specific example, a rectangular reference pixel output value binarized image is acquired for the image of the surface of the metal band including the transfer mark (linear pattern defect portion) of the welded portion (FIG. 14A). Further, the rectangular averaged output value binarizing means binarizes the average value of the output signal values in the rectangular filter (S62 in FIG. 6). As a specific example, a rectangular averaged output value binarized image is acquired for an image of a metal band surface including a transfer mark (a linear pattern defect portion) of a welded portion (FIG. 14B). Next, a difference between the binarized output signal value of the reference pixel and the average value of the binarized output signal value is acquired (rectangular output binarized difference acquisition means: S63). A specific example of the difference obtained by binarizing the rectangular output signal value is shown in FIG.
The linear reference pixel output value binarization means binarizes the output signal value of the linear reference pixel (S71 in FIG. 6). As a specific example, a linear reference pixel output value binarized image is acquired for an image of a metal band surface including a transfer mark (linear pattern defect portion) of a welded portion (FIG. 16A). Further, the linear averaged output value binarizing means binarizes the average value of the output signal values in the linear filter (S72 in FIG. 6). As a specific example, a linear averaged output value binarized image is acquired for the image of the surface of the metal band including the transfer mark (linear pattern defect portion) of the welded portion (FIG. 16B). Next, the difference between the binarized output signal value of the reference pixel and the average value of the binarized output signal value is acquired (linear output binarized difference acquisition means: S73). A specific example of the difference obtained by binarizing the linear output signal value is shown in FIG.
After that, the difference between the rectangular output binarized difference image in which both the noise part and the linear pattern defect part are emphasized and the linear output binarized difference image in which only the noise part is emphasized are taken (FIG. 6). S81). The determination image obtained by this is shown in FIG. In FIG. 18, only the linear pattern defect portion can be emphasized. As described above, since the difference other than the defect extending in the width direction can be made almost zero by offsetting the difference in the output signal value in the process of taking the difference, the defect extending in the width direction is described above. Even slight irregularities such as a linear pattern can be detected with high accuracy.
尚、当然であるが、これらの２値化において、図５に示す矩形出力差分２値化手段（Ｓ３１）と線形出力差分２値化手段（Ｓ４１）で用いられる２値化の閾値は同じ値とし、図６に示す矩形基準画素出力値２値化手段（Ｓ６１）と線形基準画素出力値２値化手段（Ｓ７１）と矩形平均化出力値２値化手段（Ｓ６２）と線形平均化出力値２値化手段（Ｓ７２）の２値化の閾値は同じ値とする。 Of course, in these binarizations, the binarization threshold values used in the rectangular output difference binarization means (S31) and the linear output difference binarization means (S41) shown in FIG. The rectangular reference pixel output value binarizing means (S61), the linear reference pixel output value binarizing means (S71), the rectangular average output value binarizing means (S62) and the linear averaged output value shown in FIG. The binarization threshold value of the binarization means (S72) is the same value.
上述の説明の金属帯の金属としては、鋼、アルミニウム、銅等、及びこれらの合金を含み、さらにこれら金属の表面にめっきしたものを含んでもよい。また、画像取得手段として、リニア画像カメラ、２次元画像カメラなどを用いることができる。 As a metal of the metal strip of the above-mentioned description, steel, aluminum, copper, etc., and alloys thereof, and those plated on the surface of these metals may be included. Moreover, a linear image camera, a two-dimensional image camera, etc. can be used as an image acquisition means.
金属帯表面について、線状模様欠陥部を含む部分を画像取得手段により撮影し、本発明で開示した画像処理を行った。なお、平均化処理として算術平均を用いた。図７（ａ）及び（ｂ）は、それぞれ本実施例で得られた矩形出力差分画像及び線形出力差分画像である。また、図８（ａ）及び（ｂ）は、それぞれ本実験で得られた図７（ａ）の矩形出力差分画像と、図７（ｂ）の線形出力差分画像とを２値化処理して得られた画像である。 On the surface of the metal strip, a portion including a linear pattern defect portion was photographed by an image acquisition means, and the image processing disclosed in the present invention was performed. In addition, arithmetic average was used as the averaging process. FIGS. 7A and 7B are a rectangular output difference image and a linear output difference image obtained in this embodiment, respectively. FIGS. 8A and 8B are obtained by binarizing the rectangular output difference image of FIG. 7A obtained in this experiment and the linear output difference image of FIG. 7B, respectively. It is the obtained image.
本発明例として、矩形出力差分画像と線形出力差分画像との差を取った画像を図９（ａ）に、２値化した画像を図９（ｂ）に示す。図９（ｂ）では、画像中のうち出力信号値の絶対値が閾値を超える部分（異常部分）を黒色で、閾値以下である部分（正常部分）を白色で示した。図９（ｂ）の矢印の位置に示されるように、幅方向に延在する線状模様欠陥部のみを抽出することができた。 As an example of the present invention, an image obtained by taking a difference between a rectangular output difference image and a linear output difference image is shown in FIG. 9A, and a binarized image is shown in FIG. 9B. In FIG. 9B, a portion (abnormal portion) where the absolute value of the output signal value exceeds the threshold in the image is shown in black, and a portion (normal portion) that is equal to or less than the threshold is shown in white. As indicated by the position of the arrow in FIG. 9B, only the linear pattern defect extending in the width direction could be extracted.
比較例１として、従来の欠陥検知方法として広く用いられている、元の表面画像をそのまま２値化した結果を図１０に示す。線状模様欠陥と、それ以外の表面全体にわたる模様等に起因するノイズとを分離することが難しく、線状模様欠陥のみを検知することは困難であった。 As Comparative Example 1, the result of binarizing the original surface image, which is widely used as a conventional defect detection method, is shown in FIG. It was difficult to separate the linear pattern defect from the noise caused by the pattern over the entire other surface, and it was difficult to detect only the linear pattern defect.
また、比較例２として、矩形フィルタのみを用いて、基準画素の出力信号値と平均化した出力信号値との差分を２値化した結果を図８（ａ）に示す。矩形フィルタでの処理で得られた出力信号値の差分を２値化するのみでは、図１０に示す元の表面画像の場合と同様に、線状模様欠陥と、それ以外の表面全体にわたる模様に起因するノイズとの分離が難しいことが分かる。 Further, as Comparative Example 2, FIG. 8A shows the result of binarizing the difference between the output signal value of the reference pixel and the averaged output signal value using only the rectangular filter. Just by binarizing the difference between the output signal values obtained by processing with the rectangular filter, as in the case of the original surface image shown in FIG. It can be seen that it is difficult to separate from the resulting noise.
１ 鋼帯
２ 通板ロール
３ 光照射手段
４ 画像取得手段
５ 演算装置
DESCRIPTION OF SYMBOLS 1 Steel strip 2 Plate roll 3 Light irradiation means 4 Image acquisition means 5 Arithmetic unit
Claims (8)
画素数及び基準画素を設定した矩形フィルタを用いて、前記表面画像における基準画素の出力信号値を取得し、前記矩形フィルタ内の平均化処理した出力信号値を取得し、さらに前記基準画素の出力信号値と前記矩形フィルタ内の平均化処理した出力信号値との差分を矩形出力差分として取得し、
前記矩形フィルタに比べて前記金属帯の長手方向の画素数を少なくした線形フィルタを用いて、前記表面画像における基準画素の出力信号値を取得し、前記線形フィルタ内の平均化処理した出力信号値を取得し、さらに前記基準画素の出力信号値と前記線形フィルタ内の平均化処理した出力信号値との差分を線形出力差分として取得し、
その後、同一の基準画素について、前記矩形出力差分と前記線形出力差分との差の絶対値を取得するとともに、所定の閾値Ａを設定し、
出力差分の差の絶対値が前記所定の閾値Ａ以上又は閾値Ａを超える画素を抽出し、当該画素に対応する金属帯表面部位を幅方向線状模様欠陥として判定することを特徴とする金属帯表面の幅方向線状模様欠陥の検査方法。 Image the surface of the metal band to obtain the surface image,
Using a rectangular filter in which the number of pixels and the reference pixel are set, an output signal value of the reference pixel in the surface image is acquired, an output signal value obtained by averaging processing in the rectangular filter is acquired, and an output of the reference pixel is further obtained. The difference between the signal value and the output signal value averaged in the rectangular filter is acquired as a rectangular output difference,
Using a linear filter in which the number of pixels in the longitudinal direction of the metal strip is reduced compared to the rectangular filter, an output signal value of a reference pixel in the surface image is obtained, and an output signal value obtained by averaging processing in the linear filter Further, the difference between the output signal value of the reference pixel and the output signal value averaged in the linear filter is acquired as a linear output difference,
Thereafter, for the same reference pixel, the absolute value of the difference between the rectangular output difference and the linear output difference is acquired, and a predetermined threshold A is set,
A metal band characterized in that a pixel whose absolute value of the difference in output difference is equal to or greater than the predetermined threshold A or exceeds the threshold A is determined, and a metal band surface portion corresponding to the pixel is determined as a widthwise linear pattern defect. Inspection method for surface widthwise linear pattern defects.
前記線形出力差分を所定の閾値Ｂを用いて２値化し、
２値化した矩形出力差分と２値化した線形出力差分との差の絶対値を２値化出力差分差として取得して、当該絶対値が所定値Ｃ以上である画素に対応する金属帯表面部分を幅方向線状模様欠陥として判定することを特徴とする請求項１に記載の金属帯表面の幅方向線状模様欠陥の検査方法。 The rectangular output difference is binarized using a predetermined threshold B,
Binarizing the linear output difference using a predetermined threshold B;
The absolute value of the difference between the binarized rectangular output difference and the binarized linear output difference is obtained as the binarized output difference difference, and the metal band surface corresponding to the pixel whose absolute value is equal to or greater than the predetermined value C 2. The method for inspecting a width direction linear pattern defect on a metal strip surface according to claim 1, wherein the portion is determined as a width direction linear pattern defect.
前記矩形フィルタ内の平均化処理した出力信号値を前記所定の閾値Ｄを用いて２値化し、
前記矩形フィルタ内の前記基準画素の出力信号２値化値と前記平均化処理した出力信号２値化値との差を矩形出力２値化差分として取得し、
前記線形フィルタ内の基準画素の出力信号値を前記所定の閾値Ｄを用いて２値化し、前記線形フィルタ内の平均化処理した出力信号値を前記所定の閾値Ｄを用いて２値化し、前記線形フィルタ内の前記基準画素の出力信号２値化値と前記平均化処理した出力信号２値化値との差を線形出力２値化差分として取得し、
前記矩形出力２値化差分と前記線形出力２値化差分との差の絶対値を取得して、
当該絶対値が所定値Ｅ以上である画素に対応する金属帯表面部位を幅方向線状模様欠陥として判定することを特徴とする請求項１に記載の金属帯表面の幅方向線状模様欠陥の検査方法。 The output signal value of the reference pixel in the rectangular filter is binarized using a predetermined threshold D,
The averaged output signal value in the rectangular filter is binarized using the predetermined threshold D,
Obtaining a difference between the binarized output signal value of the reference pixel in the rectangular filter and the binarized output signal binarized value as a rectangular output binarized difference;
The output signal value of the reference pixel in the linear filter is binarized using the predetermined threshold D, the averaged output signal value in the linear filter is binarized using the predetermined threshold D, and Obtaining a difference between the output signal binarized value of the reference pixel in the linear filter and the averaged output signal binarized value as a linear output binarized difference;
Obtain the absolute value of the difference between the rectangular output binarized difference and the linear output binarized difference;
2. The width direction linear pattern defect on the metal band surface according to claim 1, wherein a metal band surface portion corresponding to a pixel whose absolute value is equal to or greater than a predetermined value E is determined as a width direction linear pattern defect. Inspection method.
画素数および基準画素を設定した矩形フィルタを用いて、前記表面画像における基準画素の出力信号値を取得する矩形基準画素出力値取得手段と、前記矩形フィルタ内の平均化処理した出力信号値を取得する矩形平均化出力値取得手段と、前記矩形基準画素出力値取得手段で取得した基準画素の出力信号値と前記矩形平均化出力値取得手段で取得した矩形フィルタ内の平均化処理した出力信号値との差分を矩形出力差分として取得する矩形出力差分取得手段と、
前記矩形フィルタに比べて前記金属帯の長手方向の画素数を少なくした線形フィルタを用いて、前記表面画像における基準画素の出力信号値を取得する線形基準画素出力値取得手段と、前記線形フィルタ内の平均化処理した出力信号値を取得する線形平均化出力値取得手段と、前記線形基準画素出力値取得手段で取得した基準画素の出力信号値と前記線形平均化出力値取得手段で取得した線形フィルタ内の平均化処理した出力信号値との差分を線形出力差分として取得する線形出力差分取得手段と、
同一の基準画素について、前記矩形出力差分と前記線形出力差分との差の絶対値を取得する出力差分差取得手段と、
所定の閾値Ａを設定するとともに、前記出力差分差取得手段により取得した出力差分の差の絶対値が、前記所定の閾値Ａを超える画素を抽出する出力差分差画素抽出手段と、
当該出力差分差画素抽出手段で抽出した画素に対応する金属帯表面部位を幅方向線状模様欠陥として判定する判定手段と
を備えることを特徴とする金属帯表面の幅方向線状模様欠陥の検査装置。 An image acquisition means for imaging the surface of the metal strip and acquiring a surface image;
Using a rectangular filter in which the number of pixels and the reference pixel are set, a rectangular reference pixel output value acquisition means for acquiring an output signal value of the reference pixel in the surface image, and an output signal value averaged in the rectangular filter are acquired. A rectangular averaged output value acquiring unit, an output signal value of the reference pixel acquired by the rectangular reference pixel output value acquiring unit, and an output signal value obtained by averaging in the rectangular filter acquired by the rectangular averaged output value acquiring unit A rectangular output difference acquisition means for acquiring the difference between and as a rectangular output difference;
Linear reference pixel output value acquisition means for acquiring an output signal value of a reference pixel in the surface image using a linear filter in which the number of pixels in the longitudinal direction of the metal band is reduced compared to the rectangular filter, and in the linear filter Linear averaged output value acquisition means for acquiring the output signal value after averaging processing, the output signal value of the reference pixel acquired by the linear reference pixel output value acquisition means, and the linear acquired by the linear averaged output value acquisition means Linear output difference acquisition means for acquiring the difference between the averaged output signal value in the filter as a linear output difference;
For the same reference pixel, output difference difference acquisition means for acquiring the absolute value of the difference between the rectangular output difference and the linear output difference;
An output difference difference pixel extracting unit that sets a predetermined threshold A and extracts pixels in which the absolute value of the difference between the output differences acquired by the output difference difference acquiring unit exceeds the predetermined threshold A;
A determination means for determining a metal band surface portion corresponding to the pixel extracted by the output difference difference pixel extraction means as a width direction linear pattern defect, and a width direction linear pattern defect inspection on the metal band surface apparatus.
前記線形出力差分取得手段で取得した線形出力差分を所定の閾値Ｂを用いて２値化する線形出力差分２値化手段と、
２値化した矩形出力差分と２値化した線形出力差分との差の絶対値を取得する２値化出力差分差取得手段と、
当該絶対値が所定値Ｃ以上である画素に対応する金属帯表面部分を幅方向線状模様欠陥として判定する判定手段と、
を備えることを特徴とする請求項５に記載の金属帯表面の幅方向線状模様欠陥の検査装置。 A rectangular output difference binarization means for binarizing the rectangular output difference acquired by the rectangular output difference acquisition means using a predetermined threshold B;
Linear output difference binarization means for binarizing the linear output difference acquired by the linear output difference acquisition means using a predetermined threshold B;
Binarized output difference difference acquisition means for acquiring an absolute value of a difference between the binarized rectangular output difference and the binarized linear output difference;
Determination means for determining a metal strip surface portion corresponding to a pixel whose absolute value is equal to or greater than a predetermined value C as a widthwise linear pattern defect;
The inspection apparatus of the width direction linear pattern defect of the metal strip surface of Claim 5 characterized by the above-mentioned.
前記矩形平均化出力値取得手段で取得した矩形フィルタ内の平均化処理した出力信号値を前記所定の閾値Ｄを用いて２値化する線形基準画素２値化手段と、
前記矩形フィルタ内の平均化処理した出力信号値を２値化した値と前記基準画素の出力信号値を２値化した値との差を矩形出力２値化差分として取得する矩形出力２値化差分取得手段と、
前記線形フィルタ内の平均化処理した出力信号値を２値化した値と前記基準画素の出力信号値を２値化した値との差を線形出力２値化差分として取得する線形出力２値化差分取得手段と、
前記矩形出力２値化差分と前記線形出力２値化差分との差の絶対値を取得する出力２値化差分差取得手段と、
当該絶対値が所定値Ｅ以上である画素に対応する金属帯表面部位を幅方向線状模様欠陥として判定する判定手段と、
を備えることを特徴とする請求項５に記載の金属帯表面の幅方向線状模様欠陥の検査装置。 Rectangular reference pixel output binarization means for binarizing the output signal value of the reference pixel acquired by the rectangular reference pixel output value acquisition means using a predetermined threshold D;
Linear reference pixel binarization means for binarizing the output signal value averaged in the rectangular filter acquired by the rectangular average output value acquisition means using the predetermined threshold D;
Rectangular output binarization for obtaining a difference between a binarized value of the output signal value obtained by averaging processing in the rectangular filter and a binarized value of the output signal value of the reference pixel as a rectangular output binarized difference Difference acquisition means;
Linear output binarization that obtains a difference between a binarized value of the averaged output signal value in the linear filter and a binarized value of the output signal value of the reference pixel as a linear output binarized difference Difference acquisition means;
Output binarized difference difference acquisition means for acquiring an absolute value of a difference between the rectangular output binarized difference and the linear output binarized difference;
Determination means for determining a metal strip surface portion corresponding to a pixel whose absolute value is equal to or greater than a predetermined value E as a widthwise linear pattern defect;
The inspection apparatus of the width direction linear pattern defect of the metal strip surface of Claim 5 characterized by the above-mentioned.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2018042554 | 2018-03-09 | ||
| JP2018042554 | 2018-03-09 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| JP2019158857A true JP2019158857A (en) | 2019-09-19 |
| JP6809510B2 JP6809510B2 (en) | 2021-01-06 |
Family
ID=67996978
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP2018121415A Active JP6809510B2 (en) | 2018-03-09 | 2018-06-27 | Inspection method and inspection device for widthwise linear pattern defects on the surface of metal strips |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JP6809510B2 (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS57184956A (en) * | 1981-05-08 | 1982-11-13 | Toshiba Corp | Inspecting device of surface defect |
| JPS59138904A (en) * | 1983-01-28 | 1984-08-09 | Nippon Steel Corp | Checking method of surface defect of running plate body |
| JPH09138201A (en) * | 1995-11-15 | 1997-05-27 | Mitsubishi Electric Corp | Surface inspection apparatus |
| JP2000028543A (en) * | 1998-07-13 | 2000-01-28 | Dainippon Printing Co Ltd | Pattern inspecting method and device therefor |
| JP2003329596A (en) * | 2002-05-10 | 2003-11-19 | Mitsubishi Rayon Co Ltd | Apparatus and method for inspecting defect |
| US20090022391A1 (en) * | 2005-07-01 | 2009-01-22 | Abb Oy | Method and Product for Detecting Abnormalities |
-
2018
- 2018-06-27 JP JP2018121415A patent/JP6809510B2/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS57184956A (en) * | 1981-05-08 | 1982-11-13 | Toshiba Corp | Inspecting device of surface defect |
| JPS59138904A (en) * | 1983-01-28 | 1984-08-09 | Nippon Steel Corp | Checking method of surface defect of running plate body |
| JPH09138201A (en) * | 1995-11-15 | 1997-05-27 | Mitsubishi Electric Corp | Surface inspection apparatus |
| JP2000028543A (en) * | 1998-07-13 | 2000-01-28 | Dainippon Printing Co Ltd | Pattern inspecting method and device therefor |
| JP2003329596A (en) * | 2002-05-10 | 2003-11-19 | Mitsubishi Rayon Co Ltd | Apparatus and method for inspecting defect |
| US20090022391A1 (en) * | 2005-07-01 | 2009-01-22 | Abb Oy | Method and Product for Detecting Abnormalities |
Also Published As
| Publication number | Publication date |
|---|---|
| JP6809510B2 (en) | 2021-01-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10002418B2 (en) | Laser beam welding diagnosis apparatus and laser beam welding diagnosis method | |
| JP7028091B2 (en) | Surface defect detection device and surface defect detection method | |
| JP2000036044A (en) | Defect integrating processor and defect integrating processing method | |
| CN113196040A (en) | Surface defect detection method, surface defect detection device, steel product manufacturing method, steel product quality management method, steel product manufacturing facility, surface defect determination model generation method, and surface defect determination model | |
| JP2012251983A (en) | Wrap film wrinkle inspection method and device | |
| JP4244046B2 (en) | Image processing method and image processing apparatus | |
| JP4331558B2 (en) | Appearance inspection method and appearance inspection apparatus for inspection object | |
| JP5347661B2 (en) | Belt surface inspection apparatus, surface inspection method, and program | |
| JP2017062181A (en) | Surface flaw checkup apparatus, and surface flaw checkup method | |
| JP2014119432A (en) | Inspection device and inspection method | |
| KR20160097651A (en) | Apparatus and Method for Testing Pattern of Sample using Validity Image Processing Technique, and computer-readable recording medium with program therefor | |
| JP2019158857A (en) | Method and apparatus for inspecting linear pattern defect in width direction of metal strip surface | |
| JP2010287442A (en) | Inspection method for battery, and inspection device for battery | |
| JP6035124B2 (en) | Defect inspection apparatus and defect inspection method | |
| JP2017148125A (en) | Image processing device, image processing method, and program | |
| JP5533019B2 (en) | Surface inspection apparatus and method | |
| JP2005265467A (en) | Defect detection device | |
| JP5563936B2 (en) | Weld bead quality inspection method | |
| JP2008196866A (en) | Weld crack detection method and device | |
| JP5786631B2 (en) | Surface defect inspection equipment | |
| JP4563183B2 (en) | Inspection method and apparatus for streaky unevenness defect | |
| JP2017181136A (en) | Surface defect detection method and surface defect detection device | |
| JP2005265828A (en) | Flaw detection method and apparatus | |
| JP2009074828A (en) | Flaw detection method and flaw detector | |
| JP2021067588A (en) | Surface inspection device for object to be inspected and surface inspection method for object to be inspected |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| RD04 | Notification of resignation of power of attorney |
Free format text: JAPANESE INTERMEDIATE CODE: A7424 Effective date: 20190327 |
|
| A621 | Written request for application examination |
Free format text: JAPANESE INTERMEDIATE CODE: A621 Effective date: 20191025 |
|
| A977 | Report on retrieval |
Free format text: JAPANESE INTERMEDIATE CODE: A971007 Effective date: 20201009 |
|
| TRDD | Decision of grant or rejection written | ||
| A01 | Written decision to grant a patent or to grant a registration (utility model) |
Free format text: JAPANESE INTERMEDIATE CODE: A01 Effective date: 20201110 |
|
| A61 | First payment of annual fees (during grant procedure) |
Free format text: JAPANESE INTERMEDIATE CODE: A61 Effective date: 20201123 |
|
| R150 | Certificate of patent or registration of utility model |
Ref document number: 6809510 Country of ref document: JP Free format text: JAPANESE INTERMEDIATE CODE: R150 |
|
| R250 | Receipt of annual fees |
Free format text: JAPANESE INTERMEDIATE CODE: R250 |