JP4209525B2 - Method for measuring the thickness of pantograph strips - Google Patents

Method for measuring the thickness of pantograph strips Download PDF

Info

Publication number
JP4209525B2
JP4209525B2 JP35901098A JP35901098A JP4209525B2 JP 4209525 B2 JP4209525 B2 JP 4209525B2 JP 35901098 A JP35901098 A JP 35901098A JP 35901098 A JP35901098 A JP 35901098A JP 4209525 B2 JP4209525 B2 JP 4209525B2
Authority
JP
Japan
Prior art keywords
hull
boundary
sliding plate
edge
image
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
JP35901098A
Other languages
Japanese (ja)
Other versions
JP2000180128A (en
Inventor
裕三 北岡
俊繁 永尾
誠 瀬藤
寧健 大槻
秀一 香坂
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Mitsubishi Electric Corp
West Japan Railway Co
Original Assignee
Mitsubishi Electric Corp
West Japan Railway Co
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Mitsubishi Electric Corp, West Japan Railway Co filed Critical Mitsubishi Electric Corp
Priority to JP35901098A priority Critical patent/JP4209525B2/en
Publication of JP2000180128A publication Critical patent/JP2000180128A/en
Application granted granted Critical
Publication of JP4209525B2 publication Critical patent/JP4209525B2/en
Anticipated expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Images

Landscapes

  • Electric Propulsion And Braking For Vehicles (AREA)
  • Length Measuring Devices By Optical Means (AREA)

Description

【0001】
【発明の属する技術分野】
この発明は、鉄道車両のパンタグラフに搭載されトロリ線と接触して集電するすり板の使用限度超過や異常摩耗を検知するための、非接触によるパンタグラフすり板厚み測定方法に関するものである。
【0002】
【従来の技術】
図42は例えば特開平9−265524号公報に開示された従来のパンタグラフすり板厚み測定装置を示す構成図である。
図42において、201は走行する列車、202はパンタグラフ、203は列車検知センサ、204はパンタグラフ202を撮影するランダムシャッターカメラ、205はパンタグラフ検知センサ、206はパンタグラフ202を照明する照明、207は画像処理装置の本体である。この画像処理装置207はランダムシャッターカメラ204よりパンタグラフ202のアナログの映像信号を入力してデジタル変換する画像入力手段208、デジタル変換された映像信号を保存する保存手段209、保存された映像信号を表示用信号に処理する表示手段210、装置全体を制御する制御手段211、撮影されたパンタグラフ202の画像を処理してパンタグラフ202の厚みを計測する画像処理手段212より構成されている。213は撮影されたパンタグラフ202の画像を画像表示するモニタ、矢印216は列車進行方向である。
【0003】
次に従来装置の動作について説明する。
矢印216で示される方向から列車201が進入し、列車検知センサ203により列車の進入が検知された時、画像処理装置207の制御手段211は、パンタグラフ202を撮影するため照明206を点灯するとともに、ランダムシャッターカメラ204の絞りを制御する。そして、パンタグラフ検知センサ205がパンタグラフ202を検知する毎に、画像入力手段208にランダムシャッターカメラ204からのパンタグラフ202のアナログの映像信号を入力し、画像入力手段208でアナログの映像信号をディジタル変換して保存手段209に画像データ(映像信号をディジタル変換したデータを画像データと呼ぶ)を保存する。
【0004】
保存手段209に保存された画像データは列車通過後、画像処理手段212でパンタグラフ203の厚みを算出する。画像処理手段212による厚み算出は、舟体とすり板の境界を一本の直線で求め、その直線とすり板上面のエッジ間の画素を厚みに換算している。
【0005】
【発明が解決しようとする課題】
従来のパンタグラフすり板厚み測定装置は以上のように構成されており、太陽光の影響や汚れによって舟体とすり板の境界が見えにくい場合や、舟体が湾曲して舟体とすり板の境界を1本の直線で求められない舟体の場合、あるいは撮影タイミングがずれ画像上の舟***置がずれた場合等に対処する方法が示されておらず、これらの場合にすり板の厚みを正確に算出することができない。また、すり板上面の異常摩耗を検知する方法も示されていない。
【0006】
この発明は上記のような問題点を解消するためになされたもので、舟体とすり板の境界が見えない場合、湾曲した舟体の場合、撮影タイミングがずれ画像上の舟***置がずれた場合等にパンタグラフのすり板厚みを測定し、しかも異常摩耗を検知することができるパンタグラフすり板厚み測定方法を提供することを目的とする。
【0007】
【課題を解決するための手段】
発明に係るパンタグラフすり板の厚み測定方法は、太陽光や照明の強度と方向など明暗環境の異なる条件下において撮影された各種舟体の画像をテンプレートとして複数用意し、検査対象とする撮影画像に対してパターンマッチング処理を行なって最も一致度の大きいテンプレートとその位置から検査対象画像上に写った舟体の上面とすり板下面の境界を検出するための範囲を設定する。
【0010】
発明に係るパンタグラフすり板の厚み測定方法は、太陽光や照明の強度と方向など明暗環境の異なる条件下において撮影された各種舟体の画像をテンプレートとして複数用意し、検査対象とする撮影画像に対してパターンマッチング処理を行なって最も一致度の大きいテンプレートとその位置から検査対象画像上に写った舟体の下面を検出するための範囲を設定する。
【0014】
発明に係るパンタグラフすり板の厚み測定方法は、舟体長手方向の各位置毎にすり板上面側から舟体上面とすり板下面の境界側に輝度の差分をとり、正の最大値を示した箇所Aを求め、そのA点からすり板上面側に一定の範囲内で差分値がもっとも大きく、しかも所定の値より大きい箇所BがA点以外に存在するとき、B点を舟体とすり板の境界の候補点とし、また、B点が見いだせない時は、A点を候補点とする。
【0015】
発明に係るパンタグラフすり板の厚み測定方法は、舟体とすり板の境界の候補点から舟体下面方向に所定の輝度値より明るい画素が所定の距離以上連続して存在する条件を満たす候補点が、長手方向に所定以上連続して存在するとき、この条件を満たす候補点全てを舟体上面とすり板下面の境界の候補点から除外する。
【0016】
発明に係るパンタグラフすり板の厚み測定方法は、すり板下面境界抽出範囲を舟体長手方向に複数ブロックに分け、そのブロック毎に舟体上面とすり板下面の境界の候補点を代表する直線を求め、そして隣接するブロックの直線の交点で連結した折れ線を舟体上面とすり板下面との境界とする。
【0017】
発明に係るパンタグラフすり板の厚み測定方法は、舟体長手方向に複数ブロックに分けて、各ブロック毎に舟体上面とすり板下面の境界候補点を代表する直線を求め、その直線を求めた候補点数が所定値以上となった直線のみを用いて舟体上面とすり板下面の境界を設定するが、所定値未満の区間が所定値以上の区間にはさまれた場合には、隣接する区間の直線どうしの交点で連結するものとし、所定値未満の区間が端の場合は隣接する区間の直線を延長し、得られた折れ線を舟体とすり板の境界とする。
【0018】
発明に係るパンタグラフすり板の厚み測定方法は、舟体長手方向に複数ブロックに分けて、各ブロック毎に舟体上面とすり板下面の境界候補点を代表する直線を求め、その直線を求めた候補点数が所定値以上となった直線のみを用いて舟体上面とすり板下面の境界を設定するが、所定値未満の区間が舟体中央部分であるとき、舟体中央部分の下端を抽出し求めた直線を所定の距離だけ上方向に平行移動したものを舟体中央付近のブロック内の直線として採用する。
【0019】
発明に係るパンタグラフすり板の厚み測定方法は、舟体長手方向に複数ブロックに分けて、各ブロック毎に舟体上面とすり板下面の境界候補点を代表する直線を求め、その直線を求めた候補点数が所定値以上となった直線のみを用いて舟体上面とすり板下面の境界を設定するが、所定値未満の区間が舟体中央部分であるとき、舟体中央部分の下端を抽出してその平均値を求め、その平均位置を所定の距離だけ上方向に平行移動した点と、隣接するブロック内の直線端から引いた前記点を通る直線を舟体中央付近のブロック内の直線とする。
【0020】
発明に係るパンタグラフすり板の厚み測定方法は、舟体上面とすり板下面の境界部分から舟体下面まで遮蔽する長方形状の遮蔽板が装着された舟体に対しては、その遮蔽板の舟体下面側の横方向の1辺のエッジを抽出し求めた直線を所定の距離だけ上方向に平行移動したものを舟体上面とすり板下面の境界とする。
【0021】
発明に係るパンタグラフすり板の厚み測定方法は、撮影した舟体画像に対して、舟体上面とすり板下面の境界部分から舟体下面まで遮蔽する長方形状の遮蔽板の舟体下面側の横方向の1辺のエッジを抽出し求めた直線の傾きで、舟体上面とすり板下面の境界が遮蔽板に隠されていない部分の境界の平均位置を通る直線を舟体上面とすり板下面の境界とする。
【0022】
発明に係るパンタグラフすり板の厚み測定方法は、撮影した舟体画像に対して、舟体上面とすり板下面の境界部分から舟体下面まで遮蔽する長方形状の遮蔽板の舟体下面側の横方向の1辺のエッジを抽出し平均位置を求め、その平均位置を所定の寸法だけ上方向に移動した点と、舟体上面とすり板下面の境界が遮蔽板に隠されていない部分の境界の平均位置を通る直線を舟体上面とすり板下面の境界とする。
【0031】
【発明の実施の形態】
実施の形態1.
図1はこの発明の実施の形態1によるパンタグラフすり板厚み測定装置の構成図である。1は走行する列車、2は被撮影体である列車屋根上のパンタグラフ、3は列車1の車輪を検知する車輪検知センサ、4は撮影手段でありパンタグラフ2の舟体両端の舟体前側面と後側面をそれぞれ撮影するランダムシャッターカメラ、5はパンタグラフ2の舟体を検知するパンタグラフ検知センサ、6は撮影領域を照らす照明、7は画像処理装置、8は任意のタイミングで画像入力したランダムシャッターカメラ4からの映像信号をディジタルデータに変換する画像入力手段、9は画像入力手段8より入力したデータを保存する保存手段、10は画像信号を出力する表示手段、11はパンタグラフすり板厚み測定装置全体を制御する制御手段、12は入力された画像データを画像処理する画像処理手段、13は表示手段10から画像信号が入力され画像を表示するモニタ、14は列車進行方向である。尚、画像処理手段12は後述する各実施の形態におけるパンタグラフの厚み測定方法に係わる処理の実施手段においては共通である。
【0032】
次に動作について説明する。列車1は矢印14で示す方向から進入し、車輪検知センサ3bが車輪検知センサ3aより先に車輪を検知すると、列車1が進入したと制御手段11が判断する。車輪検知センサ3aが車輪検知センサ3bより先に車輪を検知すると列車1が出庫したと制御手段11が判断する。列車1が入庫したと判断した場合、以下の測定動作を行う。
【0033】
まず、照明6を点灯し、次に列車1が進み、パンタグラフ検知センサ5がパンタグラフ2を検知すると、画像入力手段8はランダムシャッターカメラ4a、4b、4c、4dを同時に起動し、ランダムシャッターカメラ4c、4dはパンタグラフ2の舟体15(図2を参照)後側のすり板17の側面をパンタグラフ2の舟体長手方向に2分割して撮影し、ランダムシャッターカメラ4a、4bはパンタグラフ2の舟体15前側のすり板17の面をパンタグラフ2の舟体長手方向に2分割して撮影する。
【0034】
撮影された舟体15の映像信号は画像入力手段8に入力され、画素ごとにその明るさに比例した0〜255のディジタルデータに変換され、保存手段9に保存される。保存された画像データは明るい画素ほど値が255に近くなり、255が一番明るい値となる。
【0035】
列車1が進入しするとき、列車1に搭載された全てのパンタグラフ2の舟体15をこのように順次ランダムシャッターカメラ4が撮影し画像入力手段8に入力して保存手段9に保存する。
【0036】
最後に、列車通過終了後、制御手段11が照明6を消灯する。
また、表示手段10は、必要に応じて以下の画像処理の状態をモニタ13に表示する。
【0037】
列車1が通過すると画像入力手段8で入力され保存された舟体15の画像データに対して、画像処理手段12が各画像毎に画像処理を実施する。
まず、画像データ上から舟体15の位置検出のためパターンマッチングを行う。
図3に示すテンプレート画像19は、カメラ毎に、そして、金属すり板、カーボンすり板毎に、さらに舟体15に照射される太陽光や照明の強度と方向など明暗環境の異なる条件下において撮影され、図2,3に示すようなすり板17の上面が明るく白く写った画像、すり板17の上面が暗く黒く写った画像、すり板17の大きく摩耗した画像等、色々な舟体状態の画像が複数枚用意されている。尚、図3に示すテンプレート画像19は、舟体部分を切り出し、しかも1/8に縮小した画像である。
【0038】
次に、パターンマッチングの動作を図4のフローチャートおよび図5に示すパターンマッチングの概要に基づいて説明する。先ず、対象画像を入力した後に測定対象とする画像を1/8に縮小し(ステップS400、401)、その画像に対して用意された複数枚のテンプレート画像19(No.1〜No.N)を順次処理していく(ステップS402)。
【0039】
その際に、まず、対象とする画像をさらに1/2に縮小(元サイズの1/16)し、対象画像の左上角からx、y方向にテンプレート画像19を移動して、最も相関の高い位置(x0、y0)を求める。そして、その位置で元画像の1/8サイズの画像に対して、周囲25画素で精密にサーチを行い、そこで、最も相関の高い位置を求める(ステップS406〜409)。
【0040】
すべてのテンプレート画像19(No.1〜No.N)でこれら処理を行い、テンプレート画像19(No.1〜No.N)のうち最大の相関値Qmaxとそのときの位置(x,y)を出力する(ステップS403)。そして、相関値のQの相対値が60%以上であるか判定する(ステップS404)。相関値が60%以上であればテンプレート画像19の位置(x0、y0)が画像上の舟***置として採用する(ステップS405)。
【0041】
即ち、図5に示すように、対象画像(I)におて位置(x,y)+テンプレート(T)での検査範囲位置(x,y)+寸法のところを検査範囲とする。しかし、相関値が60%以上でなければ、すり板上面エッジの検出でサーチし(ステップS410)、画像上の舟***置を求めて検査範囲を設定する。
【0042】
パターンマッチングでマッチしたテンプレート画像19の1点25から舟体端の所まで舟体長手方向に寸法OFFSET_Y22、およびすり板17の上面を含む所まで舟体下面18の方向に寸法OX23だけ移動した所を基準位置として、図6に示すように基準位置から舟体長手方向に舟体端から舟体中央側まで長さL1と、すり板厚み方向に舟体下面18まで達しないが舟体上面とすり板下面の境界15Aを含む幅W1を持った範囲を設定し、その範囲をすり板厚み算出、異常検出を行う検査範囲20とする。
【0043】
この範囲は舟体ごとの個体差や通過時の傾きなどの違いによらず検査すべき範囲を含む最小のサイズとする。このように限定された範囲内で処理を行なうことで無関係な模様などの誤検出の可能性を最小限にできる。
この舟体長手方向への寸法OFFSET_Y22と厚み方向への寸法OX23は、形状の違うテンプレート画像19毎に最適な値を設定する。
【0044】
実施の形態2.
実施の形態1のパターンマッチングで最も高い相関値が60%未満の場合は以下の処理を行う。
図7は金属すり板の舟体15の図であり、図8,9は金属すり板上面エッジから撮影画像上の舟体15を検知するフローを示している。
【0045】
先ず、検査範囲を見つけた後に(ステップS500)、エッジレベルを15に設定する(ステップS)。そして、最初エッジレベルは4以上であるから次のステップS503へ進む(ステップS502)。
【0046】
ステップS502では、すり板17の向こう側エッジより背景の方が暗く撮影されており画像右から左方向に見て、すり板17の向こう側エッジは輝度の暗い部分から明るい部分に変化しているため、図7の画像右端から検出を開始し、画像の右から左に向かって隣の画素の輝度値が暗い輝度から明るい輝度に変化する差が15以上となるエッジを画像左端まで求める。そして、このような処理を舟体長手方向のすべての位置で検出する(ステップS503)。
【0047】
その検出されたエッジが、すり板厚み方向(画面左右方向)に32画素の幅で、舟体長手方向(画面上下方向)に8分割した各ブロック内でエッジがN個以上存在し、しかも、そのブロックが6ブロック以上存在する時、その位置をすり板上面のエッジAとする(ステップS504)。また、ブロックが6ブロック以上存在しない時は、エッジレベルを15から1ずつ4まで減らして、すり板17上面エッジAを求める(ステップS501A)。
【0048】
次にすり板17の上面エッジAから画像の左方向に舟体15の幅分85mm先の地点で画像の左方向に向かって画素の輝度値が明るい輝度から暗い輝度になるエッジを検出する(ステップS505)。これをすべての舟体長手方向の位置に対して行う。これは、すり板17の上面よりすり板17の側面の方が暗く撮影されるため画像右から左方向に見て、すり板上面エッジ26(図6参照)は輝度の明るい部分から暗い部分に変化しているためである。
【0049】
このようにして検出されたエッジが、すり板厚み方向に32画素の幅で、舟体長手方向に8分割した各ブロック内でエッジがN個以上存在し、しかも、そのブロックが6ブロック以上存在する時、そこに直線成分が存在するとする。そして、その位置をすり板17の上面エッジBとする(ステップS506)。また、ブロックが6ブロック以上存在しない時は、エッジレベルを14から1ずつ4まで減らして、すり板17の上面エッジAから求め直す(ステップS501A)。
【0050】
次に前記方法で求め、上面エッジBが存在したすり板17の上面エッジAから画像の左方向に向かってすり板17の幅25mm先の地点で画素の輝度値が明るい輝度から暗い輝度になるエッジを検出する(ステップS507)。これをすべての舟体長手方向の位置に対して行う。これは、すり板17の上面よりすり板17の側面の方が暗く撮影されるため画像右から左方向に見て、すり板の上面エッジcの所は輝度の明るい部分から暗い部分に変化しているためである。
【0051】
このようにして検出されたエッジが、すり板厚み方向に32画素の幅で、舟体長手方向に8分割した各ブロック内でエッジがN個以上存在し、しかも、そのブロックが6ブロック以上存在した時、そこに直線成分が存在するとする。その位置をすり板17の上面エッジCとする(ステップS508:図9を参照)。そして、すり板17の上面エッジAが舟体15の向こう側のエッジであるとし、その位置が撮影された画像上での舟***置であるとする(ステップS509)。すり板17の上面エッジBは金属すり板のすり板上面エッジ26である。
【0052】
エッジレベルを15から1ずつ減らし4になってもすり板17の上面のエッジAが検出できない時(ステップS502)は、図9のステップS510へ進み画像の右端から左に向かって画素の輝度値が暗い輝度から明るい輝度、または明るい輝度から暗い輝度となるエッジを検出する(ステップS510)。このような処理を舟体長手方向のすべての位置で行う。このような輝度の差を取るのは、すり板17の向こう側エッジと背景との差が無く、背景の方が明るく写り、すり板17の向こう側エッジの部分が画像の右から左を見て明るい部分から暗い部分に変化していることがあるためである。
【0053】
そして、その検出されたエッジが、すり板厚み方向に32画素の幅で、舟体長手方向に8分割した各ブロック内でエッジがN個以上存在し、しかも、そのブロックが6ブロック以上存在した時、そこに直線成分が存在するとする。そして、そのエッジをすり板17の上面エッジA’とする(ステップS511)。
【0054】
次に、前記方法で求めたすり板17の上面エッジA’から画像の左方向の舟体15の幅分85mm先の地点で画像の左方向に向かって画素の輝度値が明るい輝度から暗い輝度になるエッジを検出する(ステップS512)。これをすべての舟体長手方向の位置に対して行う。そして、その検出されたエッジが、すり板厚み方向に32画素の幅で、舟体長手方向に8分割した各ブロック内でエッジがN個以上存在し、しかも、そのブロックが6ブロック以上存在する時、そこに直線成分が存在するとする(ステップS513)。
【0055】
そして、その位置をすり板17の上面エッジA’であるとして、すり板17上面エッジA’が舟体15の向こう側エッジであるとし、その位置が撮影された画像上での舟***置であるとする(ステップS514)。すり板17の上面エッジB’は金属すり板のすり板上面エッジ26である。
【0056】
このようにすり板17の上面エッジを求め、そして、設計値である舟体15の幅を利用することで、パターンマッチングでは検知できなかった画像上の舟***置を検知することができる。
【0057】
また、すり板17の上面エッジA’も求まらなかった時、撮影画像には舟体15が写っていないものと判断する。
【0058】
実施の形態3.
図10はすり板上面エッジ26からすり板厚み算出、異常検出を行う検査範囲設定を示した図であり、SW1はすり板上面エッジ26から舟体下面方向で舟体下面18まで達しない幅、SW2はすり板上面エッジ26から舟体下面とは逆方向ですり板17の半分程度の幅、L3は舟体長手方向で舟体中央側から舟体端までの幅である。
【0059】
実施の形態2のようにして求めた画像上のすり板上面エッジ26から求めた直線成分から舟体下面18側に舟体下面18まで達しないが舟体上面とすり板下面の境界15Aを含む一定の幅SW1とその反対側にすり板17の幅の半分程度の幅SW2を取り、また、舟体15に装着しているすり板17の全ての厚みを測定するため舟体長方向に舟体中央側から舟体端までの一定の長さL3を取った範囲をすり板厚み算出、異常検出を行う検査範囲20とする。このように範囲を設定することで、すり板厚み算出及び異常検出を行う際に、検査範囲が特定できて求めやすくなる。
【0060】
実施の形態4.
図11はすり板上面エッジ26の向こう側エッジからすり板厚み算出、異常検出を行う検査範囲設定を示した図であり、29はすり板の向こう側エッジからの寸法、W3は舟体下面方向に舟体下面までは達しない一定の幅、L3は舟体長手方向の一定の幅である。
【0061】
実施の形態2のようにして求めた画像上の舟***置の向こう側にあるすり板17の上面のエッジAまたはA’で求めた直線成分から画像の左側にすり板上面エッジ26までは達しない寸法29だけ移動した位置から、舟体下面18の方向に舟体下面18までは達しないが舟体上面とすり板下面の境界15Aを含む幅W3を取る。
【0062】
また、舟体15に装着しているすり板17の全ての厚みを測定するため舟体長方向に舟体中央側から舟体端までの長さL3を取った範囲をすり板厚み算出、異常検出を行う検査範囲20とする。
このように範囲を設定することで、すり板厚み算出及び異常検出を行う際に、検査範囲が特定できて求めやすくなる。
【0063】
実施の形態5.
図12はカーボンすり板上面エッジから撮影画像上の舟体検知を示した図であり、図13、14はカーボンすり板上面エッジから撮影画像上の舟体検知を示したフローである。
【0064】
すり板17の上面よりすり板17の側面の方が暗く撮影されるため画像右から左方向に向かってすり板上面エッジ26は輝度の明るい部分から暗い部分に変化しているため、画像右端から検出を開始して、画像右から左方向で隣の画素の輝度値が明るい輝度から暗い輝度になる差が15以上となるエッジを画像左端まで求める。そして、この処理をすべての舟体長手方向の位置で検出する(ステップS517)。
【0065】
その検出されたエッジが、すり板厚み方向に32画素の幅で、舟体長手方向に8分割した各ブロック内でエッジがN個以上存在し、しかも、そのブロックが6ブロック以上存在する時、その位置をすり板上面のエッジEとする(ステップS518)。また、ブロックが6ブロック以上存在しない時は、エッジレベルを14から1ずつ4まで減らして、すり板17の上面エッジEを求める(ステップS521)。
【0066】
次にすり板17の向こう側エッジはすり板17の上面の方が背景より明るいため画像右から左に向かって暗い部分から明るい部分に変化しており、すり板17の上面エッジEから舟体15の幅分40mm後(画像右方向)の地点で画像の右方向に向かって画素の輝度値が暗い輝度から明るい輝度になるエッジを検出する。これをすべての舟体長手方向の位置に対して行う(ステップS519)。
【0067】
そして、その検出されたエッジが、すり板厚み方向に32画素の幅で、舟体長手方向に8分割した各ブロック内でエッジがN個以上存在し、しかも、そのブロックが6ブロック以上存在する時、そこに直線成分が存在するとする(ステップS520)。そして、その位置をすり板17の上面エッジDとする(ステップS521)。
【0068】
また、すり板17の上面エッジEがカーボンすり板のすり板上面エッジ26である。ブロックが6ブロック以上存在しない時は、エッジレベルを14から1ずつ4まで減らして(ステップS521、516)、すり板上面のエッジEから求め直す。カーボンすり板の場合は、舟体15の上にすり板17が1個搭載されており、舟体15の幅とすり板17の幅は同じ幅である。
【0069】
エッジレベルを15から1ずつ減らし4になってもすり板17の上面エッジEが検出できない時、画像の右端から左に向かって画素の輝度値が明るい輝度から暗い輝度となるエッジを検出する(ステップS523)。これをすべての舟体長手方向の位置に対して行う。そして、その検出されたエッジがすり板厚み方向に32画素の幅で、舟体長手方向に8分割した各ブロック内でエッジがN個以上存在し、しかも、そのブロックが6ブロック以上存在した時、そこに直線成分が存在するとする。そして、そのエッジをすり板17の上面エッジE’とする(ステップS524)。
【0070】
そして、すり板17の上面エッジE’から舟体15の幅分40mm後(画像右方向)の地点で画像の左方向に向かって画素の輝度値が明るい輝度から暗い輝度、または暗る輝度から明るい輝度になるエッジを検出する(ステップS525)。これをすべての舟体長手方向の位置に対して行う。このような輝度の差を取るのは、すり板17の向こう側エッジと背景との差が無く、背景の方が明るく写り、すり板17の向こう側エッジの部分が画像の右から左を見て明るい部分から暗い部分に変化していることがあるためである。
【0071】
このようにして検出されたエッジが、すり板厚み方向に32画素の幅で、舟体長手方向に8分割した各ブロック内でエッジがN個以上存在し、しかも、そのブロックが6ブロック以上存在する時、そこに直線成分が存在するとする(ステップS526)。そして、その位置をすり板上面のエッジD’であるとして(ステップS527)、すり板上面のエッジD’が舟体15の向こう側エッジであるとし、その位置が撮影された画像上での舟***置であるとする。すり板17の上面のエッジE’はカーボンすり板のすり板上面エッジ26である。
【0072】
このようにすり板17の上面エッジを求め、そして、設計値である舟体15の幅を利用することで、パターンマッチングでは検知できなかった画像上の舟***置を検知することができる。
また、すり板上面のエッジD’も求まらなかった時、撮影画像には舟体15が写っていないものと判断する。
【0073】
実施の形態6.
図15はカーボンすり板のすり板上面エッジからすり板厚み算出、異常検出を行う検査範囲設定を示した図であり、SW3はすり板上面エッジ26から舟体下面方向への一定の幅、SW4はすり板上面エッジ26から舟体下面とは逆方向への一定の幅、L4は舟体長手方向の一定の幅である。
【0074】
実施の形態5のようにして求めた画像上の舟***置の上面エッジ26から求めた直線成分から舟体下面18側に舟体下面18までは達しないが舟体上面とすり板下面の境界15Aを含む幅SW3とその反対側にすり板17の幅の半分程度の幅SW4を取る。
【0075】
また、舟体15に装着されているすり板17のすべての厚みを測定するため舟体長方向に舟体中央側から舟体端までの一定の長さL4を取った範囲をすり板厚み算出、異常検出を行う検査範囲20とする。このように範囲を設定することで、すり板厚み算出、異常検出を行う際に、範囲が特定できて求めやすくなる。
【0076】
実施の形態7.
図16はすり板上面エッジの向こう側エッジからすり板厚み算出、異常検出を行う検査範囲設定を示した図であり、33はすり板の向こう側エッジからの寸法、W4は舟体下面方向に舟体下面までは達しない一定の幅、L4は舟体長手方向の一定の幅である。
【0077】
実施の形態5のようにして求めた画像上の舟***置の向こう側にあるすり板上面のエッジDまたはD’で求めた直線成分から画像の左側にすり板下面上面エッジ26まで達しない寸法OK33だけ移動した位置から、舟体下面方向に舟体下面18までは達しないが舟体上面とすり板下面の境界15Aを含む幅W4を取り、また、舟体15に装着されているすり板17のすべての厚みを測定するため舟体長方向に舟体中央側から舟体端まで一定の長さL4を取った範囲をすり板厚み算出、異常検出を行う検査範囲20とする。このように範囲を設定することで、すり板厚み算出及び異常検出を行う際に、検査範囲が特定できて求めやすくなる。
【0078】
実施の形態8.
実施の形態1で示したパターンマッチングでマッチした舟***置から舟体下面18の検査範囲設定方法を示す。図6において、テンプレート画像19のマッチした1点25から舟体下面18の方向へ舟体下面18までは達しないが、舟体上面とすり板下面の境界15Aまでは達する寸法OU24の位置を基準として、その基準位置からすり板厚み方向に舟体下面18を含む一定の幅W2と舟体長手方向に舟体中央側から一定の長さL2とを取った検査範囲を設定し、その範囲を舟体下面18を検出するための舟体下面検査範囲21とする。
【0079】
このように範囲を設定することで、舟体下面18のエッジを検出し、その下面の直線成分、あるいは平均位置を求める際に検査範囲が特定できて求めやすくなる。舟体下面方向の寸法OU24は形状の違うテンプレート画像19毎に最適な値が設定されている。
【0080】
実施の形態9.
舟体下面18の検査範囲設定は、パターンマッチングで求まったテンプレート画像の一点25から設定するのでなく、実施の形態2または実施の形態5で示したすり板17の上面エッジを求めて、そのエッジから舟体下面検査範囲21を設定しても良い。
図17はすり板上面エッジ26から舟体下面検査範囲21を設定する場合の図であり、27はすり板上面エッジ26から舟体下面18の方向への寸法OFFSET_Oである。
【0081】
金属すり板の場合はすり板上面エッジ26(図9ではBまたはB’)、カーボンすり板の場合はすり板上面エッジ26(図13ではEまたはE’)から舟体下面18の方向に舟体下面18までは達しない寸法OO27の所を基準として、その基準から舟体下面18の方向に舟体下面18を含む一定の幅W2を取る。そして、舟体長手方向は検査範囲20の舟体中央側の端から舟体長手方向に舟体下面18までの長さL2を取る。
【0082】
この一定の幅W2と一定の長さL2の範囲を舟体下面検査範囲21とする。このように範囲を設定することで、舟体下面18のエッジを検出し、その下面の直線成分、あるいは平均位置を求める際に検査範囲が特定できて求めやすくなる。
【0083】
実施の形態10.
図18は舟体上面とすり板下面の境界を検出するための範囲設定を示した図である。パターンマッチングでマッチしたテンプレート画像19の1点25からすり板厚み方向へ舟体上面とすり板下面の境界15Aの手前までの寸法OX36とテンプレート画像の1点25から舟体長手方向へ舟体端までの寸法OY35の位置を基準として、舟体15に装着しているすり板17のすべての厚みを測定するため、その基準位置から舟体長手方向に舟体中央側までの長さL5と、すり板厚み方向に舟体上面とすり板下面の境界15Aを含んだ幅W5を取った範囲を設定し、その範囲を舟体上面とすり板下面の境界15Aを検出する範囲34とする。このように範囲を設定することで、舟体下面18のエッジを検出し、その下面の直線成分、あるいは平均位置を求める際に範囲が特定できて求めやすくなる。
【0084】
テンプレート画像の1点25からのすり板厚み方向へ舟体上面とすり板下面の境界15Aの手前までの寸法OFFSET_36と、テンプレート画像の1点25から舟体長手方向へ舟体端までの寸法OY35は形状の違うテンプレート画像19毎に最適な値が設定されている。
【0085】
実施の形態11.
図19は、すり板上面エッジ26を舟体長手方向に複数ブロックに分割してそのブロックの平均位置から舟体上面とすり板下面との境界15Aのための範囲設定を示した図であり、37はすり板上面エッジ26を複数(図19では3ブロック)に分割したブロックA、38はすり板上面エッジ26を複数(図19では3ブロック)に分割したブロックB、39はすり板上面エッジを複数(図19では3ブロック)に分割したブロックC、40はブロックAのすり板上面エッジ26の平均位置、41はブロックBのすり板上面エッジ26の平均位置、42はブロックCのすり板上面エッジ26の平均位置、43はブロックAのすり板上面エッジ26の平均位置から舟体上面とすり板下面の境界15Aまでの寸法、44はブロックBのすり板上面エッジ26の平均位置から舟体上面とすり板下面の境界15Aまでの寸法、45はブロックCのすり板上面エッジ26の平均位置から舟体上面とすり板下面の境界15Aまでの寸法、F1はブロックAの平均位置40から寸法43だけ移動した位置、F2はブロックBの平均位置41から寸法44だけ移動した位置、F3はブロックCの平均位置42から寸法45だけ移動した位置、SW5はF1、F2、F3の各位置を結んだ折れ線から舟体下面18の方向で舟体下面18までは達しない幅、SW6はF1、F2、F3の各位置を結んだ折れ線からすり板17の上面方向ですり板上面エッジ26までは達しない幅、L6は舟体長手方向で舟体中央側から舟体端までの長さである。
【0086】
すり板上面エッジ26を舟体長手方向に複数ブロックに分割してそのブロックの平均位置から舟体上面とすり板下面の境界15を求める範囲設定の方法で3ブロックに分割した場合について説明する。すり板上面エッジ26を舟体長手方向に複数に分割するのは、すり板17の摩耗量が舟体端と舟体中央側で違うため、舟体長手方向のそれぞれの位置で最適な舟体上面とすり板下面の境界15Aを設定するために分割する。
【0087】
すり板上面エッジ26を3ブロックに分割した1つのブロックA内で、すり板厚み方向(×方向)ですべての舟体長手方向(y方向)の各位置毎で隣の画素との輝度差が最も大きなエッジを求め、その求められた舟体長手方向の各位置毎のエッジのX画素、Y画素から平均位置40を求める。
【0088】
これと同様にブロックB内ですり板上面エッジ26の平均位置41、ブロックC内ですり板上面エッジ26の平均位置42を求める。そして、ブロックAの平均位置40から舟体上面とすり板下面の境界15Aまでの寸法43だけ移動した位置F1と、ブロックBの平均位置41から舟体上面とすり板下面の境界15Aまでの寸法44だけ移動した位置F2と、ブロックCの平均位置42から舟体上面とすり板下面の境界15Aまでの寸法45だけ移動した位置F3の3点を結んだ折れ線を中心しとして、舟体下面18の方向で舟体下面18までは達しない幅SW5と、すり板17の上面方向ですり板上面エッジ26までは達しない幅SW6と、そして、舟体15に装着しているすり板17のすべての厚みを測定するため、舟体長手方向で舟体中央側から舟体端までの長さL6を取った範囲を舟体上面とすり板下面の境界15Aを検出する範囲34とする。
【0089】
また、各点F1、F2、F3を結んだ折れ線からの舟体下面18側とすり板上面側への寸法の幅は、各点毎に幅を変えて舟体上面とすり板下面との境界15Aを検出するための範囲としても良い。
【0090】
このようにして境界15Aを含む範囲34を予め設定することで、境界15Aを検出する処理においてそれ以外のエッジ成分を誤検出しなくなる。
【0091】
また、上記例ではすり板上面エッジ26を3ブロックに分割したが、3ブロック以外のブロック数に分割して処理しても良い。
平均位置の求め方は以下のようにして求める。
【0092】
平均位置のX位置=(×1+×2+×3+・・・・+×n)/n
平均位置のY位置=(Y1+Y2+Y3+・・・・+Yn)/n
【0093】
×1、×2、・・・・、×nは舟体長手方向の各位置毎に求めたエッジの×画素である。
Y1、Y2、・・・・、Ynは舟体長手方向の各位置毎に求めたエッジのY画素である。
nは平均位置を求めるブロック内で求めた舟体長手方向の各位置の数である。
【0094】
実施の形態12.
図20は、すり板上面エッジ26と舟体下面18から舟体上面とすり板下面との境界15Aを検出するための範囲設定を示した図であり、46は舟体端側のすり板上面エッジ26の平均位置を求める範囲、47は舟体端と舟体中央との間のすり板上面エッジ26の平均位置を求める範囲、48は舟体端側のすり板上面エッジ26の平均位置、49は舟体端と舟体中央との間のすり板上面エッジ26の平均位置、50は舟体下面18のエッジの平均位置、51は舟体端側のすり板上面エッジ26の平均位置から舟体上面とすり板下面の境界15Aまでの寸法、52は舟体端と舟体中央との間のすり板上面エッジ26の平均位置から舟体上面とすり板下面の境界15Aまでの寸法、53は舟体下面18のエッジの平均位置から舟体上面とすり板下面の境界15Aまでの寸法、F4は舟体端側の範囲46のエッジの平均位置48から寸法51だけ移動した位置、F5は舟体端と舟体中央との間の範囲47のエッジの平均位置49から寸法52だけ移動した位置、F6は舟体下面18のエッジの平均位置50から寸法53だけ移動した位置、SW7はF4、F5、F6の各位置を結んだ折れ線から舟体下面18の方向で舟体下面18までは達しない幅、SW8はF1、F2、F3の各位置を結んだ折れ線からすり板17の上面方向ですり板上面エッジ26までは達しない幅、L7は舟体長手方向で舟体中央側から舟体端までの長さである。
【0095】
舟体端のすり板上面側で一定の長さ(すり板上面エッジを3分割にしたブロックの半分程度)を取った舟体端側のすり板上面エッジ26の平均位置を求める範囲46内ですり板厚み方向(X方向)で舟体長手方向の各位置毎に隣の画素との輝度差が最も大きなエッジを求め、その求められた舟体長手方向の各位置毎のエッジの×画素、Y画素から平均位置48を求める。
【0096】
これと同様に一定の長さ(すり板上面エッジを3分割にしたブロックの半分程度)を取った舟体端と舟体中央との間のすり板上面エッジ26の平均位置を求める範囲47内から舟体端側のすり板上面エッジ26の平均位置49を求める。
【0097】
また、舟体下面検査範囲21内で舟体下面18のエッジの平均位置50を求める。そして、舟体端側のすり板上面エッジ26の平均位置48から舟体上面とすり板下面の境界15Aまで寸法51移動した位置F4と、舟体端と舟体中央との間のすり板上面エッジ26の平均位置49から舟体上面とすり板下面の境界15Aまで寸法52移動した位置F5と、舟体下面18の平均位置50から舟体上面とすり板下面の境界15Aまで寸法53移動した位置F6の3点を結んだ折れ線を中心として、舟体下面18側に寸法の幅SW7とすり板上面側に寸法の幅SW8を取った範囲と、そして、舟体15に装着しているすり板17のすべての厚みを測定するため、舟体長手方向に対して舟体中央側から舟体端まで一定の長さL7を取った範囲を舟体上面とすり板下面の境界15Aを検出するための範囲34とする。
【0098】
また、各点F4、F5、F6を結んだ折れ線からの舟体下面18側とすり板上面側への寸法の幅は、各点毎に幅を変えて舟体上面とすり板下面の境界15Aを検出するための範囲としても良い。
【0099】
すり板上面の平均値を求める範囲48、49をすり板上面エッジを3分割にしたブロックの半分程度としているが、これは、ブロックを3分割した各ブロック全体で平均値を求めても良い。さらに、舟体下面18の平均値を利用しているのは、舟体中央側のすり板17の摩耗量により、すり板上面エッジ26の平均値を境界15Aまで移動する寸法に開きがあり、境界15Aの範囲34が広がってしまうのを防ぐことができる。舟体下面18と境界15A間の寸法は一定である。
【0100】
このようにして範囲34を予め設定することで、境界15Aを検出する処理においてそれ以外のエッジ成分を誤検出しなくなる。
【0101】
実施の形態13.
図21は、すり板上面エッジ26の直線成分から舟体上面とすり板下面の境界15Aを検出するための範囲設定を示した図であり、54はすり板上面エッジ26の直線成分、55Aはすり板上面エッジ26から舟体上面とすり板下面の境界15Aまでの寸法、54Aはすり板上面エッジ26から舟体上面とすり板下面の境界15Aまでの寸法55Aだけ平行移動した直線、SW10はすり板上面エッジ26の直線54Aから舟体下面方向で舟体下面まで達しない幅、SW11はすり板上面エッジ26の直線54Aからすり板上面方向ですり板上面エッジ26まで達しないまでの幅、L8は舟体長手方向で舟体中央側から舟体端までの一定の長さである。
【0102】
検査範囲20内で舟体長手方向の各位置毎にすり板上面エッジ26を求め、そのエッジからすり板上面エッジ26の直線成分54を求める。この直線成分は、点の並びからHOUGH変換により直線成分上のエッジ数が最も多い直線を選択する。
【0103】
あるいは、すり板上面エッジ26から最小2乗法により直線成分54を求めて良い。そして、求めたすり板上面エッジ26の直線54をすり板上面エッジ26から舟体上面とすり板下面の境界15Aまでの寸法55Aだけ平行移動する。
【0104】
そして、すり板厚み方向で舟体上面とすり板下面の境界15Aを含む範囲を取るため、舟体上面とすり板下面の境界15A上に平行移動した直線54Aを中心としてすり板上面エッジ26の直線54Aから舟体下面方向で舟体下面まで達しない幅SW10と、すり板上面エッジ26の直線54Aからすり板上面方向ですり板上面エッジ26まで達しないまでの幅SW11を取り、また、舟体15に装着されているすり板17の厚みを全て測定するため舟体長手方向(Y方向)に舟体中央側から舟体端までの一定の長さL8を取った範囲を舟体上面とすり板下面の境界15Aを検出するための範囲34とする。
【0105】
このような範囲を設定することで、舟体上面とすり板下面の境界15を特定し、境界15以外のエッジを検出しないですむ。
【0106】
実施の形態14.
図22は、実施の形態13以外の方法としてすり板上面エッジ26の直線成分54から舟体上面とすり板下面の境界を検出するための範囲設定を示した図であり、55(n)は舟体長手方向の各位置毎のすり板上面エッジ26から舟体上面とすり板下面の境界15Aまでは達しない寸法、56(n)は舟体長手方向の各位置毎のすり板上面エッジ26からの寸法55(n)から舟体下面の方向で舟体上面とすり板下面の境界15Aを含んだ所までの幅、L9は舟体長手方向で舟体中央側から舟体端までの一定の長さである。
【0107】
検査範囲20内で舟体長手方向の各位置毎にすり板上面エッジ26を求め、そのエッジからすり板上面エッジ26の直線成分54を求める。この直線成分54は、点の並びからHOUGH変換により直線成分上のエッジ数が最も多い直線を選択する。
【0108】
あるいは、すり板上面エッジ26から最小2乗法により直線成分54を求めて良い。そして、求めた直線成分54上の舟体長手方向の1個目の位置からすり板上面エッジ26から舟体上面とすり板下面の境界15Aまでは達しない寸法55(1)だけ移動した所の点を基準として,その点から舟体下面18の方向で舟体上面とすり板下面の境界15Aを含んだ所までの幅56(1)を取る。これは、舟体長手方向の各位置毎にすり板厚み方向で舟体上面とすり板下面の境界15Aを含む範囲を取るために行う。
【0109】
これを2個目の位置以降、n個目まで舟体長手方向の各位置毎に直線成分54上の点から各寸法55(n)だけ移動した点を基準として、その点から一定の寸法56(n)の幅を取る。そして、これら求めた舟体長手方向の各位置毎に幅55(n)と、そして舟体15に装着されているすり板17の厚みを全て測定するため、舟体長手方向で舟体中央側から舟体端までの一定の長さL9を取った範囲を舟体上面とすり板下面との境界15を検出するための範囲34とする。
【0110】
実施の形態15.
図23は、すり板上面エッジ26の各エッジから舟体上面とすり板下面の境界15Aを検出するにあたり範囲を限定して境界15A以外を検出しないため検査範囲を設定することを示した図である。
【0111】
検査範囲20内で舟体長手方向の各位置毎にすり板上面エッジ26を求め、まず、舟体長手方向の1個目のエッジをすり板上面エッジ26から舟体上面とすり板下面の境界15A方向でその境界15Aまでは達しない寸法H(1)だけ移動した所の点を基準として、その点から舟体下面方向で舟体上面とすり板下面の境界15Aを含む幅H’(1)を取る。
【0112】
これは舟体上面とすり板下面の境界15Aを検出するためこの幅H’(1)を取る。同様に2個目以降、n個目まで舟体長手方向の各位置毎にすり板上面エッジ26から各寸法H(n)だけ移動した点を基準として、その点から各幅H(n)を取る。そして、これら求めた舟体長手方向の各位置毎の幅H’(n)と、舟体15に装着したすり板17の全ての厚みを測定するため舟体長手方向で舟体中央側から舟体端まで一定の長さL10を取った範囲を舟体上面とすり板下面との境界を検出するための範囲とする。
【0113】
実施の形態16.
図24は舟体上面とすり板下面の境界15Aの候補点を検出するための求めかたを示した図であり、図24は舟体上面とすり板下面の境界15Aのエッジ検出フローである。
【0114】
金属すり板の場合、舟体上面とすり板下面との境界15Aを検出するため範囲34内で、すべての舟体長手方向の各位置に対してすり板17の上面側の画素から隣の舟体下面18側の画素の差分を取った時、差分値が正となりその差分値が一番大きなエッジA(n)を検出する(ステップS600)。このエッジA(n)は、すり板17の上面側から舟体下面18に向かって小さな輝度(暗い)から大きな輝度(明るい)への変化が一番大きなエッジとなる(ステップS601)。
【0115】
次にそのエッジA(n)の位置からすり板17の上面に向かって一定の戻り量60(例えば2mm)の範囲ですり板17の上面側の画素から隣の舟体下面18側の画素の差分を取った時、差分値が正となりその差分値が一番大きなエッジB(n)を検出する(ステップS602)。このエッジB(n)は、舟体下面18からすり板17の上面側に向かって小さな輝度(暗い)から大きな輝度(明るい)への変化が一番大きなエッジとなる(ステップS603、604)。
【0116】
ただし、この各位置で求めたエッジB(n)が所定値以上の差分値である時、その位置ではエッジB(n)を採用し(ステップS605、607)、所定値以下である時、その位置においては最初に求めたエッジA(n)を採用する(ステップS605、606)。
【0117】
このようにして舟体長手方向の各位置毎に求めたエッジA(n)またはB(n)を舟体上面とすり板下面の境界15Aの候補とする。これは、差分値が所定値以下であるときそのエッジは境界15Aが分かりにくく、本来の境界15A以外のエッジを抽出している可能性があるために、所定値以下のエッジは除外している。
【0118】
また、カーボンすり板の場合は、舟体上面とすり板下面の境界15Aの幅が金属すり板より広く、すり板上面から舟体下面18に向かって求めたエッジA(n)が舟体上面とすり板下面の境界15Aの候補となるため、カーボンすり板はこのエッジA(n)を舟体上面とすり板下面の境界15Aの候補とする。
エッジA(n)とエッジB(n)を求める差分は、数画素離れた画素間である方が良い場合もある。
なお、nは舟体長手方向の各位置の総数である。
【0119】
実施の形態17.
図26は舟体上面とすり板下面の境界15Aの候補点の評価を示した図である。請求項で求めた舟体上面とすり板下面の境界15Aの候補点から舟体下面18の方向にある画素の輝度値が、所定の値より大きい画素が所定の幅HO以上存在し、しかもその幅HOが舟体長手方向に連続して所定の長さLOより長く存在する時、その部分62は光りが反射している部分として舟体上面とすり板下面の境界15Aの候補から除外する。
【0120】
舟体上面とすり板下面の境界15Aが太陽光等により反射し、その部分で求めた境界線は本来の境界に沿って引いていないためずれた境界線となるが、その境界を除外することでその影響を防ぐことができる。
【0121】
実施の形態18.
図27は、舟体上面とすり板下面の境界15Aを検出を示した図である。
舟体長手方向に舟体上面とすり板下面の境界15Aを複数ブロックに分割して各ブロック毎に境界線を検出し、その境界線の交点を結んだ折れ線を境界とする方法について説明する。複数ブロックに分割するのは、舟体15が湾曲し1本の直線で舟体上面とすり板下面の境界15Aを求めることができないため、湾曲した舟体15の境界を直線近似して求めるためである。
【0122】
舟体上面とすり板下面との境界15Aを舟体長手方向にn分割に分割したブロック(1)63、ブロック(2)64、ブロック(3)65、・・・・ブロック(n−1)66、ブロック(n)67はそれぞれオーバラップさせて分割する。
【0123】
そして、各ブロック毎に舟体上面とすり板下面との境界15Aの候補となるエッジの点群から、HOUGH変換により直線を求める。その求めた直線が、各ブロックの境界線(1)68、境界線(2)69、境界線(3)70、・・・・、境界線(n−1)71、境界線(n)72である。
【0124】
次に、境界線(1)68と境界線(2)69の交点(1)73、境界線(2)69と境界線(3)70の交点(2)74、・・・・、境界線(n−1)71と境界線(n)72の交点(n−1)75をつなぎ目として複数本の境界線を結ぶ。このようにして求めた境界線を舟体上面とすり板下面との境界15Aとする。
【0125】
また、各ブロックで求めた境界線の交点が無いときは以下のように接続する。舟体端または舟体中央のブロックで境界線の交点が無いとき、例えば,ブロック(1)63の境界線(1)68とブロック(2)64の境界線(2)69の交点が無いとき、境界線(1)68がブロック(1)63の端側(ブロック(2)64と反対側)に位置する点と、ブロック(2)64の境界線(2)69とブロック(3)65の境界線(3)70の交点74をつないでブロック(1)63とブロック(2)64の境界線とする。
【0126】
また、隣接する2つのブロックの境界線の交点が2つの当該ブロック内に無いときは、その両側の隣接するブロック以遠に存在する境界線の交点同士をつないで、交点のない2つのブロックの境界線とする。
【0127】
さらに、すべてのブロックの境界線で交点が無い時は、境界線(1)68がブロック(1)63の端側(ブロック(2)64と反対側)に位置する点と、境界線(n)72がブロック(n)67の端側(ブロック(n−1)66と反対側)に位置する点とをつないだ線を境界線とする。
なお、以上におけるnはブロックの分割数である。
【0128】
実施の形態19.
図28は、舟体上面とすり板下面の境界15Aの検出で、あるブロックの直線を利用しない場合を示した図である。
実施の形態18では、各ブロックの境界線を境界候補となるエッジの点群からHOUGH変換により求めたが、この境界線上のエッジの点数が所定数より少ない場合、そのブロックで求めた境界線を利用しないようにする。エッジの点数が所定数より少ないことは、舟体上面とすり板下面の境界15Aが分かりにくく、エッジがばらついたためであり、それで求めた境界線は信頼できないためである。
【0129】
エッジの点数が所定数より少ない境界線が所定数以上の境界線に挟まれている場合、例えばブロック(2)64の境界線(2)76が所定数より少ない場合、このブロック(2)64の境界線(2)76を利用しないで、その隣接するブロック(1)63とブロック(3)65で求めた境界線(1)68と境界線(3)70を延長し、その交点で結んだ直線78をその部分の境界線とする。
【0130】
また、エッジの点数が所定数より少ない境界線が端側である場合、例えばブロック(n)67の境界線(n)77が所定数より少ない場合、そのブロック(n)67の隣接するブロックはブロック(n−1)66のみであるため、このブロック(n−1)66の境界線(n−1)71をブロック(n)67まで延長し、その直線79をその部分の境界線とする。
【0131】
このようにして、エッジの点数が所定数より少ない場合は、上記のように境界線をつなぎ、舟体上面とすり板下面の境界15Aとする。
【0132】
実施の形態20.
図29は、舟体上面とすり板下面の境界15Aの検出で舟体中央側のブロックの直線が利用できない時に舟体下面18の直線を利用する場合を示した図である。
【0133】
実施の形態18では舟体上面とすり板下面の境界15Aを複数のブロックに分割してそれぞれのブロックで境界線を求め、その交点をつないで舟体上面とすり板下面の境界15Aとしたが、舟体中央部のブロック(n)67で求めた境界線(n)77のエッジの点数が所定数より少ない時、この境界線(n)77を利用しないで舟体下面範囲21内で舟体下面18のエッジを求め、そのエッジの点群をHOUGH変換により求めた直線を舟体下面18の直線80とし、この直線80を舟体上面とすり板下面の境界15Aの方向に舟体下面18から舟体上面とすり板下面の境界15Aまでの寸法だけ平行移動する。この平行移動した直線81をブロック(n)の舟体上面とすり板下面の境界15Aの境界線とし、そして、この直線81とブロック(n−1)66の境界線(n−1)71の交点を結び、舟体上面とすり板下面の境界15Aとする。
【0134】
実施の形態21.
図30は舟体上面とすり板下面の境界15Aの検出で舟体下面18の平均位置を利用した場合を示した図である。
実施の形態18では舟体上面とすり板下面の境界15Aを複数のブロックに分割してそれぞれのブロックで境界線を求め、その交点をつないで舟体上面とすり板下面の境界15Aとしたが、舟体中央部のブロック(n)67で求めた境界線(n)77のエッジの点数が所定数より少ない時、この境界線を利用しないで舟体下面18の範囲21内で舟体下面18のエッジを求め、そのエッジの平均位置82を求める。
【0135】
そして、この平均位置82を舟体上面とすり板下面の境界15Aの方向に舟体下面18と舟体上面とすり板下面の境界15Aまでの寸法だけ移動する。この移動した点83とブロック(n−1)66の境界線(n−1)71の端を結び、その結んだ線をブロック(n)67の端まで延長する。この点83と境界線(n−1)71を結んだ直線84をブロック(n)67の境界とする。
【0136】
実施の形態22.
図31は、遮蔽板により舟体上面とすり板下面の境界15Aが見えない舟体の境界の検出を示した図である。
舟体下面18の範囲21内で舟体下面18側の遮蔽板85の1辺のエッジの点群からHOUGH変換により遮蔽板85の一辺の直線80Aを求める。そして、この直線80Aを舟体上面とすり板下面の境界15A方向に舟体下面18と舟体上面とすり板下面の境界15Aまでの寸法だけ平行移動した直線86を舟体端側に延長し、直線86と延長した部分87を舟体上面とすり板下面の境界15Aとする。
【0137】
実施の形態23.
図32は、遮蔽板85により舟体上面とすり板下面の境界15Aが見えない舟体15の境界15Aを遮蔽板85の一辺と境界15Aの見える部分の平均位置を利用して求める場合を示した図である。
【0138】
舟体上面とすり板下面の境界15Aを含んで舟体下面18まで遮蔽する長方形の遮蔽板85が舟体15に装着されているとき、遮蔽板の1辺は境界15Aと平行であることから、この1辺の傾きで境界15Aの見えている部分を通る直線を引けば、見えない部分の境界15Aを決定することができる。
【0139】
まず舟体下面18の範囲21内で舟体下面18側の遮蔽板85の1辺のエッジの点群からHOUGH変換により遮蔽板85の1辺の直線80Aを求める。次に遮蔽板85で覆われていない舟体上面とすり板下面の境界15A部分のエッジを抽出し、その平均位置88を求める。
【0140】
この平均位置88を通るように舟体下面18側の遮蔽板85の1辺の直線80Aと平行な直線を引き、この直線を舟体端側と舟体中央側まで延長した直線89を舟体上面とすり板下面の境界15Aとする。
【0141】
実施の形態24.
図33は、長方形でない遮蔽板により舟体上面とすり板下面の境界15Aが見えない舟体15の境界を遮蔽板の1辺の平均位置と境界の見える部分の平均位置を利用して求める方法を示した図である。
【0142】
舟体上面とすり板下面の境界15A部分から舟体下面18まで遮蔽する遮蔽板85が舟体15に装着されているとき、舟体下面18の範囲21内で舟体下面18側の遮蔽板85の1辺のエッジの平均位置82Aを求め、この平均位置82Aを舟体上面とすり板下面の境界15A方向に遮蔽板85の1辺から舟体上面とすり板下面の境界15Aまでの寸法だけ移動し、この点を83Aとする。
【0143】
次に遮蔽板85で覆われていない舟体上面とすり板下面の境界15A部分のエッジを抽出し、その平均位置88を求める。
そして、遮蔽板の平均位置82Aを移動した点83と、遮蔽板85で覆われていない舟体上面とすり板下面の境界15A部分の平均位置88を通る直線90を引き、その直線90を舟体端側と舟体中央側まで延長した線を舟体上面とすり板下面の境界15Aとする。
【0144】
実施の形態25.
図34は、すり板上面エッジ26の求め方を示した図であり、A(1)は舟体長手方向の1個目の位置のすり板上面エッジ、A(2)は舟体長手方向の2個目の位置のすり板上面エッジ、A(n)は舟体長手方向のn個目の位置のすすり板上面エッジ、B(1)は舟体長手方向の1個目の位置のすり板の向こう側エッジ、B(2)は舟体長手方向の2個目の位置のすり板の向こう側エッジ、B(n)は舟体長手方向のn個目の位置のすり板の向こう側エッジ、97は舟体15の幅分相当の画素、98は舟体上面とすり板下面の境界線、A’(n、1)はn個目の位置ですり板上面エッジ26の1個目の候補、A’(n、2)はn個目の位置ですり板上面エッジ26の2個目の候補、A’(n、m)はn個目の位置ですり板上面エッジ26のm個目の候補、B’(n、1)はn個目の位置ですり板上面エッジ26の1個目の候補A’(n、1)から舟体幅分先の向こう側のエッジ、B’(n、2)はn個目の位置ですり板上面エッジの2個目の候補A’(n、2)から舟体幅分先の向こう側エッジ、B’(n、m)はn個目の位置ですり板上面エッジ26のm個目の候補A’(n、m)から舟体幅分先び向こう側エッジ、nは舟体長手方向の位置の総数、mは舟体長手方向ある位置ですり板上面エッジの候補数である。
【0145】
検査範囲20内で、舟体上面とすり板下面の境界15Aからすり板17の上面方向に舟体長手方向の各位置毎にすり板上面のエッジを求める。このエッジの求め方は、まず1個目の位置で舟体上面とすり板下面の境界15Aからすり板17の上面方向(*方向)にすり板17の上面側の画素の輝度値からその画素の隣で舟体上面とすり板下面の境界15A側の画素の輝度値の差分値を求める。
【0146】
そして、複数のエッジA’(1、1)、A’(1、2)、・・・・、A’(1、m)を求める。そのエッジA’(1、1)、A(1、2)、・・・・、A’(1、m)からすり板17の向こう側エッジの方向で舟体15の幅分相当の画素97先で、それぞれ、すり板上面エッジ26側の画素の輝度値からその画素の隣ですり板17の向こう側エッジ側の画素の輝度値の差分値を取りエッジB’(1、1)、B’(1、2)、・・・・、B’(1、m)を求める。次に、エッジA’(1、1)、A’(1、2)、・・・・、A’(1、m)の差分値の絶対値と、エッジB’(1、1)、B’(1、2)、・・・・、B’(1、m)の差分値の絶対値の積を求める。
【0147】

Figure 0004209525
*ここで、|A’(1、1)|、|A’(1、2)|、・・・・、|A’(1、m)|、|B’(1、1)|、|B’(1、2)|、・・・・、|B’(1、m)|は、各エッジA’(1、1)、A’(1、2)、・・・・、A’(1、m)、 B’(1、1)、B’(1、2)、・・・・、B’(1、m)の差分値の絶対値とする。
*上記式は舟体長手方向の1個目の位置での場合である。
【0148】
そして、積C(1、1)、C(1、2)、・・・・、C(1、m)の中で一番大きな値を示したものがC(1、2)であれば、エッジA’(1、2)が1個目の位置のすり板上面エッジA(1)であるとする。
【0149】
同様に2個目以降、n個目まで積が一番大きなエッジA’(i、j)(i=1〜n、j=1〜m)を求める。
【0150】
このようにして求められたエッジA’(i、j)(i=1〜n、j=1〜m)をすり板上面エッジA(1)、A(2)、・・・・、A(n)とする。この時の舟体15の幅相当画素数97先のB’(i、j)(i=1〜n、j=1〜m)が向こう側エッジB(1)、B(2)、・・・・、B(n)となる。
【0151】
また、画素の差分値を隣の画素間で求めたが、これを数画素以上離れた画素間で求めても良い。
【0152】
実施の形態26.
実施の形態25で示したすり板上面エッジ26の求め方で、舟体長手方向の各位置毎の複数のエッジA’(i、j)(i=1〜n、j=1〜m)を隣の画素の輝度の差分値で求めたが、撮影画像の特徴から舟体15の上面からすり板17の上面方向に見て、すり板上面エッジ26は暗い輝度から明るい輝度へと変化しているため、舟体長手方向の各位置毎の複数のエッジA’(i、j)(i=1〜n、j=1〜m)を舟体15の上面側の画素からすり板上面側の画素を引いた差が負の値を示したエッジのみを抽出しても良い。
【0153】
また、舟体長手方向の各位置毎の複数のエッジA’(i、j)(i=1〜n、j=1〜m)と向こう側エッジB’(i、j)(i=1〜n、j=1〜m)に、すり板上面側ほど大きな係数を賭けて積を求めても良い。これは、舟体上面とすり板下面の境界15Aからすり板上面エッジ26までの間にノイズが存在し、それを検出してしまう可能性があり、しかも、すり板17の上面側ほど本来のすり板上面エッジ26である可能性が高いため、すり板17の上面側ほど重みを重く付けている。
【0154】
Figure 0004209525
*ここで、|A’(1、1)|、|A’(1、2)|、・・・・、|A’(1、m)|、|B’(1、1)|、|B’(1、2)|、・・・・、|B’(1、m)|は、各エッジA’(1、1)、A’(1、2)、・・・・、A’(1、m)、B’(1、1)、B’(1、2)、・・・・、B’(1、m)の輝度差の絶対値とする。
kA1、kA2、kAM、kB1、kB2、kBmは係数である。
*上記式は舟体長手方向の1個目の位置での場合である。
【0155】
実施の形態27.
図35は、すり板上面エッジの検出開始範囲を示した図であり、3分割した場合の例を示している。L(1)は舟体端から1/6の部分でのすり板上面検査開始位置、L(2)はさらに1/6の部分でのすり板上面検査開始位置、L(3)は舟体15の残り4/6の部分でのすり板上面検査開始位置L(3)である。
【0156】
実施の形態25では、すり板上面エッジ26の検査開始を舟体上面とすり板下面の境界線98から開始したが、これを舟体長手方向(Y方向)の舟体長手方向の各位置毎に検査開始位置を変えて行う。
【0157】
それは、湾曲した舟体15に対して求めた舟体上面とすり板下面の境界線98が、実際の境界より舟体下面18側にずれ、そのずれた境界からすり板上面エッジ26を検査した時、本来の舟体上面とすり板下面の境界15Aをすり板上面エッジ26と誤検出する可能性があるためである。
【0158】
また、舟体上面とすり板下面の境界15Aよりすり板17の上面側からすり板上面エッジ26を求めても、すり板上面エッジ26を越えてしまわない開始位置を設定する。
【0159】
そして、舟体上面とすり板下面の境界線98からL(1)、L(2)、L(3)の位置からすり板上面エッジ26の検査を開始する。
検査開始位置を舟体長手方向に3分割して設定したが、これを複数に分割して設定しても良い。
【0160】
実施の形態28.
図36は、段摩耗の判定を示した図であり、aは段摩耗の深さ、bは段摩耗の長さ、cは段摩耗の長さ、dは段摩耗の幅、108はすり板厚みである。
図36はこれまでの図と向きを90度変えて表している。
【0161】
舟体上面とすり板下面の境界線98からすり板上面エッジ26までのすり板厚み108を舟体長手方向の各位置毎に求める。
そして、幅dの間隔で厚み値108の差を求める。その求めた厚み値の差が深さa以上である時、その摩耗した部分の左右(舟体長手方向の左右)の幅bと幅cの間で厚み値108の平均値を求める。そして、それぞれの平均値の差が深さa以上である時、その部分は段摩耗であるとする。
【0162】
実施の形態29.
図37は、舟体両側側面から段摩耗を判定する場合を示した図である。
実施の形態28で示した段摩耗の判定で、舟体15の片側側面で段摩耗と判定したD1とD2と、その反対側側面を撮影した画像で段摩耗と判定したD3とで、舟体長手方向の位置を比較する。D1と舟体長手方向でほぼ同じ位置にある反対側側面では段摩耗と判定していないため、D1は段摩耗で無いとし、次にD2と舟体長手方向でほぼ同じ位置にある反対側側面で段摩耗と判定したD3があるため、D2は段摩耗であるとする。段摩耗は、舟体25の片側側面から反対側側面まで上面に渡って発生するため段摩耗の判断を両側側面で行う。
【0163】
実施の形態30.
図38は、欠けの判定を示した図であり、kaは欠けの幅、kbは欠けの候補と判定するための欠けの深さ、kcは欠けの候補となる部分でkbから最も深い位置までの深さ、112は欠けの候補となった部分の最大厚み値、113は欠けの候補となった部分の最小厚み値、114は舟体25の向こう側エッジを手前に舟体15の幅分平行移動したエッジである。
【0164】
舟体上面とすり板下面との境界線98からすり板上面エッジ26までの厚み値108を求める。次に舟体15の向こう側エッジを求め、それを舟体15の幅分手前に平行移動したエッジ114とすり板上面エッジ26との差を求める。その差がkb以上ある部分で、しかも幅がka以上連続している時、その部分を欠けの候補とする。
【0165】
そして、欠けの候補となった部分(幅ka以上と差kb以上ある部分)で、厚みの最大値112と厚みの最小値113との差がkc以上ある時、その部分を欠けとする。欠けは舟体15の片側側面にしか発生しないため、前側画像と後側画像ごとにそれぞれ判定を行う。
【0166】
実施の形態31.
実施の形態18では舟体上面とすり板下面との境界15Aを複数のブロックに分割し各ブロック毎に境界線を求め、その境界線を舟体上面とすり板下面の境界15Aとしたが、その境界線で舟体中央側のブロック(n)67で求めた境界線(n)72の傾きを舟体15の傾きとする。
【0167】
実施の形態19では、舟体上面とすり板下面との境界15Aを複数のブロックに分割し各ブロック毎に境界線を求め、その境界線のエッジ数が所定数より少ないブロックの境界線は利用しないようにしたが、このとき、舟体中央側のブロック(n)67の境界線(n)72が所定数より少ないため、その隣接するブロック(n−1)66の境界線(n−1)71を延長しそれを境界線としたとき、その境界線の傾きを舟体15の傾きとする。
【0168】
実施の形態20では舟体上面とすり板下面の境界15Aを複数のブロックに分割してそれぞれのブロックで境界線を求め、そして舟体中央部のブロック(n)67で求めた境界線(n)77が所定のエッジ数より少ない時、舟体下面18の直線80を舟体上面とすり板下面の境界15A方向に舟体下面18と境界15Aまでの寸法だけ平行移動し、この平行移動した直線81をブロック(n)67の舟体上面とすり板下面の境界15Aの境界線とし、そして、この直線81とブロック(n−1)68の境界線(n−1)71の交点を結び、舟体上面とすり板下面の境界15Aとしたが、この場合は、この舟体下面18の直線の傾きを舟体15の傾きとする。
【0169】
実施の形態21では、舟体上面とすり板下面の境界15Aを複数のブロックに分割してそれぞれのブロックで境界線を求め、舟体中央部のブロック(n)67で求めた境界線(n)77が所定のエッジ数より少ない時、この境界線を利用しないで舟体下面18のエッジの平均位置82を求め、そして、この平均位置82を舟体上面とすり板下面の境界15A方向に舟体下面18と境界15Aまでの寸法だけ移動し、この移動した点83とブロック(n−1)66の境界線(n−1)71の端を結び、そして、その結んだ線をブロック(n)67の端まで延長した直線84をブロック(n)67の境界としたが、この場合はこの延長した直線84の傾きを舟体15の傾きとする。
【0170】
実施の形態22では、舟体上面とすり板下面の境界15A部分から舟体下面18まで遮蔽する遮蔽板85が舟体15に装着されているため、舟体上面とすり板下面の境界15Aが見えないことから、舟体下面18側の遮蔽板85の1辺の一辺の直線80Aを舟体上面とすり板下面の境界15A方向に舟体下面18と境界15Aまでの寸法だけ平行移動した直線86を舟体端側に延長し、この直線86と延長した部分87を舟体上面とすり板下面の境界15Aとしたが、この場合は、この直線86の傾きを舟体15の傾きとする。
【0171】
実施の形態23では、舟体上面とすり板下面の境界15A部分から舟体下面18まで遮蔽する遮蔽板85が舟体15に装着されているため、舟体下面18側の遮蔽板85の1辺の直線80Aの傾きで、遮蔽板85で覆われていない舟体上面とすり板下面の境界15A部分のエッジの平均位置88を通るように引いた直線を、舟体上面とすり板下面の境界15Aとしたが、この直線の傾きを舟体15の傾きとする。
【0172】
実施の形態24では、舟体上面とすり板下面の境界15A部分から舟体下面18まで遮蔽する遮蔽板85が舟体に装着されているため、舟体下面18側の遮蔽板85の1辺の平均位置82Aを求め、この平均位置82Aを舟体上面とすり板下面の境界15A方向に遮蔽板85の舟体下面18側の一辺と境界15Aまでの寸法だけ移動した点83と、遮蔽板85で覆われていない舟体上面とすり板下面の境界15A部分のエッジの平均位置88を通る直線を引き、その直線を舟体上面とすり板下面の境界15Aとしたが、この場合は、この直線の傾きを舟体15の傾きとする。
【0173】
上記のように境界線から傾きを測定し、その傾きが所定値以上である場合、舟体15が異常であると判断する。
【0174】
実施の形態32.
図39は、傾き検出のフローであり、図37は直線の傾きを表した図であり、LINE1〜LINE10はHOUGH変換より求めた直線、B1、B2は角度の範囲、A1、A2は範囲B1、B2をそれぞれ移動させる移動角度である。
【0175】
エッジを強調して抽出しやすくするため撮影した画像を1/Nに縮小し(ステップS700)、画像全体に対してエッジを検出する(ステップS701)。そして検出したエッジの点群からHOUGH変換により複数本の直線を求め、次に求めた直線上のエッジの点数が多い順番に上位10本の直線を選択し、また、その直線の傾きも算出する(ステップS702)。
【0176】
図40は選択した10本の直線LINE1〜LINE10を傾き毎に並べた図であるが、まず、−7度〜+7度の範囲内では、B1度の範囲をA1度単位でずらしていき、それぞれの角度の所でのB1度の範囲内に存在する直線を抽出する(ステップS703、704)。
【0177】
図40では、1個のB1度の範囲内にLINE5〜LINE8が存在し、次のB1度の範囲内ではLINE6〜LINE9が存在する。そして、LINE5〜LINE8の角度の平均値とLINE5〜LINE8それぞれの直線のエッジ点数の合計を求める。
【0178】
同様にLINE6〜LINE9に対しても角度の平均値とエッジ点数の合計を求める。このような処理を−7度〜+7度の範囲で行う。
【0179】
次に−90度〜−7度と+7度〜+90度の範囲では、B2度の範囲をA2度単位でずらしていき、それぞれの角度の所でのB2度の範囲内に存在する直線を抽出する(ステップS705)。図40では、1個のB2度の範囲内にLINE1〜LINE3が存在し、次のB2度の範囲内ではLINE2〜LINE4が存在する。そして、LINE1〜LINE3の角度の平均値とLINE1〜LINE3それぞれの直線のエッジ点数の合計を求める(ステップS706)。同様にLINE2〜LINE4に対しても角度の平均値とエッジ点数の合計を求める。このような処理を−90度〜−7度と+7度〜+90度の範囲で行う。
【0180】
このようにして−7度〜+7度、−90度〜−7度と+7度〜+90度の各範囲内での直線のエッジ点数の合計が一番多い範囲にある直線の角度の平均値を、舟体15の傾きとする(ステップS708)。
そして、その傾きが所定値以上である場合、舟体15が異常であると判断する。 また、上記では傾きを範囲内にある直線の平均値としたが、これを範囲内の中心値を傾きとしても良い。
また、−7度〜+7度、−90度〜−7度と+7度〜+90度の各範囲で角度の範囲A1、A2、移動させる角度B1、B2を変えているのは、−5度または+5度を境にして舟体15の傾き異常としており、その近辺での傾き精度を高めるためA1とB1をA2とB2より小さく設定している。
【0181】
実施の形態33.
図41は、1つの舟体の左右を2台のカメラで分担し撮影された左右画像の舟体上面とすり板下面の境界15Aを合わせることを示した図であり、98Aは左画像の舟体上面とすり板下面の境界15Aの境界線、98Bは右画像の舟体上面とすり板下面の境界15Aの境界線、OHは左画像の舟体上面とすり板下面の境界15Aと右画像の舟体上面とすり板下面の境界15Aのオーバーラップ部分、TAはオーバーラップ部分の左画像の厚み値、TBはオーバーラップ部分の右画像の厚み値、厚み平均値Aはオーバーラップ部分の左画像の厚み平均値、厚み平均値Bはオーバーラップ部分の右画像の厚み平均値、境界の平均位置は舟体端側での舟体上面とすり板下面の境界15Aの平均位置である。
【0182】
舟体15の側面状態、撮影条件等の悪い状態で撮影された左右画像の境界線98A、98Bの検出ずれ、または厚み値のずれ、および遮蔽板85の舟体下面18側の一辺から舟体上面とすり板下面の境界15Aまでの大きさの違い、あるいは遮蔽板18の取付方のずれ等により発生する境界線のずれを補正するための方法を示す。
【0183】
舟体15の中央部分をオーバーラップして撮影された左右それぞれの画像で、求められた左右画像それぞれの舟体上面とすり板下面の境界15Aの境界線98Aと98Bは、舟体中央部分でオーバーラップ部分OHだけオーバーラップしている。そのオーバーラップ部分OHの左右画像それぞれの厚み値TA、TBの平均値を求める。そして、その求めた左右画像それぞれの厚み平均値A、Bを比較し、厚み平均値(厚み平均値A<厚み平均値Bとする。)の大きな右側画像の舟体上面とすり板下面の境界15Aの境界線98Bを、小さな厚み平均値Aの位置と、右側画像の舟体端側での舟体上面とすり板下面の境界15Aの平均位置の2点を通る直線98Cとする。左画像の舟体上面とすり板下面の境界15Aは、そのままとする。
【0184】
このようにして左右画像での舟体上面とすり板下面の境界15Aの境界線98A、98Bの検出ずれ、または厚み値のずれ等を補正することができる。
【0185】
【発明の効果】
発明によれば、撮影した画像上に写っている舟***置を、照射される太陽光や照明の強度と方向など明暗環境の異なる条件下において撮影された画像は、舟体上面が太陽光で白く写った画像や黒く写った画像があり、また舟体の形状の違い等により撮影される画像は多種多様である。このような画像に対応できるように異なる複数枚の画像すなわちテンプレート画像をもとにそのテンプレート画像と撮影画像を重ね合わせることで、テンプレート画像と最もマッチする位置を撮影画像の舟***置とし、そして、テンプレートのマッチした位置から舟体上面とすり板下面の境界を検出するための範囲を設定することができ、また、撮影タイミングがずれて画像上の舟***置がずれても舟体上面とすり板下面の境界を検出するための範囲を設定でき、境界を検出できるという効果がある。
【0188】
発明によれば、撮影した画像上に写っている舟***置を、照射される太陽光や照明の強度と方向など明暗環境の異なる条件下において撮影された画像は、舟体上面が太陽光で白く写った画像や黒く写った画像、また舟体の形状の違い等により撮影される画像は多種多様である。このような画像に対応できるように異なる複数枚の画像すなわちテンプレート画像をもとにそのテンプレート画像と撮影画像を重ね合わせることで、テンプレート画像と最もマッチする位置を撮影画像の舟***置とし、そして、テンプレートのマッチした位置から舟体下面を検出する検査範囲を設定することができ、また、撮影タイミングがずれて画像上の舟***置がずれても舟体下面検査範囲を設定でき、境界を求めることができるという効果がある。
【0192】
発明によれば、舟体長手方向の各位置毎にすり板上面側から舟体上面とすり板下面の境界側に輝度の差分をとり、正の最大値を示した箇所Aを求め、そのA点からすり板上面側に一定の範囲内で輝度の差分値がもっとも大きく、しかも所定の値より大きい箇所BがA点以外に存在するとき、B点を舟体とすり板の境界の候補点とし、また、B点が見いだせない時は、A点を候補点とすることで、黒い帯状の影となった舟体上面とすり板下面の境界を検出でき、しかも、撮影条件、または舟体側面の状態が悪く、境界が分かりにくい画像に対しても、舟体上面とすり板下面の境界の候補となるエッジを検出できるという効果がある。
【0193】
発明によれば、舟体とすり板の境界の候補点から舟体下面方向に所定の輝度値より明るい画素が所定の距離以上連続して存在する条件を満たす候補点が、長手方向に所定以上連続して存在するとき、この条件を満たす候補点全てを舟体とすり板下面の境界の候補点から除外することで、舟体上面とすり板下面の境界が太陽光等により反射し、その部分で求めた境界線は本来の境界に沿って引いていないためずれた境界線となるが、そのような境界を除外することでその影響を防ぐことができるという効果がある。
【0194】
発明によれば、すり板下面境界抽出範囲を舟体長手方向に複数ブロックに分け、そのブロック毎に舟体上面とすり板下面の境界の候補点を代表する直線を求め、そして隣接するブロックの直線の交点で連結した折れ線を舟体上面とすり板下面との境界とすることで、湾曲した舟体に対して舟体上面とすり板下面の境界を直線で近似し境界を求めることができるという効果がある。
【0195】
発明によれば、舟体長手方向に複数ブロックに分けて、各ブロック毎に舟体上面とすり板下面の境界候補点を代表する直線を求め、その直線を求めた候補点数が所定値以上となった直線のみを用いて舟体上面とすり板下面の境界を設定するが、所定値未満の区間が所定値以上の区間にはさまれた場合には、隣接する区間の直線どうしの交点で連結するものとし、所定値未満の区間が端の場合は隣接する区間の直線を延長し、得られた折れ線を舟体とすり板の境界とすることで、舟体上面とすり板下面の境界が分かりにくいため、舟体上面とすり板下面の境界の候補点以外の点を抽出することがあり、その点群から求めた直線は本来の境界でないため、その直線を除外することができ、そして舟体上面とすり板下面の境界を求めることができるという効果がある。
【0196】
発明によれば、舟体長手方向に複数ブロックに分けて、各ブロック毎に舟体上面とすり板下面の境界候補点を代表する直線を求め、その直線を求めた候補点数が所定値以上となった直線のみを用いて舟体上面とすり板下面の境界を設定するが、所定値未満の区間が舟体中央部分であるとき、舟体中央部分の下端を抽出し求めた直線を所定の距離だけ上方向に平行移動したものを舟体中央付近のブロック内の直線として採用することで、舟体上面とすり板下面の境界が分かりにくいため、舟体上面とすり板下面の境界の候補点以外の点を抽出することがあり、その点群から求めた直線は本来の境界でないため、その直線を除外することができ、そして舟体中央部分の下端を抽出し、求めた直線を境界線とすることで、隣接したブロックの延長線が曲がっているよな場合でも舟体上面とすり板下面の境界を求めることができるという効果がある。
【0197】
発明によれば、舟体長手方向に複数ブロックに分けて、各ブロック毎に舟体上面とすり板下面の境界候補点を代表する直線を求め、その直線を求めた候補点数が所定値以上となった直線のみを用いて舟体上面とすり板下面の境界を設定するが、所定値未満の区間が舟体中央部分であるとき、舟体中央部分の下端を抽出してその平均値を求め、その平均位置を所定の距離だけ上方向に平行移動した点と、隣接するブロック内の直線端から引いた前記点を通る直線を舟体中央付近のブロック内の直線とすることで、舟体上面とすり板下面の境界が分かりにくいため、舟体上面とすり板下面の境界の候補点以外の点を抽出することがあり、その点群から求めた直線は本来の境界でないため、その直線を除外することができ、また、舟体下面が歪んでいるため、その下面の直線が下面に沿って引けない場合でも、平均値を利用することで舟体上面とすり板下面の境界を求めることができるという効果がある。
【0198】
発明によれば、撮影した舟体画像に対して、舟体上面とすり板下面の境界部分から舟体下面まで遮蔽する長方形状の遮蔽板の横方向の1辺のエッジを抽出し求めた直線を所定の距離だけ平行移動したものを舟体上面とすり板下面の境界とすることで、舟体上面とすり板下面の境界が隠されているため境界を検出できない舟体に対しても境界を設定することができるという効果がある。
【0199】
発明によれば、撮影した舟体画像に対して、舟体上面とすり板下面の境界部分から舟体下面まで遮蔽する長方形状の遮蔽板の横方向の1辺のエッジを抽出し求めた直線の傾きで、舟体上面とすり板下面の境界が遮蔽板に隠されていない部分の境界の平均位置を通る直線を舟体上面とすり板下面の境界とすることで、舟体上面とすり板下面の境界が隠されているため境界を検出できない舟体に対しても境界を設定することができ、また、大きさが違う遮蔽板が装着されているため平行移動する量が特定できない場合でも境界線を引くことができるという効果がある。
【0200】
発明によれば、撮影した舟体画像に対して、舟体上面とすり板下面の境界部分から舟体下面まで遮蔽する長方形状の遮蔽板の横方向の1辺のエッジを抽出し平均位置を求め、その平均位置を所定の寸法だけ上方向に移動した点と、舟体上面とすり板下面の境界が遮蔽板に隠されていない部分の境界の平均位置を通る直線を舟体上面とすり板下面の境界とすることで、舟体上面とすり板下面の境界が隠されているため境界を検出できない舟体に対しても境界を設定することができ、しかも遮蔽板の一辺が歪み直線が傾くような場合でも、平均値を取ることで境界線を引くことができるという効果がある。
【図面の簡単な説明】
【図1】 パンタグラフすり板厚み測定装置の概略構成図である。
【図2】 パンタグラフの撮像画像を示す図である。
【図3】 テンプレート画像を示す図である。
【図4】 パターンマッチングの動作を説明するフローチャートである。
【図5】 パターンマッチングの概要を説明する図である。
【図6】 パターンマッチング後の検査範囲設定を示した図である。
【図7】 金属すり板の舟体の図である。
【図8】 金属すり板上面エッジから撮影画像上の舟体検知方法を説明するフローチャートである。
【図9】 金属すり板上面エッジから撮影画像上の舟体検知方法を説明するフローチャートである。
【図10】 すり板上面エッジからすり板厚み算出、異常検出を行う検査範囲設定を示した図である。
【図11】 すり板上面エッジの向こう側エッジからのすり板厚み算出、異常検出を行う検査範囲設定を示した図である。
【図12】 カーボンすり板上面エッジから撮影画像上の舟体検知を示した図である。
【図13】 カーボンすり板上面エッジから撮影画像上の舟体検知方法を説明するフローチャートである。
【図14】 カーボンすり板上面エッジから撮影画像上の舟体検知方法を説明するフローチャートである。
【図15】 カーボンすり板のすり板上面エッジからのすり板厚み算出、異常検出を行う検査範囲設定を示した図である。
【図16】 カーボンすり板のすり板上面エッジの向こう側エッジからのすり板厚み算出、異常検出を行う検査範囲設定を示した図である。
【図17】 すり板上面エッジから舟体下面の検査範囲設定を示した図である。
【図18】 舟体上面とすり板下面の境界を検出するための範囲設定を示した図である。
【図19】 すり板上面エッジを複数ブロックに分割してそのブロックの平均位置から舟体上面とすり板下面との境界を検出するための範囲設定を示した図である。
【図20】 すり板上面エッジと舟体下面から舟体上面とすり板下面との境界を検出するための範囲設定を示した図である。
【図21】 すり板上面エッジの直線成分から舟体上面とすり板下面の境界を検出するための範囲設定を示した図である。
【図22】 すり板上面の直線成分から舟体上面とすり板下面の境界を検出するための範囲設定を示した図である。
【図23】 すり板上面エッジから舟体上面とすり板下面との境界を検出するための範囲設定を示した図である。
【図24】 舟体上面とすり板下面の境界エッジ検出を示した図である。
【図25】 舟体上面とすり板下面の境界エッジ検出方法を説明するフローチャートである。
【図26】 舟体上面とすり板下面の境界候補点の評価を示した図である。
【図27】 舟体上面とすり板下面の境界の検出を示した図である。
【図28】 舟体上面とすり板下面の境界の検出(あるブロックの直線を利用しない場合)を示した図である。
【図29】 舟体上面とすり板下面の境界線検出(舟体下面の直線を利用)を示した図である。
【図30】 舟体上面とすり板下面の境界の検出(舟体下面の平均位置を利用)を示した図である。
【図31】 遮蔽板により舟体上面とすり板下面の境界が見えない舟体の境界の検出(遮蔽板の一辺を利用)を示した図である。
【図32】 遮蔽板により舟体上面とすり板下面の境界が見えない舟体の境界の検出(遮蔽板の一辺と境界の見える部分の平均位置を利用)を示した図である。
【図33】 遮蔽板により舟体上面とすり板下面の境界が見えない舟体の境界の検出(遮蔽板の一辺の平均位置と境界の見える部分の平均位置を利用)を示した図である。
【図34】 すり板上面エッジの求め方を示した図である。
【図35】 すり板上面エッジの検出開始範囲を示した図である。
【図36】 段摩耗の判定を示した図である。
【図37】 舟体両側側面から段摩耗を判定を示した図である。
【図38】 欠けの判定を示した図である。
【図39】 傾き検出方法を説明するフローチャートである。
【図40】 直線の傾きを表した図である。
【図41】 左右画像の境界線を示した図である。
【図42】 従来のシステムを示した図である。
【符号の説明】
1 走行する列車、2 パンタグラフ、3 車輪検知センサ、4 ランダムシャッタカメラ、5 パンタグラフ検知センサ、6 照明、7 画像処理装置、8画像入力手段、9 保存手段、10 表示手段、11 制御手段、12 画像処理手段、13 モニタ、14 列車進行方向、15 舟体、16(15A) 舟体上面とすり板下面の境界、17 すり板、18 舟体下面、19 テンプレート画像、20 検査範囲、21 舟体下面検査範囲、22 舟体長手方向への寸法、23 舟体端までの舟体下面方向への寸法、24 舟体下面までの舟体下面方向への寸法、25 テンプレート画像の1点。[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a non-contact pantograph sliding plate thickness measuring method for detecting use limit surplus or abnormal wear of a sliding plate mounted on a pantograph of a railway vehicle and collecting current by contacting with a trolley wire.
[0002]
[Prior art]
FIG. 42 is a block diagram showing a conventional pantograph slip plate thickness measuring device disclosed in, for example, Japanese Patent Laid-Open No. 9-265524.
42, 201 is a traveling train, 202 is a pantograph, 203 is a train detection sensor, 204 is a random shutter camera that images the pantograph 202, 205 is a pantograph detection sensor, 206 is illumination that illuminates the pantograph 202, and 207 is image processing. The main body of the device. The image processing apparatus 207 receives an analog video signal of the pantograph 202 from the random shutter camera 204 and converts it into digital data, an image input unit 208 that stores the digital video signal, a storage unit 209 that stores the digitally converted video signal, and displays the stored video signal. A display unit 210 that processes the signal, a control unit 211 that controls the entire apparatus, and an image processing unit 212 that processes a captured image of the pantograph 202 and measures the thickness of the pantograph 202. Reference numeral 213 denotes a monitor that displays an image of the photographed pantograph 202, and an arrow 216 denotes a train traveling direction.
[0003]
Next, the operation of the conventional apparatus will be described.
When the train 201 enters from the direction indicated by the arrow 216 and the train detection sensor 203 detects that the train has entered, the control unit 211 of the image processing device 207 turns on the illumination 206 to photograph the pantograph 202, and The aperture of the random shutter camera 204 is controlled. Each time the pantograph detection sensor 205 detects the pantograph 202, the analog video signal of the pantograph 202 from the random shutter camera 204 is input to the image input unit 208, and the analog video signal is digitally converted by the image input unit 208. Then, image data (data obtained by digitally converting the video signal is referred to as image data) is stored in the storage unit 209.
[0004]
The image data stored in the storage unit 209 calculates the thickness of the pantograph 203 by the image processing unit 212 after passing the train. In the thickness calculation by the image processing means 212, the boundary between the boat body and the sliding board is obtained by a single straight line, and pixels between the straight line and the edge of the upper surface of the sliding board are converted into thickness.
[0005]
[Problems to be solved by the invention]
A conventional pantograph sliding plate thickness measuring device is configured as described above. If the boundary between the boat body and the sliding plate is difficult to see due to the influence of sunlight or dirt, or if the boat body curves and the boat body and the sliding plate There is no method to deal with the case where the boundary is not determined by a single straight line, or the case where the shooting timing is shifted and the position of the boat on the image is shifted. In these cases, the thickness of the sliding plate is not shown. Cannot be calculated accurately. Also, no method for detecting abnormal wear on the upper surface of the sliding plate is shown.
[0006]
The present invention has been made to solve the above-described problems. When the boundary between the hull and the sliding board is not visible, in the case of a curved hull, the shooting timing is shifted and the hull position on the image is shifted. An object of the present invention is to provide a method for measuring the thickness of a pantograph sliding plate that can measure the thickness of the pantograph sliding plate and detect abnormal wear.
[0007]
[Means for Solving the Problems]
  BookThe thickness measurement method of the pantograph slip plate according to the invention is:Prepare multiple images of various hulls taken under different light and dark conditions such as the intensity and direction of sunlight and lighting as templates, and perform pattern matching processing on the images to be inspected to obtain the highest degree of matching A large template and a range for detecting the boundary between the upper surface of the boat body and the lower surface of the sliding plate appearing on the inspection target image from the position are set.
[0010]
  BookThe method for measuring the thickness of a pantograph strip according to the present invention is to prepare a plurality of images of various hulls photographed under different light and dark conditions such as sunlight and illumination intensity and direction as templates, and to obtain a photographed image to be inspected. Pattern matching processing is performed on the template, and a range for detecting the template having the highest degree of coincidence and the lower surface of the hull reflected on the inspection target image from the position is set.
[0014]
  BookThe thickness measurement method of the pantograph sliding plate according to the invention showed a positive maximum value by taking a difference in luminance from the upper surface side of the sliding plate to the boundary side of the upper surface of the boat body and the lower surface of the sliding plate for each position in the longitudinal direction of the boat body. A point A is obtained, and when a point B having a largest difference value within a certain range from the point A to the top surface side of the sliding plate and larger than a predetermined value exists other than the point A, the point B is a hull and a ground plate If the point B cannot be found, the point A is set as a candidate point.
[0015]
  BookThe method for measuring the thickness of a pantograph sliding board according to the present invention is a candidate point that satisfies a condition that pixels brighter than a predetermined luminance value continuously exist for a predetermined distance or more from a candidate point at a boundary between a boat body and a sliding board toward a lower surface of the boat body. However, when it exists continuously for a predetermined length or more in the longitudinal direction, all candidate points that satisfy this condition are excluded from candidate points at the boundary between the upper surface of the boat body and the lower surface of the sliding plate.
[0016]
  BookThe thickness measurement method of the pantograph sliding plate according to the invention divides the sliding plate lower surface boundary extraction range into a plurality of blocks in the longitudinal direction of the hull, and a straight line representing the candidate point of the boundary between the upper surface of the hull and the lower surface of the sliding plate for each block. Then, the broken line connected at the intersection of the straight lines of the adjacent blocks is defined as the boundary between the upper surface of the boat body and the lower surface of the sliding plate.
[0017]
  BookThe thickness measurement method of the pantograph sliding plate according to the invention is divided into a plurality of blocks in the longitudinal direction of the hull, and a straight line representing a boundary candidate point between the upper surface of the hull and the lower surface of the sliding plate is obtained for each block, and the straight line is obtained. The boundary between the upper surface of the hull and the bottom surface of the sliding plate is set using only the straight line where the number of candidate points is equal to or greater than a predetermined value, but if a section less than the predetermined value is sandwiched by a section greater than the predetermined value, it is adjacent It shall be connected at the intersection of the straight lines of the sections, and if the section less than the predetermined value is the end, the straight line of the adjacent section is extended, and the obtained broken line is used as the boundary between the hull and the sliding board.
[0018]
  BookThe thickness measurement method of the pantograph sliding plate according to the invention is divided into a plurality of blocks in the longitudinal direction of the hull, and a straight line representing a boundary candidate point between the upper surface of the hull and the lower surface of the sliding plate is obtained for each block, and the straight line is obtained. The boundary between the upper surface of the hull and the bottom surface of the sliding plate is set using only the straight line where the number of candidate points is greater than or equal to the predetermined value. When the section below the predetermined value is the hull central portion, the lower end of the hull central portion is extracted. A straight line obtained by translating the obtained straight line upward by a predetermined distance is adopted as a straight line in the block near the center of the hull.
[0019]
  BookThe thickness measurement method of the pantograph sliding plate according to the invention is divided into a plurality of blocks in the longitudinal direction of the hull, and a straight line representing a boundary candidate point between the upper surface of the hull and the lower surface of the sliding plate is obtained for each block, and the straight line is obtained. The boundary between the upper surface of the hull and the bottom surface of the sliding plate is set using only the straight line where the number of candidate points is greater than or equal to the predetermined value. When the section below the predetermined value is the hull central portion, the lower end of the hull central portion is extracted. The average value is obtained, and a straight line passing through the point obtained by translating the average position upward by a predetermined distance and the point drawn from the straight line end in the adjacent block is a straight line in the block near the center of the hull. And
[0020]
  BookThe thickness measurement method of the pantograph sliding plate according to the invention is such that the boat of the shielding plate is attached to a boat body equipped with a rectangular shielding plate that shields from the boundary portion between the upper surface of the boat body and the lower surface of the sliding plate to the lower surface of the boat body. The boundary between the upper surface of the boat body and the lower surface of the sliding plate is defined as a straight line obtained by extracting one edge in the lateral direction on the lower surface side of the body and moving in parallel upward by a predetermined distance.
[0021]
  BookThe thickness measurement method of the pantograph sliding plate according to the invention is such that a rectangular shielding plate that shields from the boundary between the upper surface of the boat body and the lower surface of the sliding plate to the lower surface of the boat body on the side of the lower surface of the boat body. The straight line passing through the average position of the boundary of the part where the boundary between the upper surface of the boat and the lower surface of the sliding plate is not concealed by the shielding plate, with the slope of the straight line obtained by extracting one edge of the direction, The boundary of
[0022]
  BookThe thickness measurement method of the pantograph sliding plate according to the invention is such that a rectangular shielding plate that shields from the boundary between the upper surface of the boat body and the lower surface of the sliding plate to the lower surface of the boat body on the side of the lower surface of the boat body. Extract the edge of one side of the direction to obtain the average position, move the average position upward by a predetermined dimension, and the boundary of the part where the boundary between the hull upper surface and the bottom surface of the sliding plate is not hidden by the shielding plate A straight line passing through the average position is defined as the boundary between the upper surface of the boat and the lower surface of the sliding plate.
[0031]
DETAILED DESCRIPTION OF THE INVENTION
Embodiment 1 FIG.
1 is a configuration diagram of a pantograph slip plate thickness measuring apparatus according to Embodiment 1 of the present invention. 1 is a traveling train, 2 is a pantograph on a train roof which is a subject to be photographed, 3 is a wheel detection sensor for detecting the wheels of the train 1, 4 is a photographing means, and the front side of the hull at both ends of the pantograph 2 Random shutter camera for photographing each rear side surface, 5 for a pantograph detection sensor for detecting the hull of the pantograph 2, 6 for illumination for illuminating the photographing area, 7 for an image processing device, and 8 for a random shutter camera for inputting an image at an arbitrary timing 4 is an image input means for converting the video signal from 4 into digital data, 9 is a storage means for saving the data input from the image input means 8, 10 is a display means for outputting the image signal, and 11 is a whole pantograph slab thickness measuring device. 12 is an image processing means for image processing of input image data, and 13 is an image signal input from the display means 10. By monitor for displaying an image, 14 is the train travel direction. The image processing means 12 is common to the processing execution means related to the pantograph thickness measurement method in each embodiment described later.
[0032]
Next, the operation will be described. When the train 1 enters from the direction indicated by the arrow 14 and the wheel detection sensor 3b detects the wheel before the wheel detection sensor 3a, the control means 11 determines that the train 1 has entered. When the wheel detection sensor 3a detects the wheel before the wheel detection sensor 3b, the control means 11 determines that the train 1 has left. When it is determined that the train 1 has arrived, the following measurement operation is performed.
[0033]
First, when the illumination 6 is turned on, and then the train 1 advances and the pantograph detection sensor 5 detects the pantograph 2, the image input means 8 simultaneously activates the random shutter cameras 4a, 4b, 4c, and 4d, and the random shutter camera 4c. 4d is a picture of the pantograph 2 boat 15 (see FIG. 2), the side surface of the sliding plate 17 divided into two in the longitudinal direction of the pantograph 2, and the random shutter cameras 4a and 4b are pantograph 2 boats. The surface of the sliding plate 17 on the front side of the body 15 is divided into two in the longitudinal direction of the hull of the pantograph 2 and photographed.
[0034]
The photographed video signal of the boat body 15 is input to the image input means 8, converted into digital data of 0 to 255 proportional to the brightness for each pixel, and stored in the storage means 9. In the stored image data, the brighter the pixel, the closer the value is to 255, and 255 is the brightest value.
[0035]
When the train 1 enters, the boat bodies 15 of all the pantographs 2 mounted on the train 1 are sequentially photographed by the random shutter camera 4 in this way, input to the image input means 8 and stored in the storage means 9.
[0036]
Finally, after the train passes, the control means 11 turns off the illumination 6.
The display unit 10 displays the following image processing state on the monitor 13 as necessary.
[0037]
When the train 1 passes, the image processing means 12 performs image processing for each image on the image data of the hull 15 input and stored by the image input means 8.
First, pattern matching is performed to detect the position of the boat body 15 from the image data.
The template image 19 shown in FIG. 3 is photographed under different conditions of the light and dark environment, such as the sunlight and the intensity and direction of illumination, which are irradiated to the ship body 15 for each camera, for each of the metal and carbon friction plates. 2 and 3, various images of the hull state such as an image in which the upper surface of the sliding plate 17 appears bright and white, an image in which the upper surface of the sliding plate 17 appears dark and black, and an image in which the sliding plate 17 is greatly worn out. Multiple images are prepared. A template image 19 shown in FIG. 3 is an image obtained by cutting out the hull and reducing it to 1/8.
[0038]
Next, the pattern matching operation will be described based on the flowchart of FIG. 4 and the outline of pattern matching shown in FIG. First, after inputting the target image, the image to be measured is reduced to 1/8 (steps S400 and 401), and a plurality of template images 19 (No. 1 to No. N) prepared for the image are obtained. Are sequentially processed (step S402).
[0039]
At that time, first, the target image is further reduced to 1/2 (1/16 of the original size), and the template image 19 is moved in the x and y directions from the upper left corner of the target image, and the highest correlation is obtained. The position (x0, y0) is obtained. Then, an accurate search is performed with 25 pixels around the 1/8 size image of the original image at that position, and the position with the highest correlation is obtained (steps S406 to S409).
[0040]
These processes are performed on all the template images 19 (No. 1 to No. N), and the maximum correlation value Qmax and the position (x, y) at that time among the template images 19 (No. 1 to No. N) are obtained. Output (step S403). Then, it is determined whether the relative value Q of the correlation value is 60% or more (step S404). If the correlation value is 60% or more, the position (x0, y0) of the template image 19 is adopted as the hull position on the image (step S405).
[0041]
That is, as shown in FIG. 5, the inspection range position (x, y) + dimension in the position (x, y) + template (T) in the target image (I) is set as the inspection range. However, if the correlation value is not 60% or more, the search is performed by detecting the upper edge of the sliding plate (step S410), the hull position on the image is obtained, and the inspection range is set.
[0042]
Placed by a dimension OFFSET_Y22 in the longitudinal direction of the hull from one point 25 of the template image 19 matched by pattern matching to the end of the hull, and a dimension OX23 in the direction of the hull lower surface 18 to a place including the upper surface of the sliding board 17 6 as a reference position, as shown in FIG. 6, the length L1 from the ship end to the center of the hull in the longitudinal direction of the hull from the reference position, and the hull upper surface that does not reach the hull lower surface 18 in the sliding plate thickness direction, A range having a width W1 including the boundary 15A on the lower surface of the slip plate is set, and the range is set as an inspection range 20 in which the thickness of the slip plate is calculated and abnormality is detected.
[0043]
This range shall be the minimum size including the range to be inspected regardless of individual differences between boats or differences in inclination when passing. By performing processing within such a limited range, the possibility of erroneous detection of irrelevant patterns or the like can be minimized.
The dimension OFFSET_Y22 in the longitudinal direction of the boat body and the dimension OX23 in the thickness direction are set to optimum values for each template image 19 having a different shape.
[0044]
Embodiment 2. FIG.
When the highest correlation value is less than 60% in the pattern matching of the first embodiment, the following processing is performed.
FIG. 7 is a view of the boat body 15 of a metal slip plate, and FIGS. 8 and 9 show a flow of detecting the boat body 15 on the photographed image from the upper edge of the metal slip plate.
[0045]
First, after finding the inspection range (step S500), the edge level is set to 15 (step S). Since the initial edge level is 4 or more, the process proceeds to the next step S503 (step S502).
[0046]
In step S502, the background is photographed darker than the far side edge of the sliding plate 17, and the far side edge of the sliding plate 17 changes from a dark portion to a bright portion when viewed from the right to the left side of the image. Therefore, detection is started from the right end of the image in FIG. 7, and an edge where the difference in which the luminance value of the adjacent pixel changes from dark luminance to bright luminance is 15 or more from the right to the left of the image is obtained to the left end of the image. Such processing is detected at all positions in the longitudinal direction of the hull (step S503).
[0047]
The detected edge has a width of 32 pixels in the thickness direction of the slab (left and right direction of the screen), and there are N or more edges in each block divided into eight in the longitudinal direction of the hull (up and down direction of the screen). When there are six or more blocks, the position is set as the edge A on the upper surface of the sliding plate (step S504). When there are no more than 6 blocks, the edge level is reduced from 15 to 4 one by one to obtain the upper surface edge A of the sliding plate 17 (step S501A).
[0048]
Next, an edge where the luminance value of the pixel changes from bright to dark at the point 85 mm ahead of the width of the boat body 15 from the upper surface edge A of the sliding plate 17 toward the left of the image is detected. Step S505). This is performed for all longitudinal positions of the hull. This is because the side surface of the sliding plate 17 is photographed darker than the upper surface of the sliding plate 17, and the sliding plate upper surface edge 26 (see FIG. 6) is changed from a bright portion to a dark portion when viewed from the right to the left. This is because it has changed.
[0049]
The detected edges are 32 pixels wide in the thickness direction of the slab and there are N or more edges in each block divided into 8 in the longitudinal direction of the hull, and there are 6 or more blocks. Suppose that there is a linear component there. Then, the position is set as the upper surface edge B of the sliding plate 17 (step S506). If there are no more than 6 blocks, the edge level is reduced from 14 to 4 one by one and recalculated from the upper edge A of the sliding plate 17 (step S501A).
[0050]
Next, the luminance value of the pixel is changed from a bright luminance to a dark luminance at a point 25 mm ahead of the width of the sliding plate 17 from the upper surface edge A of the sliding plate 17 where the upper surface edge B exists to the left of the image. An edge is detected (step S507). This is performed for all longitudinal positions of the hull. This is because the side surface of the sliding plate 17 is photographed darker than the upper surface of the sliding plate 17, and the upper edge c of the sliding plate changes from a bright portion to a dark portion when viewed from the right to the left. This is because.
[0051]
The detected edges are 32 pixels wide in the thickness direction of the slab and there are N or more edges in each block divided into 8 in the longitudinal direction of the hull, and there are 6 or more blocks. When there is a linear component there. The position is defined as the upper surface edge C of the sliding plate 17 (step S508: see FIG. 9). Then, it is assumed that the upper surface edge A of the sliding plate 17 is the edge on the far side of the boat body 15 and the position thereof is the boat body position on the photographed image (step S509). The upper surface edge B of the sliding plate 17 is a sliding plate upper surface edge 26 of a metal sliding plate.
[0052]
When the edge A on the upper surface of the sliding plate 17 cannot be detected even if the edge level is decreased by 1 from 15 and becomes 4 (step S502), the process proceeds to step S510 in FIG. 9 and the luminance value of the pixel from the right end of the image toward the left. Detects an edge from dark luminance to bright luminance, or from bright luminance to dark luminance (step S510). Such a process is performed at all positions in the longitudinal direction of the hull. The difference in brightness is obtained because there is no difference between the far side edge of the sliding plate 17 and the background, the background appears brighter, and the far side edge portion of the sliding plate 17 looks from the right to the left of the image. This is because there is a case where a bright part changes to a dark part.
[0053]
The detected edge has a width of 32 pixels in the direction of the thickness of the sliding plate, and there are N or more edges in each block divided into 8 in the longitudinal direction of the boat, and there are more than 6 blocks. When there is a linear component there. Then, the edge is set as the upper surface edge A ′ of the sliding plate 17 (step S511).
[0054]
Next, the luminance value of the pixel is increased from bright to dark at the point 85 mm ahead of the width of the hull 15 in the left direction of the image from the upper surface edge A ′ of the sliding plate 17 obtained in the above-described method. Is detected (step S512). This is performed for all longitudinal positions of the hull. The detected edge has a width of 32 pixels in the direction of the thickness of the sliding plate, and there are N or more edges in each block divided into 8 in the longitudinal direction of the hull, and there are more than 6 blocks. At this time, it is assumed that there is a linear component (step S513).
[0055]
Then, assuming that the position is the upper surface edge A ′ of the sliding plate 17, the upper surface edge A ′ of the sliding plate 17 is the other side edge of the boat body 15, and the position is the boat body position on the photographed image. Suppose that there is (step S514). The upper surface edge B 'of the slip plate 17 is the upper surface edge 26 of the metal slide plate.
[0056]
Thus, by obtaining the upper surface edge of the sliding plate 17 and using the width of the hull 15 that is a design value, the hull position on the image that could not be detected by pattern matching can be detected.
[0057]
Further, when the upper surface edge A ′ of the sliding plate 17 is not obtained, it is determined that the boat body 15 is not shown in the photographed image.
[0058]
Embodiment 3 FIG.
FIG. 10 is a diagram showing an inspection range setting for calculating the sliding plate thickness from the sliding plate upper surface edge 26 and detecting an abnormality, and SW1 is a width that does not reach the boat lower surface 18 from the sliding plate upper surface edge 26 in the boat lower surface direction. SW2 is a width about half the width of the sliding plate 17 in the direction opposite to the bottom surface of the boat body from the sliding plate upper surface edge 26, and L3 is a width from the center of the boat body to the edge of the boat body in the longitudinal direction of the boat body.
[0059]
Although it does not reach the hull lower surface 18 to the hull lower surface 18 side from the straight line component obtained from the upper surface edge 26 of the slide plate on the image obtained as in the second embodiment, it includes a boundary 15A between the hull upper surface and the bottom surface of the hull plate. A constant width SW1 and a width SW2 which is about half of the width of the sliding plate 17 are taken on the opposite side, and in order to measure the total thickness of the sliding plate 17 attached to the boat body 15, the hull in the hull length direction. A range obtained by taking a certain length L3 from the center side to the end of the hull is defined as an inspection range 20 in which the thickness of the slab is calculated and abnormality is detected. By setting the range in this way, the inspection range can be specified and easily obtained when calculating the sliding plate thickness and detecting the abnormality.
[0060]
Embodiment 4 FIG.
FIG. 11 is a diagram showing the inspection range setting for calculating the sliding plate thickness from the opposite edge of the sliding plate upper surface edge 26 and detecting the abnormality, 29 is the dimension from the other edge of the sliding plate, and W3 is the direction of the boat lower surface. A constant width that does not reach the bottom of the hull, L3 is a constant width in the longitudinal direction of the hull.
[0061]
From the linear component obtained at the edge A or A ′ of the upper surface of the sliding plate 17 on the other side of the hull position on the image obtained as in the second embodiment, the upper edge 26 of the sliding plate reaches to the left side of the image. A width W3 including a boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate is taken from the position moved by the dimension 29 that does not reach the lower surface 18 of the boat body in the direction of the lower surface 18 of the boat body.
[0062]
Further, in order to measure the total thickness of the sliding plate 17 attached to the hull 15, the range obtained by taking the length L3 from the center of the hull to the end of the hull in the length direction of the hull is calculated. An inspection range 20 is set.
By setting the range in this way, the inspection range can be specified and easily obtained when calculating the sliding plate thickness and detecting the abnormality.
[0063]
Embodiment 5 FIG.
FIG. 12 is a diagram showing the hull detection on the photographed image from the upper edge of the carbon slip plate, and FIGS. 13 and 14 are flowcharts showing the hull detection on the photographed image from the upper edge of the carbon slip plate.
[0064]
Since the side surface of the sliding plate 17 is photographed darker than the upper surface of the sliding plate 17, the upper surface edge 26 of the sliding plate changes from the bright portion to the dark portion from the right to the left of the image. Detection is started, and an edge is found from the right side to the left side of the image where the luminance value of the adjacent pixel becomes 15 or more from the bright luminance to the dark luminance up to the left end of the image. And this process is detected in the position of all the hull longitudinal directions (step S517).
[0065]
When the detected edge has a width of 32 pixels in the direction of the thickness of the slab and there are N or more edges in each block divided into 8 in the longitudinal direction of the hull, and there are more than 6 blocks, This position is defined as an edge E on the upper surface of the grinding plate (step S518). If there are no more than 6 blocks, the edge level is decreased from 14 to 4 one by one to obtain the upper surface edge E of the sliding plate 17 (step S521).
[0066]
Next, since the upper surface of the sliding plate 17 is brighter than the background, the opposite edge of the sliding plate 17 changes from a dark portion to a bright portion from the right to the left of the image. An edge where the luminance value of the pixel changes from dark luminance to bright luminance in the right direction of the image is detected at a point 40 mm after the width of 15 (right direction of the image). This is performed for all the longitudinal positions of the hull (step S519).
[0067]
The detected edge has a width of 32 pixels in the direction of the thickness of the sliding plate, and there are N or more edges in each block divided into 8 in the longitudinal direction of the hull, and there are more than 6 blocks. At this time, it is assumed that there is a linear component (step S520). Then, the position is set as the upper surface edge D of the sliding plate 17 (step S521).
[0068]
Further, the upper surface edge E of the ground plate 17 is a ground plate upper surface edge 26 of the carbon ground plate. When there are no more than 6 blocks, the edge level is decreased from 14 to 4 one by one (steps S521 and 516), and the edge level is obtained again from the edge E on the upper surface of the sliding plate. In the case of a carbon sliding plate, one sliding plate 17 is mounted on the boat body 15, and the width of the boat body 15 and the width of the sliding plate 17 are the same.
[0069]
When the upper edge E of the sliding plate 17 cannot be detected even if the edge level is decreased by 1 from 15 and becomes 4, the edge whose pixel luminance value changes from bright to dark from the right end of the image to the left is detected ( Step S523). This is performed for all longitudinal positions of the hull. When the detected edge has a width of 32 pixels in the thickness direction of the slab and there are N or more edges in each block divided into eight in the longitudinal direction of the boat, and there are more than 6 blocks. Suppose that there is a linear component. Then, the edge is set as the upper surface edge E ′ of the sliding plate 17 (step S524).
[0070]
Then, the luminance value of the pixel increases from bright to dark or dark from the upper edge E ′ of the sliding plate 17 toward the left of the image at a point 40 mm after the width of the boat body 15 (right direction of the image). Edges with bright brightness are detected (step S525). This is performed for all longitudinal positions of the hull. The difference in brightness is obtained because there is no difference between the far side edge of the sliding plate 17 and the background, the background appears brighter, and the far side edge portion of the sliding plate 17 looks from the right to the left of the image. This is because there is a case where a bright part changes to a dark part.
[0071]
The detected edges are 32 pixels wide in the thickness direction of the slab and there are N or more edges in each block divided into 8 in the longitudinal direction of the hull, and there are 6 or more blocks. It is assumed that there is a linear component there (step S526). Then, assuming that the position is the edge D ′ on the top surface of the sliding plate (step S527), the edge D ′ on the top surface of the sliding plate is the far side edge of the boat body 15, and the position is the boat on the photographed image. Suppose that it is a body position. An edge E ′ on the upper surface of the slip plate 17 is a slip plate upper surface edge 26 of the carbon scrap plate.
[0072]
Thus, by obtaining the upper surface edge of the sliding plate 17 and using the width of the hull 15 that is a design value, the hull position on the image that could not be detected by pattern matching can be detected.
If the edge D 'on the upper surface of the sliding plate is not obtained, it is determined that the boat body 15 is not shown in the photographed image.
[0073]
Embodiment 6 FIG.
FIG. 15 is a diagram showing an inspection range setting for calculating and detecting the thickness of the slab from the top surface edge of the slab, and SW3 is a constant width from the top surface edge 26 of the slab to the bottom surface of the hull, SW4 A constant width in the direction opposite to the bottom surface of the boat body from the top surface edge 26 of the shroud plate, and L4 is a constant width in the longitudinal direction of the boat body.
[0074]
The boundary between the upper surface of the boat body and the lower surface of the sliding plate, although it does not reach the lower surface 18 of the boat body 18 from the straight line component obtained from the upper edge 26 of the boat position on the image obtained as in the fifth embodiment. The width SW3 including 15A and the width SW4 of about half the width of the sliding plate 17 are taken on the opposite side.
[0075]
Further, in order to measure all the thicknesses of the sliding plate 17 mounted on the hull 15, a range of a certain length L 4 from the center of the hull to the end of the hull is calculated in the hull length direction, The inspection range 20 in which abnormality detection is performed. By setting the range in this way, the range can be specified and easily obtained when calculating the sliding plate thickness and detecting the abnormality.
[0076]
Embodiment 7 FIG.
FIG. 16 is a view showing the inspection range setting for calculating the slip plate thickness and detecting the anomaly from the edge on the upper side of the slide plate, 33 is the dimension from the other side of the slide plate, and W4 is in the direction of the lower surface of the hull. A constant width that does not reach the lower surface of the hull, L4 is a constant width in the longitudinal direction of the hull.
[0077]
A dimension that does not reach the bottom edge 26 of the lower surface of the sliding plate from the straight line component obtained at the edge D or D ′ of the upper surface of the sliding plate on the other side of the hull position on the image obtained as in the fifth embodiment. A width W4 including a boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate, which does not reach the lower surface 18 of the boat body from the position moved by the OK 33, but is attached to the boat body 15. In order to measure all 17 thicknesses, a range in which a certain length L4 is taken from the center of the hull to the end of the hull in the hull length direction is defined as an inspection range 20 for calculating the slab thickness and detecting an abnormality. By setting the range in this way, the inspection range can be specified and easily obtained when calculating the sliding plate thickness and detecting the abnormality.
[0078]
Embodiment 8 FIG.
An inspection range setting method for the hull lower surface 18 from the hull position matched by the pattern matching shown in the first embodiment will be described. In FIG. 6, the position of the dimension OU24 that reaches the boundary 15A between the upper surface of the boat body and the bottom surface of the sliding plate is not used as a reference. As the inspection range, a fixed width W2 including the hull lower surface 18 in the direction of the sliding plate thickness from the reference position and a fixed length L2 from the center of the hull in the longitudinal direction of the hull are set. A hull lower surface inspection range 21 for detecting the hull lower surface 18 is set.
[0079]
By setting the range in this way, when the edge of the lower surface 18 of the hull is detected and the straight line component or average position of the lower surface is obtained, the inspection range can be specified and is easily obtained. The dimension OU24 in the boat lower surface direction is set to an optimum value for each template image 19 having a different shape.
[0080]
Embodiment 9 FIG.
The inspection range of the hull lower surface 18 is not set from one point 25 of the template image obtained by pattern matching, but the upper edge of the sliding plate 17 shown in the second or fifth embodiment is obtained and the edge is obtained. The hull bottom surface inspection range 21 may be set.
FIG. 17 is a diagram when the boat body lower surface inspection range 21 is set from the sliding plate upper surface edge 26, and 27 is a dimension OFFSET_O in the direction from the sliding plate upper surface edge 26 to the boat body lower surface 18.
[0081]
In the case of a metal sliding plate, the boat is moved in the direction from the sliding plate upper surface edge 26 (B or B ′ in FIG. 9) to the boat lower surface 18 in the case of a carbon sliding plate from the sliding plate upper surface edge 26 (E or E ′ in FIG. 13). Taking a position of the dimension OO27 that does not reach the body lower surface 18 as a reference, a certain width W2 including the boat lower surface 18 is taken in the direction of the boat lower surface 18 from the reference. The boat body longitudinal direction takes a length L2 from the end of the inspection range 20 on the center side of the boat body to the boat body longitudinal direction to the boat body lower surface 18.
[0082]
The range of the constant width W2 and the constant length L2 is referred to as a hull lower surface inspection range 21. By setting the range in this way, when the edge of the lower surface 18 of the hull is detected and the straight line component or average position of the lower surface is obtained, the inspection range can be specified and is easily obtained.
[0083]
Embodiment 10 FIG.
FIG. 18 is a diagram showing a range setting for detecting the boundary between the upper surface of the hull and the lower surface of the sliding plate. Dimension OX36 from the point 25 of the template image 19 matched by pattern matching to the front of the boundary 15A between the upper surface of the boat body and the lower surface of the slide plate in the thickness direction of the slab and the edge of the boat body from the point 25 of the template image to the longitudinal direction of the craft body In order to measure all the thicknesses of the sliding plate 17 attached to the boat body 15 on the basis of the position of the dimension OY35 until, the length L5 from the reference position to the boat body center side in the boat body longitudinal direction, A range having a width W5 including a boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate is set in the thickness direction of the sliding plate, and the range is set as a range 34 for detecting the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate. By setting the range in this way, the edge of the lower surface 18 of the hull is detected, and the range can be specified when the linear component or average position of the lower surface is determined, which makes it easy to determine.
[0084]
The dimension OFFSET_36 from the point 25 of the template image to the front of the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate in the thickness direction of the slab, and the dimension OY35 from the point 25 of the template image to the end of the boat body in the longitudinal direction. Is set to an optimum value for each template image 19 having a different shape.
[0085]
Embodiment 11 FIG.
FIG. 19 is a diagram showing a range setting for a boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate from the average position of the block by dividing the sliding plate upper surface edge 26 into a plurality of blocks in the longitudinal direction of the boat body. 37 is a block A obtained by dividing the top surface edge 26 of the sliding plate into a plurality (three blocks in FIG. 19), 38 is a block B obtained by dividing the top surface edge 26 of the sliding plate into a plurality (three blocks in FIG. 19), and 39 is a top surface edge of the sliding plate Is divided into a plurality of blocks (3 blocks in FIG. 19), 40 is the average position of the upper surface edge 26 of the sliding plate of block A, 41 is the average position of the upper surface edge 26 of the sliding plate of block B, and 42 is the sliding plate of block C Average position of the upper surface edge 26, 43 is a dimension from the average position of the upper surface edge 26 of the sliding plate of the block A to the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate, 44 is the upper surface of the sliding plate of the block B 45 is a dimension from the average position of the wedge 26 to the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate, 45 is a dimension from the average position of the upper surface edge 26 of the sliding plate of the block C to the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate, F1 Is a position moved by the dimension 43 from the average position 40 of the block A, F2 is a position moved by the dimension 44 from the average position 41 of the block B, F3 is a position moved by the dimension 45 from the average position 42 of the block C, SW5 is F1 , F2 and F3, the width from the broken line connecting the positions of F1, F2, and F3 to the lower surface 18 of the hull, SW6 is the direction of the upper surface of the sliding plate 17 from the broken line connecting the positions F1, F2, and F3 The width L6 that does not reach the top surface edge 26 of the slab plate is the length from the center of the hull to the end of the hull in the longitudinal direction of the hull.
[0086]
A case will be described in which the sliding plate upper surface edge 26 is divided into a plurality of blocks in the longitudinal direction of the hull and divided into 3 blocks by a range setting method for obtaining the boundary 15 between the hull upper surface and the sliding plate lower surface from the average position of the blocks. The reason why the upper surface edge 26 of the sliding plate is divided into a plurality of longitudinal directions of the boat body is that the wear amount of the sliding plate 17 is different between the edge of the boat body and the center side of the boat body. Dividing to set a boundary 15A between the upper surface and the lower surface of the sliding plate.
[0087]
In one block A in which the top surface edge 26 of the sliding plate is divided into three blocks, the luminance difference with the adjacent pixel is different for each position in the longitudinal direction (y direction) of all the hulls in the sliding plate thickness direction (x direction). The largest edge is obtained, and the average position 40 is obtained from the X pixel and Y pixel of the edge for each position in the longitudinal direction of the obtained hull.
[0088]
Similarly, the average position 41 of the upper surface edge 26 of the plate within the block B and the average position 42 of the upper surface edge 26 of the plate within the block C are obtained. Then, the position F1 moved by a dimension 43 from the average position 40 of the block A to the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate, and the dimension from the average position 41 of the block B to the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate. The hull bottom surface 18 is centered on a broken line connecting three points: a position F2 moved by 44 and a position F3 moved by a dimension 45 from the average position 42 of the block C to the boundary 15A between the hull upper surface and the bottom surface of the sliding plate. The width SW5 that does not reach the boat lower surface 18 in the direction of the width SW6, the width SW6 that does not reach the upper plate edge 26 in the direction of the upper surface of the sliding plate 17, and all the sliding plates 17 attached to the boat body 15 In order to measure the thickness of the boat, the range obtained by taking the length L6 from the center of the hull to the end of the hull in the longitudinal direction of the hull is defined as a range 34 for detecting the boundary 15A between the hull upper surface and the bottom of the sliding plate.
[0089]
In addition, the width of the dimension from the broken line connecting the points F1, F2, and F3 to the hull lower surface 18 side and the sliding plate upper surface side is changed at each point, and the boundary between the hull upper surface and the sliding plate lower surface is changed. It is good also as a range for detecting 15A.
[0090]
By previously setting the range 34 including the boundary 15A in this way, other edge components are not erroneously detected in the process of detecting the boundary 15A.
[0091]
In the above example, the top surface edge 26 of the sliding plate is divided into three blocks, but the processing may be performed by dividing the number of blocks other than three blocks.
The average position is obtained as follows.
[0092]
X position of average position = (× 1 + × 2 + × 3 +... + × n) / n
Average position Y position = (Y1 + Y2 + Y3 +... + Yn) / n
[0093]
X1, x2,..., Xn are x pixels of edges obtained for each position in the longitudinal direction of the hull.
Y1, Y2,..., Yn are Y pixels at the edge obtained for each position in the longitudinal direction of the hull.
n is the number of each position in the longitudinal direction of the hull obtained in the block for obtaining the average position.
[0094]
Embodiment 12 FIG.
FIG. 20 is a diagram showing a range setting for detecting a boundary 15A between the upper surface of the boat body and the lower surface of the slide plate from the upper surface edge 26 of the slide plate and the lower surface 18 of the boat body. A range for obtaining an average position of the edge 26, 47 a range for obtaining an average position of the upper surface edge 26 of the sliding plate between the edge of the hull and the center of the hull, 48, an average position of the upper surface edge 26 of the sliding plate at the hull end side, 49 is an average position of the upper surface edge 26 of the sliding board between the edge of the hull and the center of the hull, 50 is an average position of the edge of the lower surface 18 of the hull, and 51 is an average position of the upper edge 26 of the sliding board on the hull end side. Dimension to the boundary 15A between the upper surface of the hull and the lower surface of the sliding plate, 52 is a dimension from the average position of the upper surface 26 of the sliding plate between the edge of the hull and the center of the hull to the boundary 15A between the upper surface of the hull and the lower surface of the sliding plate, 53 is the average position of the edge of the lower surface 18 of the hull and the upper surface of the hull and below the sliding plate F4 is the position moved by the dimension 51 from the average edge 48 of the edge 46 in the range 46 on the edge of the hull, and F5 is the average position of the edge in the range 47 between the edge of the hull and the center of the hull. The position moved from the position 49 by the dimension 52, F6 moved from the average position 50 of the edge of the boat lower surface 18 by the dimension 53, SW7 from the broken line connecting the positions F4, F5, F6 to the direction of the boat lower surface 18 The width SW8 does not reach the lower surface 18 of the boat, SW8 is the width from the broken line connecting the positions F1, F2, and F3 to the upper surface of the sliding plate 17 and does not reach the upper surface edge 26 of the sliding plate, L7 is the longitudinal direction of the boat The length from the center of the hull to the end of the hull.
[0095]
It is within the range 46 in which the average position of the top edge 26 of the hull edge on the edge of the hull that takes a certain length (about half of the block obtained by dividing the upper edge of the hull board into three parts) on the upper face side of the hull edge The edge having the largest luminance difference with the adjacent pixel is obtained for each position in the ship body longitudinal direction in the plate thickness direction (X direction), and the obtained edge x pixel for each position in the ship body longitudinal direction, An average position 48 is obtained from the Y pixel.
[0096]
Similarly, within the range 47 for obtaining the average position of the upper surface edge 26 of the sliding plate between the edge of the hull and the center of the hull, which has a fixed length (about half of the block obtained by dividing the upper surface edge of the sliding plate into three). The average position 49 of the upper edge 26 of the sliding plate on the edge of the hull is obtained.
[0097]
Further, an average position 50 of the edge of the boat lower surface 18 is obtained within the boat lower surface inspection range 21. And the position F4 which moved the dimension 51 from the average position 48 of the upper surface edge 26 of the sliding board upper surface to the boundary 15A of the upper surface of the boat body and the lower surface of the sliding board, and the upper surface of the sliding board between the edge of the boat body and the center of the boat body. The position F5 moved by the dimension 52 from the average position 49 of the edge 26 to the boundary 15A between the upper surface of the hull and the bottom of the sliding board, and the dimension 53 moved from the average position 50 of the lower surface of the hull 18 to the boundary 15A of the upper surface of the hull and the bottom of the sliding board. Centering on a broken line connecting three points of position F6, a range having a width SW7 of a dimension on the lower surface 18 side of the hull and a width SW8 of a dimension on the upper surface side of the sliding plate, and a slip attached to the hull 15 In order to measure all the thicknesses of the plate 17, a boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate is detected within a range in which a certain length L7 is taken from the center of the boat body to the edge of the boat body with respect to the longitudinal direction of the boat body. For this reason, the range 34 is used.
[0098]
Further, the width of the dimension from the broken line connecting the points F4, F5, F6 to the boat lower surface 18 side and the sliding plate upper surface side is changed for each point, and the boundary 15A between the boat upper surface and the sliding plate lower surface is changed. It is good also as a range for detecting.
[0099]
The ranges 48 and 49 for obtaining the average value of the upper surface of the slab are set to about half of the block obtained by dividing the edge of the upper surface of the slab into three, but this may be obtained for the entire block obtained by dividing the block into three. Further, the average value of the lower surface 18 of the hull is utilized because of the amount of wear of the sliding plate 17 on the center side of the hull, there is an opening in the dimension that moves the average value of the upper surface edge 26 of the sliding plate to the boundary 15A. It is possible to prevent the range 34 of the boundary 15A from expanding. The dimension between the hull lower surface 18 and the boundary 15A is constant.
[0100]
By previously setting the range 34 in this way, other edge components are not erroneously detected in the process of detecting the boundary 15A.
[0101]
Embodiment 13 FIG.
FIG. 21 is a diagram showing a range setting for detecting a boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate from the linear component of the upper surface edge 26 of the sliding plate, 54 is a linear component of the upper surface edge 26 of the sliding plate, and 55A The dimension from the upper edge 26 of the sliding plate to the boundary 15A between the upper surface of the boat and the lower surface of the sliding plate, 54A is a straight line translated by a dimension 55A from the upper edge 26 of the sliding plate to the boundary 15A between the upper surface of the boat and the lower surface of the sliding plate, SW10 The width from the straight line 54A of the sliding plate upper surface edge 26 to the lower surface of the boat body in the direction of the lower surface of the boat, SW11 is the width from the straight line 54A of the sliding plate upper surface edge 26 to the upper surface edge 26 of the sliding plate in the direction of the upper surface of the sliding plate, L8 is a certain length from the center of the hull to the end of the hull in the longitudinal direction of the hull.
[0102]
The top surface edge 26 of the sliding plate is obtained for each position in the longitudinal direction of the boat within the inspection range 20, and the linear component 54 of the top surface edge 26 of the sliding plate is obtained from that edge. As the straight line component, a straight line having the largest number of edges on the straight line component is selected from the arrangement of the points by HOUGH conversion.
[0103]
Alternatively, the linear component 54 may be obtained from the top surface edge 26 of the sliding plate by the least square method. Then, the obtained straight line 54 of the sliding plate upper surface edge 26 is translated by a dimension 55A from the sliding plate upper surface edge 26 to the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate.
[0104]
Then, in order to take a range including the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate in the direction of the thickness of the sliding plate, A width SW10 that does not reach from the straight line 54A to the bottom surface of the boat body in the bottom direction of the boat body and a width SW11 that extends from the straight line 54A of the top surface edge 26 of the sliding plate to the top surface edge 26 of the sliding plate in the top direction of the sliding plate. In order to measure all the thicknesses of the sliding plate 17 attached to the body 15, a range obtained by taking a certain length L8 from the center of the hull to the end of the hull in the longitudinal direction (Y direction) of the hull is defined as the hull upper surface. A range 34 for detecting the boundary 15 </ b> A on the bottom surface of the sliding plate is used.
[0105]
By setting such a range, it is possible to specify the boundary 15 between the upper surface of the boat body and the lower surface of the sliding plate and not detect edges other than the boundary 15.
[0106]
Embodiment 14 FIG.
FIG. 22 is a diagram showing a range setting for detecting the boundary between the upper surface of the boat body and the lower surface of the sliding plate from the linear component 54 of the upper surface edge 26 of the sliding plate as a method other than the embodiment 13, and 55 (n) A dimension that does not reach the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate, and 56 (n) is the upper surface edge 26 of the sliding plate at each position in the longitudinal direction of the boat body. Width from the dimension 55 (n) to the position including the boundary 15A between the upper surface of the hull and the bottom surface of the sliding board in the direction of the lower surface of the hull, L9 is constant from the center of the hull to the end of the hull in the longitudinal direction of the hull Is the length of
[0107]
The top surface edge 26 of the sliding plate is obtained for each position in the longitudinal direction of the boat within the inspection range 20, and the linear component 54 of the top surface edge 26 of the sliding plate is obtained from that edge. As the straight line component 54, the straight line having the largest number of edges on the straight line component is selected from the arrangement of the points by the HOUGH conversion.
[0108]
Alternatively, the linear component 54 may be obtained from the top surface edge 26 of the sliding plate by the least square method. Then, from the first position in the longitudinal direction of the hull on the obtained linear component 54, the dimension 55 (1) that has moved from the sliding plate upper surface edge 26 to the boundary 15A between the upper surface of the boat and the lower surface of the sliding plate has been moved. Taking a point as a reference, a width 56 (1) is taken from the point in the direction of the hull lower surface 18 to a place including the boundary 15A between the hull upper surface and the bottom surface of the sliding plate. This is performed to take a range including a boundary 15A between the upper surface of the boat body and the lower surface of the slide plate in the thickness direction of the slide plate for each position in the longitudinal direction of the boat body.
[0109]
Starting from the second position, up to the nth position, each point in the longitudinal direction of the hull is moved by a dimension 55 (n) from the point on the straight line component 54, and a fixed dimension 56 is determined from that point. Take the width of (n). Then, in order to measure all the obtained width 55 (n) and the thickness of the sliding plate 17 attached to the hull 15 for each position in the hull longitudinal direction, the hull center side in the hull longitudinal direction is measured. A range obtained by taking a certain length L9 from the ship body to the edge of the hull is defined as a range 34 for detecting the boundary 15 between the hull upper surface and the bottom surface of the sliding plate.
[0110]
Embodiment 15 FIG.
FIG. 23 is a diagram showing that the inspection range is set in order to detect only the boundary 15A when detecting the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate from each edge of the upper surface edge 26 of the sliding plate. is there.
[0111]
The upper surface edge 26 of the sliding plate is obtained for each position in the longitudinal direction of the hull within the inspection range 20, and first, the first edge in the longitudinal direction of the hull is defined as the boundary between the upper surface of the sliding plate 26 and the upper surface of the hull and the lower surface of the sliding plate. With reference to a point that has moved by a dimension H (1) that does not reach the boundary 15A in the 15A direction, the width H ′ (1 )I take the.
[0112]
This takes this width H '(1) in order to detect the boundary 15A between the upper surface of the hull and the lower surface of the sliding plate. Similarly, from the second point to the nth point, each width H (n) is determined from that point by moving each dimension H (n) from the sliding plate upper surface edge 26 for each position in the longitudinal direction of the hull. take. Then, in order to measure the width H ′ (n) at each position in the longitudinal direction of the hull and the thickness of all the sliding plates 17 attached to the hull 15, the boat from the center side of the hull in the longitudinal direction of the hull is measured. A range in which a certain length L10 is taken to the end of the body is a range for detecting the boundary between the upper surface of the boat body and the lower surface of the sliding plate.
[0113]
Embodiment 16 FIG.
FIG. 24 is a view showing how to obtain candidate points for the boundary 15A between the upper surface of the hull and the lower surface of the sliding plate, and FIG. 24 is an edge detection flow of the boundary 15A between the upper surface of the hull and the lower surface of the sliding plate. .
[0114]
In the case of a metal sliding plate, in order to detect the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate, the adjacent boats from the pixels on the upper surface side of the sliding plate 17 to all positions in the longitudinal direction of the boat body within the range 34. When the pixel difference on the body lower surface 18 side is taken, the edge A (n) having the largest difference value is detected (step S600). In the edge A (n), a change from a small luminance (dark) to a large luminance (bright) from the upper surface side of the sliding plate 17 toward the boat lower surface 18 becomes the largest edge (step S601).
[0115]
Next, within a range of a fixed return amount 60 (for example, 2 mm) from the position of the edge A (n) toward the upper surface of the sliding plate 17, the pixels on the lower surface 18 side of the adjacent boat body from the pixels on the upper surface side of the sliding plate 17. When the difference is taken, the edge B (n) having the largest difference value is detected (step S602). In the edge B (n), a change from a small luminance (dark) to a large luminance (bright) becomes the largest edge from the boat body lower surface 18 toward the upper surface side of the sliding plate 17 (steps S603 and S604).
[0116]
However, when the edge B (n) obtained at each position is a difference value greater than or equal to a predetermined value, the edge B (n) is adopted at that position (steps S605 and 607). At the position, the edge A (n) obtained first is adopted (steps S605 and 606).
[0117]
Thus, the edge A (n) or B (n) obtained for each position in the longitudinal direction of the hull is set as a candidate for the boundary 15A between the upper surface of the hull and the lower surface of the sliding plate. This is because when the difference value is equal to or smaller than a predetermined value, the boundary 15A is difficult to understand, and an edge other than the original boundary 15A may be extracted. .
[0118]
Further, in the case of the carbon sliding plate, the width of the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate is wider than the metal sliding plate, and the edge A (n) obtained from the upper surface of the sliding plate toward the lower surface of the boat body 18 is Since this is a candidate for the boundary 15A on the bottom surface of the sliding plate, the carbon sliding plate uses this edge A (n) as a candidate for the boundary 15A between the upper surface of the boat body and the bottom surface of the sliding plate.
In some cases, the difference for obtaining the edge A (n) and the edge B (n) should be between several pixels apart.
Note that n is the total number of positions in the longitudinal direction of the hull.
[0119]
Embodiment 17. FIG.
  FIG. 26 is a diagram showing evaluation of candidate points on the boundary 15A between the upper surface of the hull and the lower surface of the sliding plate. Claim3There are pixels whose brightness value in the direction from the candidate point of the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate in the direction of the boat lower surface 18 exceeds a predetermined width HO, and the width HO Is longer than the predetermined length LO in the longitudinal direction of the hull, the portion 62 is excluded from the candidates for the boundary 15A between the upper surface of the hull and the lower surface of the sliding plate as a portion where light is reflected.
[0120]
The boundary 15A between the upper surface of the hull and the lower surface of the sliding plate is reflected by sunlight, etc., and the boundary line obtained at that portion is not drawn along the original boundary, but is a shifted boundary line, but the boundary is excluded. Can prevent the effect.
[0121]
Embodiment 18 FIG.
FIG. 27 is a diagram showing detection of a boundary 15A between the upper surface of the hull and the lower surface of the sliding plate.
A method of dividing the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate into a plurality of blocks in the longitudinal direction of the boat body, detecting the boundary line for each block, and using the broken line connecting the intersections of the boundary lines as the boundary will be described. The reason for the division into a plurality of blocks is that the boat body 15 is curved and the boundary 15A between the upper surface of the boat body and the lower surface of the sliding board cannot be obtained with a single straight line. It is.
[0122]
Block (1) 63, block (2) 64, block (3) 65,..., Block (n-1) in which the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate is divided into n sections in the longitudinal direction of the boat body 66 and block (n) 67 are divided by overlapping each other.
[0123]
Then, for each block, a straight line is obtained by HOUGH conversion from a point group of edges that are candidates for the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate. The obtained straight lines are the boundary line (1) 68, boundary line (2) 69, boundary line (3) 70,..., Boundary line (n-1) 71, boundary line (n) 72 of each block. It is.
[0124]
Next, the intersection (1) 73 between the boundary line (1) 68 and the boundary line (2) 69, the intersection (2) 74 between the boundary line (2) 69 and the boundary line (3) 70,. A plurality of boundary lines are connected by using the intersection (n-1) 75 of (n-1) 71 and the boundary line (n) 72 as a joint. The boundary line thus obtained is defined as a boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate.
[0125]
When there is no intersection of the boundary lines obtained in each block, the connection is made as follows. When there is no intersection of the boundary line at the edge of the hull or at the center of the hull, for example, when there is no intersection of the boundary line (1) 68 of the block (1) 63 and the boundary line (2) 69 of the block (2) 64 The boundary line (1) 68 is located on the end side of the block (1) 63 (the side opposite to the block (2) 64), the boundary line (2) 69 of the block (2) 64 and the block (3) 65 The intersection 74 of the boundary line (3) 70 is connected to form a boundary line between the block (1) 63 and the block (2) 64.
[0126]
In addition, when there is no intersection point between two adjacent blocks in the two corresponding blocks, the boundary point between two adjacent blocks on both sides is connected to each other, and the boundary between the two blocks without an intersection point A line.
[0127]
Further, when there are no intersections on the boundary lines of all the blocks, the boundary line (1) 68 is located on the end side of the block (1) 63 (the side opposite to the block (2) 64), and the boundary line (n ) 72 is defined as a boundary line connecting a point located on the end side of block (n) 67 (on the opposite side of block (n-1) 66).
Note that n in the above is the number of blocks divided.
[0128]
Embodiment 19. FIG.
FIG. 28 is a diagram illustrating a case where a straight line of a certain block is not used in detection of the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate.
In the eighteenth embodiment, the boundary line of each block is obtained from the edge point group serving as the boundary candidate by HOUGH conversion. If the number of edge points on this boundary line is less than the predetermined number, the boundary line obtained in that block is Do not use. The reason why the number of edge points is less than the predetermined number is that the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate is difficult to understand and the edges vary, and the boundary line thus obtained is not reliable.
[0129]
When a boundary line having a number of edge points smaller than a predetermined number is sandwiched between a predetermined number or more boundary lines, for example, when the boundary line (2) 76 of the block (2) 64 is smaller than the predetermined number, this block (2) 64 The boundary line (1) 68 and the boundary line (3) 70 obtained in the adjacent block (1) 63 and block (3) 65 are extended without using the boundary line (2) 76 of the block, and connected at the intersection. The straight line 78 is the boundary line of that portion.
[0130]
In addition, when a boundary line having a smaller number of edge points than the predetermined number is on the end side, for example, when the boundary line (n) 77 of the block (n) 67 is smaller than the predetermined number, an adjacent block of the block (n) 67 is Since only the block (n-1) 66 is present, the boundary line (n-1) 71 of the block (n-1) 66 is extended to the block (n) 67, and the straight line 79 is used as the boundary line of the portion. .
[0131]
In this way, when the number of edge points is smaller than the predetermined number, the boundary line is connected as described above to obtain the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate.
[0132]
Embodiment 20. FIG.
FIG. 29 is a diagram showing a case where the straight line of the hull lower surface 18 is used when the straight line of the block on the center side of the hull cannot be used by detecting the boundary 15A between the hull upper surface and the bottom surface of the sliding plate.
[0133]
In the eighteenth embodiment, the boundary 15A between the upper surface of the hull and the lower surface of the sliding plate is divided into a plurality of blocks, and the boundary line is obtained for each block, and the boundary 15A between the upper surface of the hull and the lower surface of the sliding plate is connected. When the number of edges of the boundary line (n) 77 obtained in the block (n) 67 in the center of the hull is less than a predetermined number, the ship is not used in the hull lower surface range 21 without using the boundary line (n) 77. An edge of the body lower surface 18 is obtained, and a straight line obtained by performing HOUGH conversion on the point cloud of the edge is defined as a straight line 80 of the boat lower surface 18, and the straight line 80 is directed toward the boundary 15A between the boat upper surface and the sliding plate lower surface. The distance from 18 to the boundary 15A between the upper surface of the hull and the lower surface of the sliding plate is translated. The translated straight line 81 is defined as a boundary line 15A between the upper surface of the boat body of the block (n) and the lower surface of the sliding plate, and the boundary line (n-1) 71 of the straight line 81 and the block (n-1) 66 Intersections are connected to form a boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate.
[0134]
Embodiment 21. FIG.
FIG. 30 is a diagram showing a case where the average position of the boat lower surface 18 is used for detecting the boundary 15A between the boat upper surface and the bottom surface of the sliding plate.
In the eighteenth embodiment, the boundary 15A between the upper surface of the hull and the lower surface of the sliding plate is divided into a plurality of blocks, and the boundary line is obtained for each block, and the boundary 15A between the upper surface of the hull and the lower surface of the sliding plate is connected. When the number of edges of the boundary line (n) 77 obtained in the block (n) 67 at the center of the hull is less than a predetermined number, the lower surface of the hull is within the range 21 of the hull lower surface 18 without using this boundary line. 18 edges are obtained, and an average position 82 of the edges is obtained.
[0135]
Then, the average position 82 is moved in the direction of the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate by a dimension up to the boundary 15A between the lower surface of the boat 18 and the upper surface of the boat body and the lower surface of the sliding plate. The moved point 83 is connected to the end of the boundary line (n-1) 71 of the block (n-1) 66, and the connected line is extended to the end of the block (n) 67. A straight line 84 connecting the point 83 and the boundary line (n−1) 71 is defined as a boundary of the block (n) 67.
[0136]
Embodiment 22. FIG.
FIG. 31 is a diagram illustrating detection of the boundary of the boat body in which the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate cannot be seen by the shielding plate.
A straight line 80A on one side of the shielding plate 85 is obtained from the point group of one edge of the shielding plate 85 on the boat lower surface 18 side within the range 21 of the boat lower surface 18 by HOUGH conversion. Then, a straight line 86 obtained by translating the straight line 80A in the direction of the boundary 15A between the upper surface of the hull and the lower surface of the sliding plate and the boundary 15A between the lower surface of the hull 18 and the upper surface of the hull and the lower surface of the sliding plate is extended to the end of the hull. A straight line 86 and an extended portion 87 are defined as a boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate.
[0137]
Embodiment 23. FIG.
FIG. 32 shows a case in which the boundary 15A of the boat body 15 where the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate cannot be seen by the shielding plate 85 is obtained using the average position of one side of the shielding plate 85 and the portion where the boundary 15A is visible. It is a figure.
[0138]
When a rectangular shielding plate 85 that includes the boundary 15A between the upper surface of the hull and the lower surface of the sliding plate and shields up to the lower surface 18 of the hull is mounted on the hull 15, one side of the shielding plate is parallel to the boundary 15A. By drawing a straight line passing through the visible portion of the boundary 15A with the inclination of this one side, the boundary 15A of the invisible portion can be determined.
[0139]
First, a straight line 80A on one side of the shielding plate 85 is determined by HOUGH conversion from the point group of one edge of the shielding plate 85 on the boat lower surface 18 side within the range 21 of the boat lower surface 18. Next, the edge of the boundary 15A portion between the upper surface of the boat body and the lower surface of the sliding plate that is not covered with the shielding plate 85 is extracted, and the average position 88 is obtained.
[0140]
A straight line 89 is drawn parallel to the straight line 80A on one side of the shielding plate 85 on the boat lower surface 18 side so as to pass through the average position 88, and a straight line 89 extending from the boat end side to the boat center side is obtained. A boundary 15A between the upper surface and the lower surface of the sliding plate is used.
[0141]
Embodiment 24. FIG.
FIG. 33 shows a method for obtaining the boundary of the boat body 15 where the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate cannot be seen by the non-rectangular shielding plate using the average position of one side of the shielding plate and the average position of the portion where the boundary is visible. FIG.
[0142]
When a shielding plate 85 that shields the boundary between the upper surface of the boat body and the lower surface of the sliding plate from the boundary 15A portion to the lower surface of the boat body 18 is attached to the boat body 15, the shielding plate on the lower surface of the boat body 18 within the range 21 of the lower surface of the boat body 18. An average position 82A of one edge of 85 is obtained, and the average position 82A is measured in the direction of the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate from one side of the shielding plate 85 to the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate. This point is 83A.
[0143]
Next, the edge of the boundary 15A portion between the upper surface of the boat body and the lower surface of the sliding plate that is not covered with the shielding plate 85 is extracted, and the average position 88 is obtained.
Then, a straight line 90 passing through the point 83 that has moved the average position 82A of the shielding plate and the average position 88 of the boundary 15A portion between the upper surface of the boat body and the lower surface of the sliding plate that is not covered by the shielding plate 85 is drawn. A line extending from the body end side to the boat center side is defined as a boundary 15A between the boat upper surface and the bottom surface of the sliding plate.
[0144]
Embodiment 25. FIG.
FIG. 34 is a view showing how to obtain the top surface edge 26 of the sliding plate, A (1) is the top surface edge of the sliding plate at the first position in the longitudinal direction of the boat body, and A (2) is the longitudinal direction of the boat body. The top edge of the sliding board at the second position, A (n) is the top edge of the sliding board at the nth position in the longitudinal direction of the boat, and B (1) is the sliding board at the first position in the longitudinal direction of the boat. B (2) is the second edge of the sliding board at the second position in the longitudinal direction of the boat, and B (n) is the opposite edge of the sliding board at the nth position in the longitudinal direction of the boat. , 97 are pixels corresponding to the width of the hull 15, 98 is a boundary line between the upper surface of the hull and the lower surface of the sliding plate, A ′ (n, 1) is the nth position, and the first edge of the upper surface 26 of the upper plate 26. Candidate A ′ (n, 2) is the second candidate of the top surface edge 26 of the plate at the nth position, and A ′ (n, m) is the mth position of the top surface edge 26 of the plate at the nth position. The eye candidate, B ′ (n, 1) is the nth position, and the edge on the far side of the hull width from the first candidate A ′ (n, 1) of the top surface edge 26 of the board, B ′ (N, 2) is the n-th position, and the second edge A ′ (n, 2) of the top edge of the top surface of the board is the far side edge of the boat width, and B ′ (n, m) is n At the eye position, the edge on the far side from the mth candidate A ′ (n, m) of the top surface edge 26 of the board 26, n is the total number of positions in the longitudinal direction of the hull, m is the longitudinal direction of the hull It is the number of candidates for the top edge of the board at a certain position.
[0145]
Within the inspection range 20, the edge of the upper surface of the sliding board is obtained for each position in the longitudinal direction of the boat body from the boundary 15 A between the upper surface of the boat body and the lower surface of the sliding board toward the upper surface of the sliding board 17. First, the edge is obtained from the luminance value of the pixel on the upper surface side of the sliding plate 17 from the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate to the upper surface direction (* direction) of the sliding plate 17 at the first position. Next, the difference value of the luminance value of the pixel on the boundary 15A side between the upper surface of the boat body and the lower surface of the sliding plate is obtained.
[0146]
Then, a plurality of edges A ′ (1,1), A ′ (1,2),..., A ′ (1, m) are obtained. A pixel 97 corresponding to the width of the boat body 15 in the direction from the edge A ′ (1,1), A (1,2),..., A ′ (1, m) to the other edge of the sliding plate 17. First, the difference value of the luminance value of the pixel on the opposite edge side of the sliding plate 17 adjacent to the pixel from the luminance value of the pixel on the upper surface edge 26 side of the sliding plate is obtained, respectively, and edges B ′ (1, 1), B '(1,2), ..., B' (1, m) is obtained. Next, the absolute value of the difference value between the edges A ′ (1, 1), A ′ (1, 2),..., A ′ (1, m) and the edges B ′ (1, 1), B The product of the absolute values of the difference values of '(1, 2), ..., B' (1, m) is obtained.
[0147]
Figure 0004209525
* Here, | A ′ (1,1) |, | A ′ (1,2) |,..., | A ′ (1, m) |, | B ′ (1,1) |, | B ′ (1,2) |,..., | B ′ (1, m) | are the edges A ′ (1,1), A ′ (1,2),. (1, m), B ′ (1, 1), B ′ (1, 2),..., B ′ (1, m) is the absolute value of the difference value.
* The above formula is for the first position in the longitudinal direction of the hull.
[0148]
If the product C (1,1), C (1,2),..., C (1, m) having the largest value is C (1,2), Assume that the edge A ′ (1,2) is the top surface edge A (1) of the sliding plate at the first position.
[0149]
Similarly, the edge A ′ (i, j) (i = 1 to n, j = 1 to m) having the largest product from the second to the nth is obtained.
[0150]
Edges A ′ (i, j) (i = 1 to n, j = 1 to m) obtained in this way are used as the sliding plate upper surface edges A (1), A (2),. n). At this time, B ′ (i, j) (i = 1 to n, j = 1 to m) ahead of the width equivalent pixel number 97 of the boat body 15 is the far side edge B (1), B (2),. .., B (n).
[0151]
Moreover, although the difference value of the pixel is obtained between adjacent pixels, it may be obtained between pixels separated by several pixels or more.
[0152]
Embodiment 26. FIG.
In the method of obtaining the top surface edge 26 of the sliding plate shown in the twenty-fifth embodiment, a plurality of edges A ′ (i, j) (i = 1 to n, j = 1 to m) for each position in the longitudinal direction of the hull are obtained. The difference between the luminance values of the adjacent pixels was obtained. From the characteristics of the photographed image, the sliding plate upper surface edge 26 changed from dark luminance to bright luminance when viewed from the upper surface of the boat body 15 toward the upper surface of the sliding plate 17. Therefore, a plurality of edges A ′ (i, j) (i = 1 to n, j = 1 to m) for each position in the longitudinal direction of the hull are moved from the pixels on the upper surface side of the hull 15 to the upper surface side of the sliding plate. It is possible to extract only edges where the difference obtained by subtracting the pixels shows a negative value.
[0153]
Further, a plurality of edges A ′ (i, j) (i = 1 to n, j = 1 to m) and far side edges B ′ (i, j) (i = 1 to 1) for each position in the longitudinal direction of the hull. n, j = 1 to m), the product may be obtained by betting a larger coefficient toward the upper surface of the sliding plate. This is because noise exists between the boundary 15A between the upper surface of the hull and the lower surface of the sliding plate and the upper edge 26 of the sliding plate, and may be detected. Since there is a high possibility that the edge is the upper surface edge 26 of the sliding plate, the upper surface side of the sliding plate 17 is weighted more heavily.
[0154]
Figure 0004209525
* Here, | A ′ (1,1) |, | A ′ (1,2) |,..., | A ′ (1, m) |, | B ′ (1,1) |, | B ′ (1,2) |,..., | B ′ (1, m) | are the edges A ′ (1,1), A ′ (1,2),. The absolute value of the luminance difference between (1, m), B ′ (1, 1), B ′ (1, 2),..., B ′ (1, m).
kA1, kA2, kAM, kB1, kB2, and kBm are coefficients.
* The above formula is for the first position in the longitudinal direction of the hull.
[0155]
Embodiment 27. FIG.
FIG. 35 is a diagram showing a detection start range of the upper edge of the sliding plate, and shows an example of dividing into three. L (1) is the start position for inspection of the top surface of the slab at 1/6 from the edge of the boat, L (2) is the start position for inspection of the top surface of the slab at 1/6, and L (3) is the position of the hull. 15 is a sliding plate upper surface inspection start position L (3) in the remaining 4/6 portion.
[0156]
In the twenty-fifth embodiment, the inspection of the sliding plate upper surface edge 26 is started from the boundary line 98 between the upper surface of the boat body and the lower surface of the sliding plate, and this is performed for each position in the longitudinal direction of the boat body (Y direction). Change the inspection start position.
[0157]
The boundary line 98 between the upper surface of the boat body and the lower surface of the sliding plate obtained for the curved boat body 15 is shifted to the lower surface of the boat body 18 from the actual boundary, and the upper surface edge 26 of the sliding plate is inspected from the shifted boundary. This is because the boundary 15A between the original upper surface of the hull and the bottom surface of the sliding plate may be erroneously detected as the top surface edge 26 of the sliding plate.
[0158]
Further, even if the upper surface edge 26 of the sliding plate 17 is obtained from the upper surface side of the sliding plate 17 from the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate, a starting position that does not cross the upper surface edge 26 of the sliding plate is set.
[0159]
Then, the inspection of the upper surface edge 26 of the ground plate is started from the position of L (1), L (2), L (3) from the boundary line 98 between the upper surface of the boat body and the lower surface of the sliding plate.
Although the inspection start position is set by dividing it into three in the longitudinal direction of the hull, it may be set by dividing it into a plurality.
[0160]
Embodiment 28. FIG.
FIG. 36 is a diagram showing determination of step wear, where a is the depth of step wear, b is the length of step wear, c is the length of step wear, d is the width of step wear, and 108 is a slide plate. It is thickness.
FIG. 36 shows a 90-degree change from the previous figure.
[0161]
A thickness 108 of the ground plate from the boundary line 98 between the upper surface of the boat body and the lower surface of the sliding plate to the upper surface edge 26 of the sliding plate is obtained for each position in the longitudinal direction of the boat body.
And the difference of the thickness value 108 is calculated | required by the space | interval of the width | variety d. When the difference between the obtained thickness values is greater than or equal to the depth a, an average value of the thickness values 108 is obtained between the width b and the width c of the left and right (left and right in the hull longitudinal direction) of the worn portion. And when the difference of each average value is more than the depth a, the part shall be step wear.
[0162]
Embodiment 29. FIG.
FIG. 37 is a diagram showing a case where step wear is determined from both side surfaces of the hull.
According to the step wear determination shown in the twenty-eighth embodiment, D1 and D2 determined to be step wear on one side surface of the hull 15, and D3 determined to be step wear from an image obtained by photographing the opposite side surface. Compare hand position. D1 is not determined to be step wear on the opposite side surface located at substantially the same position in the longitudinal direction of the hull, so D1 is assumed not to be step wear, and then the opposite side surface located at substantially the same position in the longitudinal direction of D2 and the hull. Since there is D3 determined as step wear, it is assumed that D2 is step wear. Since the step wear occurs over the upper surface from one side surface to the opposite side surface of the boat body 25, the step wear is judged on both side surfaces.
[0163]
Embodiment 30. FIG.
FIG. 38 is a diagram showing the determination of a chip, where ka is the width of the chip, kb is the chip depth for determining a chip candidate, and kc is a candidate part of the chip from kb to the deepest position. , 112 is the maximum thickness value of the part that is a candidate for chipping, 113 is the minimum thickness value of the part that is a candidate for chipping, and 114 is the width of the boat body 15 with the far side edge of the boat body 25 in front. This is a translated edge.
[0164]
A thickness value 108 from a boundary line 98 between the upper surface of the boat body and the lower surface of the sliding plate to the upper surface edge 26 of the sliding plate is obtained. Next, the far side edge of the hull 15 is obtained, and the difference between the edge 114 that has been translated in front of the width of the hull 15 and the upper edge 26 of the sliding plate is obtained. When the difference is more than kb and the width is continuous more than ka, that part is determined as a candidate for lack.
[0165]
If a difference between the maximum thickness value 112 and the minimum thickness value 113 is kc or more in a part that is a candidate for chipping (a part having a width ka or more and a difference kb or more), the part is regarded as a chipping. Since chipping occurs only on one side of the hull 15, determination is performed for each of the front image and the rear image.
[0166]
Embodiment 31. FIG.
In the eighteenth embodiment, the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate is divided into a plurality of blocks, and the boundary line is obtained for each block. The boundary line is set as the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate. The inclination of the boundary line (n) 72 obtained by the block (n) 67 on the center side of the hull is defined as the inclination of the hull 15.
[0167]
In the nineteenth embodiment, the boundary 15A between the upper surface of the boat body and the lower surface of the sliding board is divided into a plurality of blocks, and a boundary line is obtained for each block, and the boundary line of the block whose number of edges is smaller than a predetermined number is used. At this time, since the boundary line (n) 72 of the block (n) 67 on the center side of the hull is smaller than a predetermined number, the boundary line (n−1) of the adjacent block (n−1) 66 ) When 71 is extended and used as a boundary line, the inclination of the boundary line is set as the inclination of the boat body 15.
[0168]
In the twentieth embodiment, the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate is divided into a plurality of blocks, and the boundary line is obtained for each block, and the boundary line (n) obtained by the block (n) 67 at the center of the boat body ) When 77 is less than the predetermined number of edges, the straight line 80 on the lower surface 18 of the hull is translated in the direction of the boundary 15A between the upper surface of the hull and the bottom surface of the sliding plate by the dimension up to the lower surface 18 and the boundary 15A. A straight line 81 is used as a boundary line 15A between the upper surface of the boat body of the block (n) 67 and the lower surface of the sliding plate, and the intersection of the straight line 81 and the boundary line (n-1) 71 of the block (n-1) 68 is connected. Although the boundary 15A between the upper surface of the hull and the lower surface of the sliding plate is used, in this case, the inclination of the straight line of the lower surface 18 of the hull is defined as the inclination of the hull 15.
[0169]
In the twenty-first embodiment, the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate is divided into a plurality of blocks, and the boundary line is obtained for each block, and the boundary line (n) obtained by the block (n) 67 at the center of the boat body ) When 77 is smaller than the predetermined number of edges, the average position 82 of the edge of the hull lower surface 18 is obtained without using this boundary line, and this average position 82 is determined in the direction of the boundary 15A between the hull upper surface and the bottom surface of the sliding plate. The distance between the lower surface 18 of the boat body and the boundary 15A is moved, the moved point 83 is connected to the end of the boundary line (n-1) 71 of the block (n-1) 66, and the connected line is connected to the block ( n) The straight line 84 extended to the end of 67 is defined as the boundary of the block (n) 67. In this case, the inclination of the extended straight line 84 is set as the inclination of the boat body 15.
[0170]
In the twenty-second embodiment, since the shielding plate 85 that shields from the boundary 15A portion between the upper surface of the boat body and the lower surface of the sliding plate to the lower surface 18 of the boat body is attached to the boat body 15, the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate is Since it is not visible, a straight line 80A on one side of the shielding plate 85 on the boat lower surface 18 side is translated in the direction of the boundary 15A between the boat upper surface and the bottom surface of the boat by the size of the boat lower surface 18 and the boundary 15A. 86 is extended to the boat body end side, and this straight line 86 and the extended portion 87 are used as the boundary 15A between the boat upper surface and the bottom surface of the sliding plate. In this case, the inclination of the straight line 86 is the inclination of the boat body 15. .
[0171]
In Embodiment 23, since the shielding plate 85 that shields from the boundary 15A portion between the upper surface of the boat body and the lower surface of the sliding plate to the lower surface 18 of the boat body is attached to the boat body 15, one of the shielding plates 85 on the lower surface 18 side of the boat body. The straight line drawn through the average position 88 of the edge 15A of the boundary between the upper surface of the boat body and the lower surface of the sliding plate that is not covered with the shielding plate 85 due to the inclination of the straight line 80A of the side, Although the boundary 15A is set, the inclination of the straight line is defined as the inclination of the boat body 15.
[0172]
In Embodiment 24, since the shielding plate 85 that shields from the boundary 15A portion between the upper surface of the boat body and the lower surface of the sliding plate to the lower surface 18 of the boat body is attached to the boat body, one side of the shielding plate 85 on the lower surface 18 side of the boat body The average position 82A is obtained, and the average position 82A is moved in the direction of the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate by a dimension to one side of the lower surface 18 side of the shielding plate 85 and the boundary 15A. A straight line that passes through the average position 88 of the edge of the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate that is not covered with 85 is drawn as the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate. The inclination of this straight line is the inclination of the boat body 15.
[0173]
As described above, the inclination is measured from the boundary line, and when the inclination is equal to or greater than a predetermined value, it is determined that the boat body 15 is abnormal.
[0174]
Embodiment 32. FIG.
FIG. 39 is a flowchart of inclination detection, FIG. 37 is a view showing the inclination of a straight line, LINE1 to LINE10 are straight lines obtained by HOUGH conversion, B1 and B2 are angle ranges, and A1 and A2 are range B1. It is a movement angle which moves each B2.
[0175]
The captured image is reduced to 1 / N in order to enhance the edge for easy extraction (step S700), and the edge is detected for the entire image (step S701). Then, a plurality of straight lines are obtained from the detected edge point group by HOUGH conversion, and the top 10 straight lines are selected in descending order of the number of edges on the obtained straight line, and the inclination of the straight line is also calculated. (Step S702).
[0176]
FIG. 40 is a diagram in which ten selected straight lines LINE1 to LINE10 are arranged for each inclination. First, within the range of −7 degrees to +7 degrees, the range of B1 degrees is shifted in units of A1 degrees. A straight line existing within the range of B1 degrees at the angle of is extracted (steps S703 and 704).
[0177]
In FIG. 40, LINE5 to LINE8 exist within one B1 degree range, and LINE6 to LINE9 exist within the next B1 degree range. Then, the sum of the average value of the angles of LINE5 to LINE8 and the number of edge points of each straight line of LINE5 to LINE8 is obtained.
[0178]
Similarly, for LINE6 to LINE9, the average value of angles and the total number of edge points are obtained. Such processing is performed in the range of -7 degrees to +7 degrees.
[0179]
Next, in the range of -90 degrees to -7 degrees and +7 degrees to +90 degrees, the range of B2 degrees is shifted in units of A2 degrees, and a straight line existing within the range of B2 degrees at each angle is extracted. (Step S705). In FIG. 40, LINE1 to LINE3 exist within one B2 degree range, and LINE2 to LINE4 exist within the next B2 degree range. Then, the sum of the average value of the angles of LINE1 to LINE3 and the number of edge points of each line of LINE1 to LINE3 is obtained (step S706). Similarly, for LINE2 to LINE4, an average value of angles and the total number of edge points are obtained. Such processing is performed in a range of −90 degrees to −7 degrees and +7 degrees to +90 degrees.
[0180]
In this way, the average value of the angles of the straight lines in the range where the total number of edge points of the straight lines in each range of −7 degrees to +7 degrees, −90 degrees to −7 degrees and +7 degrees to +90 degrees is the largest is obtained. The inclination of the boat body 15 is set (step S708).
And when the inclination is beyond a predetermined value, it judges that hull 15 is abnormal. In the above description, the slope is the average value of straight lines within the range, but the center value within the range may be the slope.
In addition, the angle ranges A1 and A2 and the moving angles B1 and B2 are changed to −5 degrees or −7 degrees to +7 degrees, −90 degrees to −7 degrees, and +7 degrees to +90 degrees, respectively. The boat body 15 has an inclination abnormality at +5 degrees as a boundary, and A1 and B1 are set smaller than A2 and B2 in order to improve the inclination accuracy in the vicinity thereof.
[0181]
Embodiment 33. FIG.
FIG. 41 is a diagram showing alignment of the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate in the left and right images taken by sharing the left and right sides of one boat body with two cameras. A boundary line 15A between the upper surface of the body and the lower surface of the sliding plate, 98B is a boundary line between the upper surface of the upper surface of the boat and the lower surface of the sliding plate 15A, and OH is a boundary image 15A between the upper surface of the upper surface of the left image and the lower surface of the sliding plate. The overlap portion of the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate, TA is the thickness value of the left image of the overlap portion, TB is the thickness value of the right image of the overlap portion, and the thickness average value A is the left of the overlap portion The average thickness value of the image and the average thickness value B are the average thickness value of the right image of the overlap portion, and the average boundary position is the average position of the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate on the boat body end side.
[0182]
The hull from one side of the hull lower surface 18 side of the shielding plate 85 and the detection deviation of the boundary lines 98A and 98B of the left and right images photographed in a bad state such as the side surface state and the photographing condition of the hull 15 and the deviation of the thickness value. A method for correcting the deviation of the boundary line caused by the difference in size between the upper surface and the lower surface of the sliding plate up to the boundary 15A, or the deviation of the mounting method of the shielding plate 18 will be described.
[0183]
In the left and right images taken by overlapping the central part of the hull 15, the boundary lines 98A and 98B of the boundary 15A between the upper surface of the hull and the lower surface of the sliding board of the obtained left and right images are the central part of the hull. The overlap part OH overlaps. The average values of the thickness values TA and TB of the left and right images of the overlap portion OH are obtained. Then, the thickness average values A and B of the obtained right and left images are compared, and the boundary between the upper surface of the boat body and the lower surface of the sliding plate of the right image having a large thickness average value (thickness average value A <thickness average value B). The boundary line 98B of 15A is defined as a straight line 98C passing through two points: the position of the small thickness average value A and the average position of the boundary 15A of the upper surface of the hull and the lower surface of the sliding plate on the hull edge side of the right image. The boundary 15A between the upper surface of the hull and the lower surface of the sliding plate in the left image is left as it is.
[0184]
In this way, detection deviation of the boundary lines 98A, 98B of the boundary 15A between the upper surface of the boat body and the lower surface of the sliding plate in the left and right images, or deviation of the thickness value can be corrected.
[0185]
【The invention's effect】
  BookAccording to the invention, the image of the hull position shown on the photographed image under the different conditions of the light and dark environment such as the irradiated sunlight and the intensity and direction of the illumination, There are white images and black images, and there are a wide variety of images taken due to differences in the shape of the boat. A plurality of different images, i.e. template images, so as to support such imagesBased on the above, the template image and the captured image are overlapped, and the position that best matches the template image is taken as the hull position of the captured image, and the boundary between the upper surface of the hull and the bottom surface of the sliding board is determined from the position where the template matches. A range for detection can be set, and even if the shooting timing is shifted and the hull position on the image is shifted, a range for detecting the boundary between the upper surface of the hull and the lower surface of the sliding plate can be set. Can detectThere is an effect.
[0188]
  BookAccording to the invention, the image of the hull position shown on the photographed image under the different conditions of the light and dark environment such as the irradiated sunlight and the intensity and direction of the illumination, There are a wide variety of images taken in white, black, and images taken due to differences in the shape of the boat. By superimposing the template image and the photographed image based on a plurality of different images, that is, template images so as to correspond to such an image, the position that best matches the template image is set as the hull position of the photographed image, and , You can set the inspection range to detect the lower surface of the hull from the matching position of the template, and you can set the inspection range of the lower surface of the hull even if the shooting timing shifts and the hull position on the image shifts, There is an effect that it can be obtained.
[0192]
  BookAccording to the invention, for each position in the longitudinal direction of the hull, a difference in luminance is taken from the upper surface side of the sliding plate to the boundary side between the upper surface of the hull and the lower surface of the sliding plate, and a location A showing a positive maximum value is obtained. When there is a point B other than point A where the difference in brightness is the largest within a certain range from the point to the upper surface side of the sliding plate and is larger than the predetermined value, point B is a candidate point for the boundary between the hull and the sliding plate When point B cannot be found, the point A can be used as a candidate point to detect the boundary between the upper surface of the boat body and the lower surface of the sliding plate, which is a black belt-like shadow. There is an effect that an edge that is a candidate for the boundary between the upper surface of the boat body and the lower surface of the sliding plate can be detected even for an image having a poor side surface state and the boundary is difficult to understand.
[0193]
  BookAccording to the invention, a candidate point that satisfies a condition that pixels brighter than a predetermined luminance value continuously exist for a predetermined distance or more from the candidate point of the boundary between the boat body and the sliding board in the lower surface direction of the boat body is a predetermined value or more in the longitudinal direction. When all the candidate points satisfying this condition are excluded from the candidate points at the boundary between the hull and the bottom surface of the sliding plate, the boundary between the upper surface of the boat body and the lower surface of the sliding plate is reflected by sunlight, etc. Since the boundary line obtained in the part is not drawn along the original boundary, it becomes a deviated boundary line. However, it is possible to prevent the influence by excluding such a boundary.
[0194]
  BookAccording to the invention, the bottom surface boundary extraction range of the sliding plate is divided into a plurality of blocks in the longitudinal direction of the hull, and a straight line representing a candidate point of the boundary between the upper surface of the hull and the lower surface of the sliding plate is obtained for each block, and By defining a broken line connected at the intersection of straight lines as the boundary between the upper surface of the hull and the lower surface of the sliding plate, the boundary between the upper surface of the hull and the lower surface of the sliding plate can be approximated by a straight line with respect to the curved hull. There is an effect.
[0195]
  BookAccording to the invention, divided into a plurality of blocks in the longitudinal direction of the hull, a straight line representing the boundary candidate point between the upper surface of the hull and the lower surface of the sliding board is obtained for each block, and the number of candidate points for obtaining the straight line is equal to or greater than a predetermined value. The boundary between the upper surface of the hull and the bottom surface of the sliding plate is set using only the straight line that has become, but if a section less than the predetermined value is sandwiched by a section that is greater than or equal to the predetermined value, the intersection of the straight lines of adjacent sections If the section less than the predetermined value is the end, the straight line of the adjacent section is extended, and the obtained broken line is used as the boundary between the boat body and the sliding board, so that the boundary between the upper surface of the boat body and the lower surface of the sliding board Since it is difficult to understand, points other than the candidate points of the boundary between the upper surface of the hull and the bottom surface of the sliding plate may be extracted, and the straight line obtained from the point cloud is not the original boundary, so the straight line can be excluded, And the boundary between the upper surface of the hull and the lower surface of the sliding plate can be obtained. There is an effect that.
[0196]
  BookAccording to the invention, divided into a plurality of blocks in the longitudinal direction of the hull, a straight line representing the boundary candidate point between the upper surface of the hull and the lower surface of the sliding board is obtained for each block, and the number of candidate points for obtaining the straight line is equal to or greater than a predetermined value. The boundary between the upper surface of the hull and the lower surface of the sliding plate is set using only the straight line that has become, but when the section less than the predetermined value is the central part of the hull, the straight line obtained by extracting the lower end of the central part of the hull is specified By adopting a straight line translated upward by a distance as a straight line in the block near the center of the hull, the boundary between the upper surface of the hull and the lower surface of the sliding plate is difficult to understand. Points other than points may be extracted, and since the straight line obtained from the point cloud is not the original boundary, the straight line can be excluded, and the lower end of the center part of the hull is extracted, and the obtained straight line is the boundary. By making a line, it is an extension of an adjacent block Even if Yo is bent there is an effect that it is possible to determine the collector head top and sliding plate lower surface of the boundary.
[0197]
  BookAccording to the invention, divided into a plurality of blocks in the longitudinal direction of the hull, a straight line representing the boundary candidate point between the upper surface of the hull and the lower surface of the sliding board is obtained for each block, and the number of candidate points for obtaining the straight line is equal to or greater than a predetermined value. The boundary between the upper surface of the hull and the lower surface of the sliding plate is set using only the straight line. When the section below the specified value is the hull central part, the lower end of the hull central part is extracted and the average value is obtained. The hull is defined as a straight line in the block near the center of the hull, which is a straight line passing through the point obtained by translating the average position upward by a predetermined distance and the straight line in the adjacent block. Since the boundary between the upper surface and the lower surface of the sliding plate is difficult to understand, points other than the candidate points for the boundary between the upper surface of the boat and the lower surface of the sliding plate may be extracted, and the straight line obtained from the point group is not the original boundary, so the straight line The bottom of the hull is distorted Because, straight its lower surface, even if no shrinkage along the lower surface, there is an effect that it is possible to determine the collector head top and sliding plate lower surface of the boundary by using an average value.
[0198]
  BookAccording to the present invention, a straight line obtained by extracting an edge of one side in a lateral direction of a rectangular shielding plate that shields from a boundary portion between the upper surface of the boat body and the lower surface of the sliding plate to the lower surface of the boat body from the photographed boat body image. Is the boundary between the upper surface of the hull and the lower surface of the sliding plate, and the boundary between the upper surface of the boat and the lower surface of the sliding plate is hidden. There is an effect that can be set.
[0199]
  BookAccording to the present invention, a straight line obtained by extracting an edge of one side in a lateral direction of a rectangular shielding plate that shields from a boundary portion between the upper surface of the boat body and the lower surface of the sliding plate to the lower surface of the boat body from the photographed boat body image. By setting the straight line passing through the average position of the boundary of the part where the boundary between the upper surface of the boat and the lower surface of the sliding plate is not hidden by the shielding plate as the boundary between the upper surface of the boat and the lower surface of the sliding plate, If the boundary of the lower surface of the plate is hidden, the boundary can be set even for hulls where the boundary cannot be detected, and the amount of translation cannot be specified because a shield with a different size is attached However, there is an effect that a boundary line can be drawn.
[0200]
  BookAccording to the invention, the edge of one side in the lateral direction of the rectangular shielding plate that shields from the boundary between the upper surface of the boat body and the lower surface of the sliding plate to the lower surface of the boat body is extracted from the photographed boat body image, and the average position is calculated. The straight line that passes through the average position of the point where the average position is moved upward by a predetermined dimension and the boundary between the upper surface of the boat body and the lower surface of the sliding plate is not hidden by the shielding plate is rubbed with the upper surface of the boat body. By setting the boundary of the lower surface of the plate, the boundary between the upper surface of the hull and the lower surface of the sliding plate is hidden, so the boundary can be set even for the hull where the boundary cannot be detected. Even when the angle is inclined, the boundary line can be drawn by taking the average value.
[Brief description of the drawings]
FIG. 1 is a schematic configuration diagram of a pantograph slip plate thickness measuring apparatus.
FIG. 2 is a diagram showing a captured image of a pantograph.
FIG. 3 is a diagram showing a template image.
FIG. 4 is a flowchart illustrating an operation of pattern matching.
FIG. 5 is a diagram illustrating an outline of pattern matching.
FIG. 6 is a diagram showing inspection range setting after pattern matching.
FIG. 7 is a diagram of a boat body of metal slip plates.
FIG. 8 is a flowchart for explaining a hull detection method on a photographed image from an upper surface edge of a metal slip plate.
FIG. 9 is a flowchart for explaining a hull detection method on a photographed image from an upper surface edge of a metal slip plate.
FIG. 10 is a diagram showing inspection range setting for calculating the thickness of the slab and detecting the abnormality from the top edge of the slab.
FIG. 11 is a diagram showing an inspection range setting for calculating the thickness of the strip from the other edge of the top surface edge of the strip and detecting the abnormality.
FIG. 12 is a view showing hull detection on a photographed image from the upper edge of the carbon slip plate.
FIG. 13 is a flowchart for explaining a hull detection method on a photographed image from the upper edge of a carbon slip plate.
FIG. 14 is a flowchart for explaining a hull detection method on a photographed image from the upper edge of a carbon slip plate.
FIG. 15 is a diagram showing inspection range setting for calculating the thickness of the scraper from the upper edge of the scraper of the carbon scraper and detecting the abnormality.
FIG. 16 is a diagram showing the setting of the inspection range for calculating the thickness of the scraper from the opposite edge of the upper surface edge of the scraper of the carbon scraper and detecting the abnormality.
FIG. 17 is a diagram showing the inspection range setting from the upper edge of the sliding plate to the lower surface of the hull.
FIG. 18 is a diagram showing a range setting for detecting the boundary between the upper surface of the hull and the lower surface of the sliding plate.
FIG. 19 is a diagram showing a range setting for detecting the boundary between the upper surface of the boat body and the lower surface of the sliding plate from the average position of the block by dividing the upper surface edge of the sliding plate into a plurality of blocks.
FIG. 20 is a diagram showing a range setting for detecting the boundary between the upper surface of the boat body and the lower surface of the ground plate from the upper surface edge of the ground plate and the lower surface of the boat body.
FIG. 21 is a diagram showing a range setting for detecting the boundary between the upper surface of the hull and the lower surface of the sliding plate from the linear component of the upper surface edge of the sliding plate.
FIG. 22 is a diagram showing a range setting for detecting the boundary between the upper surface of the hull and the lower surface of the sliding plate from the linear component on the upper surface of the sliding plate.
FIG. 23 is a diagram showing a range setting for detecting the boundary between the upper surface of the hull and the lower surface of the sliding plate from the upper surface edge of the sliding plate.
FIG. 24 is a diagram showing boundary edge detection between the upper surface of the hull and the lower surface of the sliding plate.
FIG. 25 is a flowchart for explaining a method of detecting a boundary edge between the upper surface of the hull and the lower surface of the sliding plate.
FIG. 26 is a diagram showing evaluation of boundary candidate points between the upper surface of the hull and the lower surface of the sliding plate.
FIG. 27 is a diagram showing detection of the boundary between the upper surface of the hull and the lower surface of the sliding plate.
FIG. 28 is a diagram showing detection of the boundary between the upper surface of the hull and the lower surface of the sliding plate (when a straight line of a certain block is not used).
FIG. 29 is a diagram showing detection of a boundary line between the upper surface of the hull and the lower surface of the sliding board (using a straight line on the lower surface of the hull).
FIG. 30 is a diagram showing detection of the boundary between the upper surface of the hull and the lower surface of the sliding board (using the average position of the lower surface of the hull).
FIG. 31 is a diagram showing detection of the boundary of the boat body (using one side of the shield plate) where the boundary between the upper surface of the boat body and the lower surface of the sliding plate cannot be seen by the shield plate.
FIG. 32 is a diagram showing detection of a boundary of a boat body in which the boundary between the upper surface of the boat body and the lower surface of the sliding plate cannot be seen by the shielding plate (using an average position of one side of the shielding plate and a portion where the boundary is visible).
FIG. 33 is a diagram showing detection of the boundary of the hull in which the boundary between the upper surface of the boat body and the lower surface of the sliding plate cannot be seen by the shielding plate (using the average position of one side of the shielding plate and the average position of the portion where the boundary is visible). .
FIG. 34 is a view showing how to obtain the upper edge of the slidable plate.
FIG. 35 is a diagram showing a detection start range of a top surface edge of a slip plate.
FIG. 36 is a diagram showing determination of step wear.
FIG. 37 is a diagram showing determination of step wear from both sides of a hull.
FIG. 38 is a diagram showing determination of a chipping.
FIG. 39 is a flowchart illustrating an inclination detection method.
FIG. 40 is a diagram showing the slope of a straight line.
FIG. 41 is a diagram illustrating a boundary line between left and right images.
FIG. 42 shows a conventional system.
[Explanation of symbols]
1 Train to travel, 2 Pantograph, 3 Wheel detection sensor, 4 Random shutter camera, 5 Pantograph detection sensor, 6 Illumination, 7 Image processing device, 8 Image input means, 9 Storage means, 10 Display means, 11 Control means, 12 images Processing means, 13 monitor, 14 train traveling direction, 15 hull, 16 (15A) boundary between hull upper surface and bottom of sliding plate, 17 slip plate, 18 hull lower surface, 19 template image, 20 inspection range, 21 hull lower surface Inspection range, 22 dimensions in the hull longitudinal direction, 23 dimensions in the hull lower surface direction to the hull edge, 24 dimensions in the hull lower surface direction to the hull lower surface, 25 one point of the template image.

Claims (11)

通過する電車屋根上に搭載されたパンタグラフの舟体を前方あるいは後方からカメラにより撮影し、得られた画像から前記舟体の形状を特徴付ける部分を切取り所定の比率で縮小して保存する作業を、前記舟体に照射される太陽光や照明の強度と方向など明暗環境の異なる条件下において、異なる個体の複数の舟体を用いて複数回実施し、第1〜第nまでの複数枚の画像データによるテンプレートを取得し、次に検査対象となるパンタグラフの舟体を前記カメラで撮影して得られた画像を取得サイズのまま画像Aとして保存するとともに、前記画像Aを所定の比率で縮小して得られた画像Bに対して、画像B上に前記テンプレートを配置する位置を変えながら画像Bとテンプレートとの相関値Qを計算し、この相関値Qの最大値Qmとそのときのテンプレート位置P(x,y)を、第1〜第nのテンプレートに対してそれぞれ求め、前記各テンプレートごとの最大値Qmが最も大きい値を示したテンプレートの位置に対応する前記画像A上の点からx,y方向にそれぞれ所定寸法だけ離れた位置を基準として所定の大きさをもった範囲を設定し、前記舟体上のすり板下面と舟体上面の境界線を検出する画像処理を前記設定範囲内で行ない、前記舟体に搭載された前記すり板の側面の厚さを、検出した前記境界線から前記すり板の上面までの距離として、横(長)方向各位置について求めることを特徴とするパンタグラフすり板の厚み測定方法。  Photographing the pantograph hull mounted on the passing train roof with a camera from the front or the rear, cutting the portion characterizing the shape of the hull from the obtained image and reducing and storing it at a predetermined ratio, A plurality of images from the first to the nth are carried out a plurality of times using a plurality of hulls of different individuals under different light and dark environments such as sunlight and intensity and direction of illumination irradiated to the hull. A template based on data is acquired, and then an image obtained by photographing the pantograph hull to be inspected with the camera is stored as the image A in the acquired size, and the image A is reduced at a predetermined ratio. The correlation value Q between the image B and the template is calculated while changing the position where the template is placed on the image B, and the maximum value Qm of the correlation value Q and Template positions P (x, y) are obtained for the first to nth templates, respectively, on the image A corresponding to the position of the template that shows the largest maximum value Qm for each template. Image processing for setting a range having a predetermined size on the basis of positions separated from each point by a predetermined dimension in the x and y directions and detecting a boundary line between the bottom surface of the sliding board on the boat body and the upper surface of the boat body Performing within the set range, and obtaining the thickness of the side surface of the sliding plate mounted on the hull as the distance from the detected boundary line to the upper surface of the sliding plate for each position in the lateral (long) direction A method for measuring the thickness of a pantograph slip plate characterized by the following. 通過する電車屋根上に搭載されたパンタグラフの舟体を前方あるいは後方からカメラにより撮影し、得られた画像から前記舟体の形状を特徴付ける部分を切取り所定の比率で縮小して保存する作業を、前記舟体に照射される太陽光や照明の強度と方向など明暗環境の異なる条件下において、異なる個体の複数の舟体を用いて複数回実施し、第1〜第nまでの複数枚の画像データによるテンプレートを取得し、次に検査対象となるパンタグラフの舟体を前記カメラで撮影して得られた画像を取得サイズのまま画像Aとして保存するとともに、前記画像Aを所定の比率で縮小して得られた画像Bに対して、この画像B上に前記テンプレートを配置する位置を変えながら画像Bとテンプレートとの相関値Qを計算し、相関値Qの最大値Qmとそのときのテンプレート位置P(x,y)を、前記第1〜第nのテンプレートに対してそれぞれ求め、前記各テンプレートごとの最大値Qmが最も大きい値を示したテンプレートの位置に対応する前記画像A上の点からx,y方向にそれぞれ所定寸法だけ離れた位置を基準として所定の大きさをもった範囲を設定し、前記舟体の下端を検出する画像処理を前記設定範囲内で行なうことを特徴とする請求項1に記載のパンタグラフすり板の厚み測定方法。  Shooting the pantograph hull mounted on the passing train roof with a camera from the front or the rear, cutting the portion characterizing the shape of the hull from the obtained image and saving it by a predetermined ratio, A plurality of images from the first to the nth are carried out a plurality of times using a plurality of hulls of different individuals under different light and dark environments such as sunlight and intensity and direction of illumination irradiated to the hull. A template based on data is acquired, and then an image obtained by photographing the pantograph hull to be inspected with the camera is stored as the image A in the acquired size, and the image A is reduced at a predetermined ratio. The correlation value Q between the image B and the template is calculated while changing the position where the template is arranged on the image B, and the maximum value Qm of the correlation value Q and Template position P (x, y) is obtained for each of the first to nth templates, and the template position P on the image A corresponding to the position of the template having the largest maximum value Qm for each template. A range having a predetermined size is set on the basis of positions separated from each other by a predetermined dimension in the x and y directions, and image processing for detecting the lower end of the hull is performed within the set range. The thickness measurement method of the pantograph slip board according to claim 1. 通過する電車屋根上に搭載されたパンタグラフの舟体を前方あるいは後方からカメラにより撮影して得られた画像に対し、前記設定範囲として予め定めたすり板下面境界抽出範囲内の上端から下端までの間、上下方向に走査しながら順次所定の画素間隔で下側画素の輝度から上側画素の輝度を引いた差分をとり、差分値の極点のうち正の最大値を示した箇所(これをA点と称する)を検出し、次にA点から上側所定の範囲内に存在する差分値の正の極大値のうち、所定の値より大きい箇所(これをB点と称する)が前記A点以外に存在するとき、前記B点を前記舟体の上面と前記すり板の下面との境界の候補点として用い、前記条件を満たすB点が見出せなかったときは最初に見出されたA点を前記候補点として用いることを特徴とする請求項1または2に記載のパンタグラフすり板の厚み測定方法。  For the image obtained by photographing the pantograph hull mounted on the passing train roof with the camera from the front or the rear, from the upper end to the lower end in the sliding plate lower surface boundary extraction range predetermined as the setting range During this period, the difference obtained by subtracting the luminance of the upper pixel from the luminance of the lower pixel sequentially at a predetermined pixel interval while scanning in the vertical direction is taken, and the point indicating the maximum positive value among the extreme points of the difference value (this is the point A Next, a point that is larger than a predetermined value (referred to as point B) among positive maximum values of difference values existing within a predetermined range on the upper side from point A other than point A is detected. When present, the point B is used as a candidate point for the boundary between the upper surface of the hull and the lower surface of the sliding plate. When the point B satisfying the condition is not found, the point A first found is Claims used as candidate points Pantograph contact strip thickness measuring method according to 1 or 2. 舟体上面とすり板下面との境界の候補点の位置から舟体下方に所定の範囲内に所定の輝度値より明るい画素が所定の距離以上連続して存在する条件を満たす候補点が、横(長)方向に所定以上連続して存在するとき、前記条件を満たす候補点全てを前記舟体上面と前記すり板下面との境界の候補から除外することを特徴とする請求項1または2に記載のパンタグラフすり板の厚み測定方法。  Candidate points satisfying the condition that pixels brighter than a predetermined brightness value continuously exist within a predetermined range from the position of the candidate point at the boundary between the upper surface of the hull and the lower surface of the sliding board within a predetermined range below the hull are 3. The method according to claim 1, wherein when there is a predetermined length or more in the (long) direction, all candidate points that satisfy the condition are excluded from candidates for a boundary between the upper surface of the hull and the lower surface of the sliding plate. The thickness measurement method of the pantograph slip board of description. 前記設定範囲として予め定めたすり板下面境界抽出範囲を横(長)方向に複数ブロックに分割し、それぞれのブロックごとに抽出された境界候補点を代表する直線をHough変換によって求め、得られたそれぞれの直線を隣接するものどうし交点で連結して得られた折れ線を、舟体上面とすり板下面との境界として用いることを特徴とする請求項1または2に記載のパンタグラフすり板の厚み測定方法。  The above-mentioned setting range is obtained by dividing the bottom surface boundary extraction range predetermined as the set range into a plurality of blocks in the horizontal (long) direction and obtaining straight lines representative of boundary candidate points extracted for each block by Hough transform. 3. A thickness measurement of a pantograph sliding plate according to claim 1 or 2, wherein a broken line obtained by connecting each straight line at an intersection of adjacent ones is used as a boundary between the upper surface of the hull and the lower surface of the sliding plate. Method. 前記設定範囲として予め定めたすり板下面境界抽出範囲を横(長)方向に複数ブロックに分割し、それぞれのブロックごとに抽出された境界候補点を代表する直線をHough変換によって求め、Hough変換時に境界候補点との一致度が所定値以上となった直線のみを用いて、前記一致度が所定値未満の区間が所定値以上の区間にはさまれている場合には両側に隣接する区間の直線どうし交点で連結するものとし、また前記一致度が所定値未満の区間が端の場合は隣接する区間の直線を延長するものとし、得られた折れ線を、舟体上面とすり板下面との境界として用いることを特徴とする請求項1または2に記載のパンタグラフすり板の厚み測定方法。  As the setting range, a sliding plate lower surface boundary extraction range is divided into a plurality of blocks in the horizontal (long) direction, straight lines representing boundary candidate points extracted for each block are obtained by Hough transformation, and at the time of Hough transformation If only a straight line with a matching degree with a boundary candidate point equal to or greater than a predetermined value is used and a section with the matching degree less than the predetermined value is sandwiched by a section with a predetermined value or more, The straight lines of the adjacent sections shall be extended when the section where the degree of coincidence is less than the predetermined value is an end, and the obtained broken line is connected between the upper surface of the boat body and the lower surface of the sliding board. It uses as a boundary, The thickness measuring method of the pantograph slip board of Claim 1 or 2 characterized by the above-mentioned. 前記設定範囲として予め定めたすり板下面境界抽出範囲を舟体横(長)方向に複数ブロックに分け、Hough変換時に境界候補点との一致度が所定値以下となったブロックが舟体横(長)方向での中央付近であるとき、画像処理によって前記舟体中央部の下端を抽出しHough変換によって得られた直線を所定の距離だけ上方向に平行移動したものを前記舟体中央付近のブロック内の直線として採用することを特徴とする請求項1または2に記載のパンタグラフすり板の厚み測定方法。As the set range , a sliding plate lower surface boundary extraction range is divided into a plurality of blocks in the horizontal direction of the hull (longitudinal direction). When it is near the center in the (long) direction, the lower end of the center part of the hull is extracted by image processing, and a straight line obtained by Hough transformation is translated upward by a predetermined distance. The method for measuring the thickness of a pantograph slide plate according to claim 1, wherein the method is adopted as a straight line in a block. 前記設定範囲として予め定めたすり板下面境界抽出範囲を舟体横(長)方向に複数ブロックに分け、Hough変換時に境界候補点との一致度が所定値以下となったブロックが舟体横(長)方向での中央付近であるとき、画像処理によって前記舟体中央部の下端を抽出してその平均位置を求め、その位置から所定の距離だけ上方向に移動した点Aを定め、隣接するブロック内の直線端から引いた前記点Aを通る直線を前記舟体中央付近のブロック内の直線として採用することを特徴とする請求項1または2に記載のパンタグラフすり板の厚み測定方法。As the set range , a sliding plate lower surface boundary extraction range is divided into a plurality of blocks in the horizontal direction (long) of the hull, and a block whose degree of coincidence with a boundary candidate point becomes a predetermined value or less at the time of Hough conversion is When it is near the center in the (long) direction, the lower end of the central part of the hull is extracted by image processing to obtain the average position, and a point A moved upward by a predetermined distance from that position is determined and adjacent The method for measuring the thickness of a pantograph slide according to claim 1 or 2, wherein a straight line passing through the point A drawn from a straight line end in the block is adopted as a straight line in the block near the center of the hull. 前記舟体に搭載されたすり板と舟体の境界部分から舟体下端までを隠蔽する長方形状の遮蔽板が存在するために、前記設定範囲として予め定めたすり板下面境界抽出範囲内で前記境界線が検出できない場合には、通過する電車屋根上に搭載されたパンタグラフの舟体を前方あるいは後方からカメラにより撮影して得られた画像に対し、前記遮蔽板の横方向の1辺のエッジを画像処理によって抽出しHough変換によって得られた直線を所定の距離だけ平行移動したものを舟体上面とすり板下面との境界として用いることを特徴とする請求項1または2に記載のパンタグラフすり板の厚み測定方法。Since there is a rectangular shielding plate that conceals from the boundary between the sliding board mounted on the hull and the hull to the lower end of the hull, the above-mentioned setting range within the sliding board lower surface boundary extraction range If the boundary line cannot be detected, the edge of one side in the horizontal direction of the shielding plate with respect to the image obtained by photographing the pantograph boat mounted on the passing train roof from the front or rear with the camera A pantograph slide according to claim 1 or 2, wherein a straight line obtained by Hough transformation is translated by a predetermined distance and used as a boundary between the upper surface of the boat body and the lower surface of the sliding plate. A method for measuring the thickness of a plate. 前記舟体に搭載されたすり板と舟体の境界部分から舟体下端までを隠蔽する長方形状の遮蔽板が存在するために、前記設定範囲として予め定めたすり板下面境界抽出範囲内で前記境界線が検出できない場合には、通過する電車屋根上に搭載されたパンタグラフの舟体を前方あるいは後方からカメラにより撮影して得られた画像に対し、前記遮蔽板の横方向の1辺のエッジを画像処理によって抽出しHough変換によって直線を求め、次に前記遮蔽板に隠蔽されていない部分の舟体上面とすり板下面との境界を画像処理によって抽出してその平均位置を求め、その位置を通り前記直線に平行な直線を舟体上面とすり板下面との境界として用いることを特徴とする請求項1または2に記載のパンタグラフすり板の厚み測定方法。Since there is a rectangular shielding plate that conceals from the boundary between the sliding board mounted on the hull and the hull to the lower end of the hull, the above-mentioned setting range within the sliding board lower surface boundary extraction range If the boundary line cannot be detected, the edge of one side in the horizontal direction of the shielding plate with respect to the image obtained by photographing the pantograph boat mounted on the passing train roof from the front or rear with the camera Is extracted by image processing, a straight line is obtained by Hough transform, and then the boundary between the upper surface of the hull and the lower surface of the sliding plate that is not concealed by the shielding plate is extracted by image processing to obtain the average position, 3. A method for measuring the thickness of a pantograph slip plate according to claim 1 or 2, wherein a straight line passing through the straight line and parallel to the straight line is used as a boundary between the upper surface of the boat body and the lower surface of the slide plate. 前記舟体に搭載されたすり板と舟体の境界部分から舟体下端までを隠蔽する長方形状の遮蔽板が存在するために、前記設定範囲として予め定めたすり板下面境界抽出範囲内で前記境界線が検出できない場合には、通過する電車屋根上に搭載されたパンタグラフの舟体を前方あるいは後方からカメラにより撮影して得られた画像に対し、前記遮蔽板の横方向の1辺のエッジを画像処理によって抽出して平均位置Aを求め、次に前記遮蔽板に隠蔽されていない部分の舟体上面とすり板下面との境界を画像処理によって抽出してその平均位置Bを求め、前記平均位置Aを所定量上下方向に移動した点と前記平均位置Bを通る直線を舟体上面とすり板下面との境界として用いることを特徴とする請求項1または2に記載のパンタグラフすり板の厚み測定方法。Since there is a rectangular shielding plate that conceals from the boundary between the sliding board mounted on the hull and the hull to the lower end of the hull, the above-mentioned setting range within the sliding board lower surface boundary extraction range If the boundary line cannot be detected, the edge of one side in the horizontal direction of the shielding plate with respect to the image obtained by photographing the pantograph boat mounted on the passing train roof from the front or rear with a camera Is extracted by image processing to determine the average position A, and then the boundary between the upper surface of the hull and the bottom surface of the sliding plate that is not concealed by the shielding plate is extracted by image processing to determine the average position B, 3. The pantograph slide board according to claim 1, wherein a straight line passing through a point where the average position A is moved up and down by a predetermined amount and the average position B is used as a boundary between the upper surface of the boat body and the lower surface of the slide board. Thickness measurement Method.
JP35901098A 1998-12-17 1998-12-17 Method for measuring the thickness of pantograph strips Expired - Fee Related JP4209525B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP35901098A JP4209525B2 (en) 1998-12-17 1998-12-17 Method for measuring the thickness of pantograph strips

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP35901098A JP4209525B2 (en) 1998-12-17 1998-12-17 Method for measuring the thickness of pantograph strips

Publications (2)

Publication Number Publication Date
JP2000180128A JP2000180128A (en) 2000-06-30
JP4209525B2 true JP4209525B2 (en) 2009-01-14

Family

ID=18462283

Family Applications (1)

Application Number Title Priority Date Filing Date
JP35901098A Expired - Fee Related JP4209525B2 (en) 1998-12-17 1998-12-17 Method for measuring the thickness of pantograph strips

Country Status (1)

Country Link
JP (1) JP4209525B2 (en)

Families Citing this family (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2006118900A (en) * 2004-10-20 2006-05-11 Hitachi Industries Co Ltd Inspection device for pantograph contact strip
WO2009018612A1 (en) * 2007-08-06 2009-02-12 Qr Limited Pantograph damage and wear monitoring system
DE102009006392A1 (en) * 2009-01-28 2010-07-29 Siemens Aktiengesellschaft Monitoring arrangement and method for monitoring the state of the sliding contact devices of electric locomotives
JP5629478B2 (en) * 2010-03-03 2014-11-19 東海旅客鉄道株式会社 Pantograph monitoring device
JP2012015698A (en) * 2010-06-30 2012-01-19 Meidensha Corp Pantograph detection device and method thereof
TWI475187B (en) 2010-10-27 2015-03-01 Hitachi High Tech Corp Image processing devices and computer programs
CN107588732B (en) * 2016-07-07 2024-03-26 苏州华兴致远电子科技有限公司 Rail side train part height measurement method and system
JP6644720B2 (en) * 2017-01-20 2020-02-12 公益財団法人鉄道総合技術研究所 Train wire fitting detection system and its detection method
CN107554303B (en) * 2017-09-13 2023-11-03 大同新成新材料股份有限公司 Device for monitoring carbon sliding plate loss of pantograph in real time
CN111127381B (en) * 2018-10-15 2022-08-16 西南交通大学 Non-parallel detection method for pantograph slide plate
CN113048893B (en) * 2019-12-27 2023-01-10 苏州华兴致远电子科技有限公司 Train brake pad abrasion measuring method and system
CN117333526A (en) * 2023-09-06 2024-01-02 重庆中车四方所智能装备技术有限公司 Method for measuring thickness of copper-based slider of pantograph of monorail car, storage medium and system

Also Published As

Publication number Publication date
JP2000180128A (en) 2000-06-30

Similar Documents

Publication Publication Date Title
JP4209525B2 (en) Method for measuring the thickness of pantograph strips
JP5175528B2 (en) Tunnel lining crack inspection system
USRE37610E1 (en) Running guide apparatus for vehicle capable of keeping safety at passing through narrow path and the method thereof
JP6620477B2 (en) Method and program for detecting cracks in concrete
JP3167752B2 (en) Vehicle distance detection device
CN109241976B (en) Method for estimating oil spilling area based on image processing and laser ranging
Li et al. Nighttime lane markings recognition based on Canny detection and Hough transform
GB2265779A (en) Obstacle warning system for vehicle
CN107798293A (en) A kind of crack on road detection means
CN108725511B (en) Real-time position correction method for rail corrugation measuring point
CN106996748A (en) A kind of wheel footpath measuring method based on binocular vision
JP4032843B2 (en) Monitoring system and monitoring method, distance correction device and distance correction method in the monitoring system
JPH11242745A (en) Method for measuring and processing facial image
WO2010061779A1 (en) Device for measuring trolley wire wear and deviation by image processing
JP3710548B2 (en) Vehicle detection device
CN115147566A (en) Modeling method, system and device of tunnel lining three-dimensional structure based on multi-line laser
JP4586571B2 (en) Object judgment device
CN112949482B (en) Non-contact type rail sleeper relative displacement real-time measurement method based on deep learning and visual positioning
CN113160132A (en) Detection processing method and system for weld defect image
JPH05157558A (en) Vehicular gap detector
EP1398060A1 (en) Stroke information measuring apparatus and stroke information measuring method
JP2001043353A (en) Device for creation of developed image of tunnel wall surface
JP3662218B2 (en) Lane boundary detection device
JP2001108434A (en) Method and apparatus for measuring distance
JPH02242382A (en) Defect checking method

Legal Events

Date Code Title Description
A621 Written request for application examination

Free format text: JAPANESE INTERMEDIATE CODE: A621

Effective date: 20051128

A977 Report on retrieval

Free format text: JAPANESE INTERMEDIATE CODE: A971007

Effective date: 20071101

A131 Notification of reasons for refusal

Free format text: JAPANESE INTERMEDIATE CODE: A131

Effective date: 20071113

A521 Written amendment

Free format text: JAPANESE INTERMEDIATE CODE: A523

Effective date: 20080111

A131 Notification of reasons for refusal

Free format text: JAPANESE INTERMEDIATE CODE: A131

Effective date: 20080205

A521 Written amendment

Free format text: JAPANESE INTERMEDIATE CODE: A523

Effective date: 20080404

A131 Notification of reasons for refusal

Free format text: JAPANESE INTERMEDIATE CODE: A131

Effective date: 20080507

A521 Written amendment

Free format text: JAPANESE INTERMEDIATE CODE: A523

Effective date: 20080704

A02 Decision of refusal

Free format text: JAPANESE INTERMEDIATE CODE: A02

Effective date: 20080729

A521 Written amendment

Free format text: JAPANESE INTERMEDIATE CODE: A523

Effective date: 20080922

A521 Written amendment

Free format text: JAPANESE INTERMEDIATE CODE: A821

Effective date: 20080828

A911 Transfer of reconsideration by examiner before appeal (zenchi)

Free format text: JAPANESE INTERMEDIATE CODE: A911

Effective date: 20081003

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: 20081021

A01 Written decision to grant a patent or to grant a registration (utility model)

Free format text: JAPANESE INTERMEDIATE CODE: A01

A61 First payment of annual fees (during grant procedure)

Free format text: JAPANESE INTERMEDIATE CODE: A61

Effective date: 20081023

R150 Certificate of patent or registration of utility model

Free format text: JAPANESE INTERMEDIATE CODE: R150

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

Free format text: PAYMENT UNTIL: 20111031

Year of fee payment: 3

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

Free format text: PAYMENT UNTIL: 20121031

Year of fee payment: 4

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

Free format text: PAYMENT UNTIL: 20131031

Year of fee payment: 5

R250 Receipt of annual fees

Free format text: JAPANESE INTERMEDIATE CODE: R250

R250 Receipt of annual fees

Free format text: JAPANESE INTERMEDIATE CODE: R250

R250 Receipt of annual fees

Free format text: JAPANESE INTERMEDIATE CODE: R250

LAPS Cancellation because of no payment of annual fees