1306814 九、發明說明： 【發明所屬之技術領域】 本發明一般係關於基於影像處理之架空接觸電線的耗 損測量，特別是關於架空接觸電線的耗損部份之寬度的測 '量。 【先前技術】 用於供應電力至一電氣鐵路載具的一架空接觸電線在 每次該載具通過時便會接觸於一電流收集器。因此，當電 φ 氣鐵路載具的操作重複進行時，該接觸電線便會逐漸耗 損。除非該耗損架空接觸電線被一新的所替換，否則有可 能發生的是該架空接觸電線最後會斷裂並引發一場意外。 傳統上，一接觸電線皆具有一耗損極限，而該接觸電線係 根據該耗損極限而被替換，以確保電氣鐵路載具的安全性。 測量架空接觸電線耗損的方法被粗略地分爲兩類，一 類是直接測量一架空接觸電線的厚度，另一類則是測量一 架空接觸電線的一耗損部份之寬度（以下稱其爲架空接觸 Φ 電線耗損部份寬度），並將該架空接觸電線耗損部份寬度轉 換爲一架空接觸電線的厚度。 直接測量一架空接觸電線之厚度的一種方法係使用諸 如一滑動卡尺（slide caliper)的一量尺以測量一接觸電線的 厚度。 . 直接測量一架空接觸電線之厚度的另一種方法係使用 一光學偵測器。具體來說，一測量裝置包括夾持一接觸電 線的上、下滾軸、附屬於該上、下滾軸用於發射一雷射光 束的一雷射光束發射裝置、以及用於接收該雷射光束並測 1306814 量所接收之雷射光束的量的一雷射光束接收裝置，位於該 夾持部份之所接收雷射光束的量會被轉換爲一架空接觸電 線的厚度。利用這種方法，便可以自動地且連續地測量出 ' 一架空接觸電線的厚度。 '測量該架空接觸電線耗損部份寬度的一種方法係利用 一鈉燈或是一雷射光束以照射一架空接觸電線，這種方法 所基於的事實在於一耗損架空接觸電線之橫截面的形式係 爲具有一直的底部的一圓，並且當該架空接觸電線逐漸耗 損時，一耗損部份的寬度會增加。這個架空接觸電線耗損 部份會被轉換爲該架空接觸電線的厚度。測量該架空接觸 電線耗損部份寬度的這個方法是精確地調整一所接收部份 的一線偵測器之位置、利用鏡面反射以接收由來自諸如一 鈉燈以及一雷射光束的一光源之該架空接觸電線的一光線 之反射光線、利用鏡面反射藉由接收一強烈光線而改變該 接觸耗損部份使其成爲一所謂的白矇狀態、以及根據接收 該強烈光線之該白朦部份的寬度而測量該架空接觸電線耗 損部份寬度。由於這個方法係爲非接觸型態，因此能夠達 •成高速測量。 【發明內容】 在上述傳統的方法中具有一些問題如下。 在使用諸如一滑動卡尺的一量尺以直接測量一架空接 觸電線之厚度的方法中，係直接地且可靠地測量出一架空 接觸電線的一所需部份。然而，由於此方法無法自動地執 行，因此難以測量一架空接觸電線的一長區域。 在使用一光學偵測器以直接測量一架空接觸電線之厚 1306814 度的方法中，由於該測量裝置係接觸於一架空接觸電線， 因此需要在低速度下執行測量。此外，由於該測量裝置夾 持一架空接觸電線，因此並沒有辦法將此一測量裝置使用 ' 於該測量裝置妨礙到諸如點、空氣區域、以及錨狀物等已 ' 存在之結構的情形。是故，在所述已存在之結構的地方， 有需要暫時地置放該測量裝置於遠離進行測量之位置，藉 以防止該測量裝置與該已存在結構之間的衝突。 在利用一鈉燈或一雷射光束照射一架空接觸電線以測 量該架空接觸電線耗損部份的方法中，有需要使用諸如一 鈉燈的特別光線於一光源，並且有需要將特別光線對於人 體的影響納入考慮。此外，這個方法很容易受到雜訊的影 響；諸如由夾持一架空接觸電線的一鉗子或是出現在一所 偵測影像之背景中的已存在結構所引發的雜訊。在雜訊的 影響之下當獲得一不正確測量結果時，此時的該影像便不 會被記錄，並且因此沒有方法能夠針對其進行檢查。是故， 該不正確測量結果的該部份需要藉由使用直接測量一架空 接觸電線之厚度的該方法而被檢查。亦有需要利用鏡面反 β 射接收該架空接觸電線之該耗損部份所賴以照射之一光線 的該反射光線，並且因此有需要精確地針對一光源及一透 光裝置進行定位。 因此，本發明之一目的係提供一架空接觸電線耗損測 量設備，用以配置來藉由對正確地具有被提升之工作效率 的一架空接觸電線的一耗損部份之寬度進行影像處理而進 行測量。 根據本發明的一觀點，一種架空接觸電線耗損測量設 1306814 備，包括：一線偵測器，用以掃瞄一架空接觸電線；以及 一資料處理單元，連接於該線偵測器，該資料處理單元包 括：一線偵測器影像產生部，結合來自該線偵測器資料以 產生一線偵測器影像；一影像二値化處理部，二値化 (binarize)該線偵測器影像以產生一二値化線偵測器影像； 一雜訊消除處理部，自該二値化線偵測器影像中消除雜訊 以產生一雜訊已消除二値化線偵測器影像；一架空接觸電 線耗損部份邊緣偵測部，偵測該雜訊已消除二値化線偵測 器影像中該架空接觸電線之一耗損部份的邊緣；以及一架 空接觸電線耗損部份寬度計算部，根據所偵測之該接觸電 線之該耗損部份的邊緣以及該架空接觸電線的高度，來計 算該架空接觸電線之該耗損部份的寬度。該架空接觸電線 耗損部份寬度計算部係用以執行下列步驟：於該線偵測器 影像中的一掃瞄線上選擇一邊緣點作爲一參考邊緣點；於 該參考邊緣點之鄰接的一預定範圍內決定該等邊緣點的一 近似直線；決定包括該參考邊緣點並垂直於該近似直線的 一第二直線；決定一邊緣交叉點，該第二直線係與相對該 參考邊緣點之該架空接觸電線的該耗損部份的另一邊緣交 叉於該邊緣交叉點；以及根據該參考邊緣點與該邊緣交叉 點之間的距離和該架空接觸電線的高度，來計算該架空接 觸電線之該耗損部份的寬度。該架空接觸電線耗損部份寬 度計算部係用以執行下列步驟：於該線偵測器影像中的一 掃瞄線上的一第一邊緣點之鄰接的一預定範圍內決定一第 一組邊緣點的一第一近似直線；於該掃瞄線上的一第二邊 緣點之鄰接的一預定範圍內決定一第二組邊緣點的一第二 Ί306814 近似直線；決定與該第一組邊緣點相關的該第一近似直線 的一誤差程度；決定與該第二組邊緣點相關的該第二近似 直線的一誤差程度；於該第一邊緣點及該第二邊緣點選擇 ' 該相關誤差程度較小的其中之一作爲一參考邊緣點；決定 ^ 包括該參考邊緣點並垂直於該相關近似直線的一第三直線； 決定一邊緣交叉點，該第三直線係與相對該參考邊緣點之 該架空接觸電線的該耗損部份的另一邊緣交叉於該邊緣交 叉點；以及根據該參考邊緣點與該邊緣交叉點之間的距離 和該架空接觸電線的高度，來計算該架空接觸電線之該耗 ® 損部份的寬度。該架空接觸電線耗損測量設備更包括：一 第二線偵測器，用以掃瞄一導電弓（pantograph)的一頂端； 以及一'第二·資料處理單元，連接於該第二線偵測器，該第 二資料處理單元包括：一線偵測器影像產生部，由該第二 線偵測器結合資料以產生一線偵測器影像；以及一導電弓 頂端高度計算部，根據該線偵測器影像計算該架空接觸電 線的高度。 根據本發明的另一觀點，提供一種架空接觸電線耗損 ® 檢查車輛，包括：一裝有車輪之本體；一線偵測器，裝設 於該本體的一車頂上並指向上方；以及一資料處理單元， 連接於該線偵測器，該資料處理單元包括：一線偵測器影 像產生部，由該線偵測器結合資料以產生一線偵測器影像； 一影像二値化處理部，二値化該線偵測器影像以產生一二 値化線偵測器影像；一雜訊消除處理部，自該二値化線偵 測器影像中消除雜訊以產生一雜訊已消除二値化線偵測器 影像；一架空接觸電線耗損部份邊緣偵測部，偵測該雜訊 Ί306814 消除二値化線偵測器影像中該架空接觸電線之一耗: 的邊緣；以及一架空接觸電線耗損部份寬度計算部 所偵測之該架空接觸電線之該耗損部份的邊緣以及 ' 接觸電線的高度，來計算該架空接觸電線之該耗損 - 寬度。 根據本發明的再一觀點，提供一種架空接觸電 測量方法，包括下列步驟：由一線偵測器結合資料 一線偵測器影像；二値化該線偵測器影像以產生一 線偵測器影像；自該二値化線偵測器影像中消除雜 ® 生一雜訊消除二値化線偵測器影像；偵測該雜訊消 化線偵測器影像中一接觸電線之一耗損部份的邊緣 根據所偵測之該架空接觸電線之該耗損部份的邊緣 架空接觸電線的高度，來計算該架空接觸電線之該 份的寬度。 【實施方式】 請參閱第1〜5圖、第10圖及第1 1圖，其中顯 明第一實施例之架空接觸電線耗損測量設備。第1 ® 發明第一實施例之架空接觸電線耗損測量設備的透 如第1圖所示，一線偵測器1 1、一線偵測器12、一 13、以及一導電弓14係裝設於具有一車輪本體之一 輛1 0的頂部。照射燈1 3係爲一普通白光源，用以 白光以照射一架空接觸電線1 6。一第一資料處理單 連接係用於通往線偵測器1 1的信號通訊。一第二資 單元2的連接係用於通往線偵測器1 2的信號通訊。 料處理單元1及第二資料處理單元2係彼此相連接 員部份 ，根據 該架空 部份的 線耗損 以產生 二値化 訊以產 除二値 ；以及 以及該 耗損部 示本發 圖爲本 視圖。 照射燈 檢查車 發射一 元1的 料處理 第一資 -10- 1306814 線偵測器11係裝設於檢查車輛1 〇的頂部之上，方向 係朝上垂直於檢查車輛1 0之頂部的表面。線偵測器u的 配置方式爲：掃瞄方向18係與枕木的方向相同、以及掃瞄 線與架空接觸電線1 6相交。另一方面，線偵測器1 2係裝 設於檢査車輛10的頂部之上，方向係朝上而傾斜於檢查車 輛10之頂部的表面。線偵測器12的配置方式爲：掃瞄方 向19係沿著檢查車輛1〇的垂直方向而延伸、以及掃猫線 相交於導電弓1 4。 第5圖爲第一實施例之架空接觸電線耗損測量設備的 示意系統方塊圖。第一資料處理單元1包括元件50〜57，如 第5圖所示，線偵測器11由每一掃瞄線獲得一線偵測器影 像信號F50並發送線偵測器影像信號F50至一線偵測器影 像產生部50。線偵測器影像產生部50裝設成依照時序配設 線偵測器影像信號F50以產生一線偵測器影像F5 1，配設 成線偵測器影像F5 1將儲存於一記憶體5 1。儲存於記憶體 5 1中的線偵測器影像F5 1係經由一資料傳輸線5 7被傳遞並 被儲存於一記憶體52。 一影像二値化處理部53之配設成將一影像二値化處 理施加於儲存在記憶體5 2的線偵測器影像F5 1以產生一二 値化線偵測器影像F52，及配設成二値化線偵測器影像F5 2 儲存於記憶體5 2。 一雜訊消除處理部54之配設成將一雜訊消除處理施 加於儲存在記憶體52的二値化線偵測器影像F52以產生一 雜訊消除二値化線偵測器影像F5 3，及配設成將雜訊消除 二値化線偵測器影像F53儲存於記憶體52。 1306814 一架空接觸電線耗損部邊緣偵測部55之裝設成將一 架空接觸電線耗損部邊緣偵測處理施加於儲存在記憶體5 2 的雜訊消除二値化線偵測器影像F 5 3以產生架空接觸電線 耗損部邊緣資料F54，及配設成將架空接觸電線耗損部邊 ' 緣資料F 5 4儲存於記憶體5 2。 一架空接觸電線耗損部份寬度計算部5 6配設成將一 架空接觸電線耗損部份寬度計算處理施加於儲存在記憶體 52的架空接觸電線耗損部邊緣資料Ρ54以及架空接觸電線 0 高度資料F55，以產生架空接觸電線耗損部份寬度資料 F5 8,及配設成將架空接觸電線耗損部份寬度資料F58儲存 於記憶體52。導電弓頂端高度資料F116如下所述被使用作 爲架空接觸電線高度資料F5 5。儲存於一記憶體112中的導 電弓頂端高度資料F1 16經由資料傳輸線57以及連接於資 料傳輸線5 7的一資料傳輸線1 1 8被傳輸且被儲存於記億體 5 2，如第1 1圖所示。 以下參考第4圖之流程圖以說明第一實施例之架空接 觸電線1 6的耗損部份之寬度的測量處理。如第4圖所示， 首先，在步驟S 1 1中，每一掃瞄線（沿X軸）的線偵測器影 像信號F50爲線偵測器1 1所獲得並被依照時序（沿t軸）配 置而在線偵測器影像產生部50中產生線偵測器影像F5U二 維影像），並且線偵測器影像F5 1被儲存於記憶體5 1作爲 —輸入影像。 繼步驟S1 1之後，在步驟S12中，儲存於記憶體51中 的線偵測器影像信號F5 0係經由資料傳輸線5 7而被傳遞並 被儲存於記憶體52。接著，爲了二値化設定一門檻値，及 Ί306814 出現在線偵測器影像F5 1中成一條帶狀及其他背景部份的 該架空接觸電線耗損部份的一影像，經由基於用以在影像 二値化處理部53中產生二値化線偵測器影像F52之該門檻 * 値的該影像二値化處理’而被彼此分離。由於接觸電線16 - 的該耗損部份係藉由被導電弓1 4耗損而形成，因此架空接 觸電線16的該耗損部份相較於其他未耗損部份具有一較 強健的總體。是故，在線偵測器影像F 5 1中，架空接觸電 線16的該耗損部份顯露成爲不同於該背景部份的一發光 値之一帶狀部份。該影像處理產生二値化線偵測器影像 ® F52，如第2圖所示。在二値化線偵測器影像26中，架空 接觸電線1 6的該耗損部份係表示爲一白區域20，且該背景 部份係表示爲一黑區域21。 繼步驟S12之後，在步驟S13中，雜訊消除處理部54 藉由執行一膨脹處理以及一收縮處理而移除來自由二値化 線偵測器影像F5 1所產生之二値化線偵測器影像F5 2的雜 訊。這個處理稱爲雜訊消除處理。這個處理係用於自二値 化線偵測器影像F5 2上移除掉根據架空接觸電線1 6的該耗 ® 損部份及該背景部份的狀況所引起之雜訊的散佈點。 繼步驟S 1 3之後，在步驟S 1 4中，接觸電線耗損部邊 緣偵測部55會偵測在雜訊消除二値化線偵測器影像f5 3中 標示爲白區域20之架空接觸電線16的左及右邊緣22。這 個處理稱爲架空接觸電線耗損部份邊緣偵測處理。當在雜 訊消除二値化線偵測器影像F5 3中沿著一掃瞄線23搜尋 時’會偵測到一個點並將其當作一左邊緣24(上邊緣），此 處顏色係由標示爲該背景部份的黑區域2 1改變爲標示爲 -13- Ί306814 該耗損部份的白區域20;而當在雜訊消除二値化線偵測器 影像F53中沿著掃瞄線23捜尋時，會偵測到另一個點並將 其當作一右邊緣25(下邊緣），此處顏色係由標示爲該耗損 ' 部份的白區域20改變爲標示爲該背景部份的黑區域21。這 - 個處理會在每條掃瞄線由上到下進行，藉以在雜訊消除二 値化線偵測器影像F5 3中偵測架空接觸電線1 6之該耗損部 的邊緣3 0。 繼步驟S14之後，在步驟S15中，基於雜訊消除二値 化線偵測器影像F53所產生接觸電線16之該耗損部份的邊 I 緣的該等邊緣的架空接觸電線耗損部份邊緣資料F54，計 算線偵測器1 1之每一掃瞄線的左與右邊緣3 0之間邊緣至 邊緣距離31當做影像上的一距離。這個處理稱爲架空接觸 電線耗損部份寬度計算處理。基於從線偵測器n到架空接 觸電線16的高度、一透鏡焦點距離、一偵測器掃瞄寬度、 以及偵測器畫素的數目’可以計算出定義爲影像上每一畫 素[pix]的一實際尺寸[mm]之該影像解析度[mm/pix]。藉由 加倍該架空接觸電線耗損部份之影像上的寬度以及該影像 B解析度’便可以計算出該架空接觸電線耗損部份的寬度。 該計算之邊緣資料、該計算之架空接觸電線耗損部份寬 度、指出用於計算該邊緣資料之該線偵測器影像的該資 料、以及確認所使用之該對應掃瞄線的該資料，會被記錄 在一記錄區域（未顯示）之上。藉由測量導電弓頂端丨5之高 度作爲線偵測器1 1至架空接觸電線1 6之高度，計算使用 於上述計舁之線偵測器11至架空接觸電線16的高度F55。 導電弓頂端15之高度係經由資料傳輸線57而被傳輸並被 1306814 儲存於記憶體5 2。 請參閱第11圖，以下說明用於測量導電弓ι4之高度 的該架空接觸電線耗損測量設備的一配置。線偵測器1 2係 裝設於檢查車輛10之頂部上的導電弓14之前側上，藉以 測量導電弓14的高度’利用此方式該掃瞄方向便沿著導電 弓14的垂直方向而延伸。第二資料處理2包括元件11〇至 1 1 8。如第1 1圖所示’線偵測器1 2取得用於每—掃瞄線的 —線偵測器影像信號F 1 1 〇並發送線偵測器影像信號F〗i 〇 到一線偵測器影像產生部1 1 〇。線偵測器影像產生部1工〇 之裝設係用於依照時序處理線偵測器影像信號F丨丨〇以產 生一線偵測器影像F 1 1 1，其裝設還用於儲存線偵測器影像 F 1 1 1於一記憶體1 1 1。儲存於記憶體η1的線偵測器影像 F 1 1 1係經由一資料傳輸線丨18而被傳遞並被儲存於—記憶 體 1 1 2。 記憶體112裝設成儲存並維持用於一該架空接觸電線 之一連串區域的一連串線偵測器影像，其可用以在架空接 觸電線耗損測量之操作後，針對由具有一問題的一架空接 觸電線耗損部份所確認的該部份進行檢查。 一影像二値化處理部1 1 3裝設成施加一影像二値化處 理至儲存在記憶體1 1 2的線偵測器影像F 1 1 1，以產生一二 値化線偵測器影像F 1 1 2，及裝設成儲存二値化線偵測器影 像F 1 1 2於記憶體1 1 2。 一導電弓寬度濾波處理部114裝設成施加一導電弓寬 度濾波處理至儲存在記憶體1 1 2的二値化線偵測器影像 F112，以產生一導電弓所在影像Fii3，及裝設成儲存導電 1306814 弓所在影像F 1 1 3於記憶體1 1 2。 一導電弓所在軌跡濾择處理部115裝設成施加一導電 弓所在軌跡濾波處理至儲存在記憶體1 1 2的導電弓所在影 ' 像F113，以產生一經導電弓一所在一軌跡濾波處理的導電 . 弓所在影像F 1 1 4，及裝設成儲存經導電弓一所在一軌跡濾 波處理的導電弓所在影像F 1 1 4於記憶體1 1 2。 一導電弓奇特點濾波處理部1 1 6裝設成施加一導電弓 奇特點濾波處理至儲存在記憶體1 1 2的導電弓所在軌跡濾 波處理導電弓所在影像F114，以產生一導電弓奇特點濾波 ^ 處理導電弓所在影像F115，及裝設成儲存導電弓奇特點濾 波處理導電弓所在影像F 1 1 5至記憶體1 1 2。 一導電弓頂端高度計算部117裝設成施加一導電弓頂 端高度計算處理至儲存在記憶體1 1 2的經導電弓一奇特點 一濾波處理的導電弓所在影像F 1 1 5，以產生導電弓頂端高 度資料F116,及裝設成儲存導電弓頂端高度資料F116於記 憶體1 1 2。 以下參考第1 0圖之流程圖以說明第一實施例之導電 •弓14的頂端之高度的測量處理。如第10圖所示’首先’ 在步驟S21中，每一掃瞄線（沿y軸）的線偵測器影像信號 F 1 1 0爲線偵測器1 2所獲得並被依照時序（沿t軸）進行處理 而在線偵測器影像產生部1 10中產生線偵測器影像F1 11 (二 維影像）’及線偵測器影像F 1 1 1被儲存於記憶體1 1 1作爲 一輸入影像。 繼步驟S 2 1之後，在步驟S 2 2中，儲存於記憶體1 1 1 中的線偵測器影像F 1 1 〇係經由資料傳輸線11 8而被傳遞並 -16- 1306814 被儲存於記憶體1 1 2。接著，爲了二値化設定一門檻値，及 出現在線偵測器影像F 1 1 1中成一條帶狀及其他背景部份 的導電弓14影像，經由基於用以在影像二値化處理部113 * 中產生二値化線偵測器影像FI 1 2之該門檻値的該影像二 . 値化處理，而被彼此分離。由於導電弓1 4具有不同於該背 景部份的一總體，是故，導電弓14的該影像顯露成爲不同 於線偵測器影像F 1 1 4之該背景部份的一發光値之一帶狀 部份。該影像處理產生二値化線偵測器影像F 1 1 2。在二値 化線偵測器影像F112中，導電弓14的該影像係表示爲一 ® 白區域20，且該背景部份係表示爲一黑區域。 繼步驟S22之後，在步驟S23中，實現該導電弓寬度 濾波處理。特別是，二値化線偵測器影像F 1 1 2的每條掃瞄 線皆在掃瞄方向上被檢查，以偵測一上邊緣及一下邊緣。 當導電弓14的該影像出現成爲具有對應於二値化線偵測 器影像F112中導電弓14之厚度的某一恆定寬度的一帶狀 時，藉由偵測具有與導電弓14之影像上的寬度一樣寬的一 白部份便可以確認出導電弓1 4。當一條掃瞄線中位於該上 ® 邊緣與該下邊緣之間的該間隔係處於一預定範圍的寬度 時，其便可被偵測及被確認爲導電弓1 4。該下邊緣被保持 爲白以指出導電弓1 4的該頂端，而其他部份則被改變爲 黑。藉由重覆上述處理能夠確認出導電弓頂端15的所在， 此處理稱爲導電弓寬度濾波處理。顯示出產生在此處理之 該導電弓頂端位置之所在的該影像稱爲導電弓所在影像 F113。然而，可能的是，當導電弓頂端15的該所在被導電 弓寬度濾波處理所確認時，未對應於導電弓頂端位置1 5的 1306814 部份事實上亦可以會被偵測爲導電弓頂端位置丄5 會發生在諸如架空接觸電線16、絕緣體、以及用 空接觸電線的電線等不是導電弓的物體以具有寬 • 二値化線偵測器影像F 1 1 2中導電弓1 4之寬度的 • 方式而出現的情形下。爲了消除這些殘留爲導電 像F 1 1 3中之雜訊的錯誤地被偵測的部份，須執 驟。 繼步驟S23之後，在步驟S24中，執行該導 軌跡濾波處理。導電弓1 4出現成爲在二値化線偵 ® F1 12中具有一恆定寬度的一帶狀，因此，導電弓 爲由導電弓所在影像F113中導電弓頂端15的一 構成的一線。在此步驟中，相較於一預定門檻具 度之一連串個點被當作雜訊而消除，此處理稱爲 在軌跡濾波處理。然而，該導電弓所在軌跡濾波 消除由短串的點所構成的一雜訊，卻不能消除滿 認該導電弓頂端位置之長度的狀況之雜訊部份。 雜訊消除處理的以下步驟之執行所基於的事實爲 • 1 4出現成爲該線偵測器影像上的單一個，使得導 1 5亦出現成爲線偵測器1 2之該掃瞄線上的一單-使得由該導電弓寬度濾波處理所偵測之大部份的 指出該導電弓所在影像中導電弓頂端15的點。 繼步驟S24之後，在步驟S25中，執行該導 點濾波處理。當兩或多個所在點爲該導電弓所在 偵測器1 2的一掃瞄線所偵測到’最接近該串所在 間位置的該所偵測所在點便會被確認爲—真實者 ，這可以 於升起架 度爲接近 一影像的 弓所在影 行下述步 電弓所在 測器影像 1 4出現成 連串點所 有較短長 導電弓所 處理能夠 足用於確 包括另一 ，導電弓 電弓頂端 -點，以及 點正確地 電弓奇特 影像中線 點中該中 ，消除其 -18- 1306814 他所偵測所在點。這個處理稱爲導電弓奇特點濾波處理。 繼步驟S25之後，在步驟S26中，執行該導電弓頂端 高度計算處理》該雜訊消除導電弓所在影像被轉換以計算 出該導電弓頂端的一真實三維高度，這個轉換包括了基於 線偵測器1 2的該裝設位置及型態的一投影轉換。如第丄！ 圖所75之連接資料傳輸線118以及如第5圖所示之資料傳 輸線57’如上述所決定的該導電弓頂端高度被使用作爲架 空接觸電線耗損部寬度計算部5 6中的架空接觸電線高度 資料F55。在一架空接觸電線之高度爲了維護與檢查而被 預先測量的情形下，此測量値會被使用作爲架空接觸電線 局度資料F55。特別是’記憶體112之裝設成在架空接觸電 線耗損測量之前儲存關於架空接觸電線1 6的高度之資料。 δ靑參閱·第6圖及第8圖，其顯示本發明第二實施例之 —架空接觸電線耗損測量設備。第6圖爲第二實施例之架 空接觸電線耗損測量設備的系統方塊圖。在第6圖所示的 架空接觸電線耗損測量設備，在第一實施例之第4圖的流 程圖所示之步驟S15，架空接觸電線耗損部份寬度計算部 6〇利用一架空接觸電線之傾斜的補償而執行—架空接觸電 線耗損部份寬度計算處理。 在第8圖中，作爲一目標的一掃瞄線87包括—左邊緣 點80。首先’在左邊緣點8〇之附近的一預定範圍81內的 ''緣’..ί被近似而成爲一近似直線8 2 ’在包括左邊緣點8 〇 及垂直近似直線82之一直線83上的一右邊緣點被定義成 爲—邊緣交叉點85。左邊緣點80上的該接觸耗擯部份寬度 被定義成爲左邊緣點80與邊緣交叉點85之間的〜gg離86。 -19- 1306814 請參閱第7圖及第9圖，顯示本發明第三實施例之 架空接觸電線耗損測量設備。第7圖爲第三實施例之架 _觸電線耗損測量設備的系統方塊圖。在第7圖所示的 空接觸電線耗損測量設備’在第一實施例之第4圖的流 ' 圖所示之步驟S15，架空接觸電線耗損部份寬度計算部 利用一架空接觸電線之傾斜的補償而執行另一架空接觸 線耗·損部份寬度計算處理。 在第9圖中，作爲一目標的一掃瞄線1〇2包括一左 φ 緣點90以及一右邊緣點91。首先，在左與右邊緣點90 91之附近92的一預定範圍內的邊緣點被近似而分別成 近似直線93與94。當近似直線93的近似誤差95的總和 於近似直線94的近似誤差96的總和時，一邊緣交叉點 ®被計算成爲在包括左邊緣點90及垂直近似直線之一 f泉98上的一右邊緣點。在掃瞄線1〇2上的該接觸耗損部 寬度被定義成爲左邊緣點90與邊緣交叉點99之間的一 離1 〇〇。另一方面，當近似誤差96的總和小於近似誤差 的總和時，一邊緣交叉點便被計算成爲在包括右邊緣點 W 及垂直近似直線94之一直線101上的一左邊緣點。在掃 線102上的該接觸耗損部份寬度被定義成爲左邊緣交叉 與右邊緣點9 1之間的一距離。 使用所示實施例之影像處理的該架空接觸電線耗損 量設備係應用於用以供應電力至一電氣火車.或是車輛的 架空接觸電線之耗損測量。 以下說明所示實施例之架空接觸電線耗損測量設 產生的優點與效果。 空 架 程 70 電 邊 與 爲 小 99 直 份 距 95 91 猫 點 測 所 -20 - 1306814 (1)由於該架空接觸電線耗損測量設備之操作不需要 直接接觸於架空接觸電線，因此該架空接觸電線耗損測量 設備可執行耗損測量的一高速操作，使得架空接觸電線的 ' 一長距離可以在一短時間內被測量出來。 • (2)由於該架空接觸電線耗損測量設備包括了位置係 遠離諸如點、空氣區域、以及錨狀物等已存在之結構的一 偵測器，因此該架空接觸電線耗損測量設備無須考慮與已 存在之結構的衝突，使得該架空接觸電線耗損測量設備可 應用於具有已存在之結構的地方，與使用一交替滾軸與一 β 光學偵測器以直接測量一架空接觸電線之厚度的方法形成 對比。 (3) 由於該架空接觸電線耗損測量設備在所有所需區 域中產生該線偵測器影像，因此能夠獲得在所有該等所需 區域中之該等架空接觸電線的鄰近區域附近之架空接觸電 線的該影像資料與已存在結構。 (4) 該架空接觸電線耗損測量設備無須使用特殊光 線，諸如用於一光源的一鈉燈。 ® (5)由於該架空接觸電線耗損測量設備不使用雷射 光，因此該架空接觸電線耗損測量設備無須考慮一雷射光 對於人體的影響，以使能夠更輕易地處理該設備，與使用 一雷射光束的方法形成對比。 (6)由於該架空接觸電線耗損測量設備並非利用鏡面 反射而接收架空接觸電線之該耗損部份所照射之一光的該 反射光，因此該架空接觸電線耗損測量設備無須精確地針 對一光源以及一透光裝置進行定位。 -21- 1306814 (7) 由於測量區域中的該影像資料被儲存，因此有可能 利用使用該所儲存之影像而檢查出被確認成爲具有問題的 一架空接觸電線耗損部份。 (8) 基於該第二實施例之該架空接觸電線耗損測量設 *備所產生的一優點在於’當該架空接觸電線相對於一檢查 車輛之移動的方向傾斜時，該傾斜度會爲該架空接觸電線 所補償’這是除了該第一實施例以外所具有的優點。 (9) 基於該第三實施例之該架空接觸電線耗損測量設 備所產生的一優點在於，當該接觸電線相對於一檢査車輛 之移動的方向傾斜時，該傾斜度會被補償以更爲正確地決 定該架空接.觸電線之耗損部份的寬度，這是除了該第一實 施例以外所具有的優點。特別是’即使當該架空接觸電線 的一耗損部份之一邊緣出現波狀耗損或是出現一連串凹口 形狀的耗損’仍有可能更爲正確地決定該架空接觸電線之 耗損部份的寬度j 本案係基於2005年03月11日申請之日本專利申請案 第2005 -68796號’此一日本專利申請案第2005 _6 8796號的 春全份資料於此處隨附作爲參考。 雖然本發明已藉由參考本發明之某些實施例而於上述 進行說明’但本發明並不限定於上述之該等實施例，上述 之該等實施例的調整及變化可爲熟習該項技術者根據上述 教導而完成’本發明之範圍爲下列申請專利範圍所定義。 -22- 1306814 【圖式簡單說明】 第1圖爲本發明第一至第三實施例之架空接觸電線耗 損測量設備的透視圖。 ' 第2圖爲第一至第三實施例之二値化線偵測器影像的 - 示意圖。 第3圖爲第一至第三實施例之邊緣偵測的處理的示意 圖。 第4圖爲第一實施例之架空接觸電線耗損測量設備的 操作處理的流程圖。 B 第5圖爲第一實施例之架空接觸電線耗損測量設備的 系統方塊圖。 第6圖爲第二實施例之架空接觸電線耗損測量設備的 系統方塊圖。 第7圖爲第三實施例之架空接觸電線耗損測量設備的 系統方塊圖。 第8圖爲第二實施例之應用於架空接觸電線的斜度之 補償處理的示意圖。 β 第9圖爲第三實施例之應用於架空接觸電線的斜度之 補償處理的示意圖。 第1 0圖爲測量第一至第三實施例中導電弓之高度的處 理的流程圖。 第11圖爲用於測量第一至第三實施例中導電弓之高度 的架空接觸電線耗損測量設備的系統方塊圖。 -23 - 1306814 【元件符號說明】1306814 IX. Description of the Invention: TECHNICAL FIELD OF THE INVENTION The present invention generally relates to loss measurement of overhead contact wires based on image processing, and more particularly to the measurement of the width of the lossy portion of overhead contact wires. [Prior Art] An empty contact wire for supplying electric power to an electric railway carrier is in contact with a current collector each time the carrier passes. Therefore, when the operation of the electric φ gas railway carrier is repeated, the contact wire is gradually worn out. Unless the worn overhead contact wire is replaced by a new one, it may happen that the overhead contact wire will eventually break and cause an accident. Traditionally, a contact wire has a wear limit that is replaced according to the wear limit to ensure the safety of the electrical railway carrier. The methods for measuring the loss of overhead contact wires are roughly divided into two categories, one is to directly measure the thickness of an overhead contact wire, and the other is to measure the width of a lossy portion of an overhead contact wire (hereinafter referred to as overhead contact Φ). The wire consumes part of the width) and converts the width of the overhead contact wire to the thickness of an empty contact wire. One method of directly measuring the thickness of an overhead contact wire is to measure the thickness of a contact wire using a scale such as a slide caliper. . Another method of directly measuring the thickness of an overhead contact wire is to use an optical detector. Specifically, a measuring device includes an upper and a lower roller that clamps a contact wire, a laser beam emitting device attached to the upper and lower rollers for emitting a laser beam, and a laser beam for receiving the laser beam A laser beam receiving device that measures the amount of the received laser beam by the amount of 1,068,014, the amount of the received laser beam at the clamping portion is converted to the thickness of an overhead contact wire. With this method, the thickness of an overhead contact wire can be measured automatically and continuously. 'One method of measuring the width of the overhead portion of the overhead contact wire is to use a sodium lamp or a laser beam to illuminate an overhead contact wire. This method is based on the fact that the cross-section of a lossy overhead contact wire is in the form of A circle having a constant bottom, and the width of a lossy portion increases as the overhead contact wire is gradually worn out. The lossy portion of the overhead contact wire is converted to the thickness of the overhead contact wire. The method of measuring the width of the lossy portion of the overhead contact wire is to precisely adjust the position of a line detector of a receiving portion, using specular reflection to receive the overhead from a source such as a sodium lamp and a laser beam. The reflected light of a light contacting the electric wire, by specular reflection, changes the contact loss portion by receiving a strong light to make it a so-called white state, and according to the width of the white portion receiving the intense light. Measure the width of the overhead portion of the overhead contact wire. Since this method is a non-contact type, it can achieve high-speed measurement. SUMMARY OF THE INVENTION There are some problems in the above conventional methods as follows. In a method of directly measuring the thickness of an empty contact wire using a measuring rule such as a sliding caliper, a desired portion of an empty contact wire is directly and reliably measured. However, since this method cannot be performed automatically, it is difficult to measure a long area of an overhead contact wire. In the method of using an optical detector to directly measure the thickness of an overhead contact wire of 1306814 degrees, since the measuring device is in contact with an overhead contact wire, it is necessary to perform measurement at a low speed. Furthermore, since the measuring device holds an overhead contact wire, there is no way to use this measuring device to prevent the situation in which the measuring device interferes with a structure such as a point, an air region, and an anchor. Therefore, where the existing structure is present, there is a need to temporarily place the measuring device away from the location where the measurement is taken, thereby preventing a collision between the measuring device and the existing structure. In a method of illuminating an overhead contact wire with a single sodium lamp or a laser beam to measure the lossy portion of the overhead contact wire, it is necessary to use a special light such as a sodium lamp for a light source, and there is a need to influence the special light on the human body. Take it into consideration. In addition, this method is susceptible to noise; such as a pliers that holds an empty contact wire or an existing structure that appears in the background of a detected image. When an incorrect measurement result is obtained under the influence of noise, the image at this time is not recorded, and thus there is no way to check it. Therefore, this portion of the incorrect measurement needs to be inspected by using this method of directly measuring the thickness of an overhead contact wire. There is also a need to use specular anti-β radiation to receive the reflected light of a portion of the overhead contact wire to illuminate a portion of the light, and thus there is a need to accurately position a light source and a light permeable device. Accordingly, it is an object of the present invention to provide an overhead contact wire wear measuring apparatus for configuring to perform image processing by image processing a width of a lossy portion of an overhead contact wire that has the improved working efficiency. . According to one aspect of the present invention, an overhead contact wire loss measurement device 1306814 includes: a line detector for scanning an overhead contact wire; and a data processing unit coupled to the line detector, the data processing The unit includes: a line detector image generating unit that combines data from the line detector to generate a line detector image; and an image binarization processing unit that binarizes the line detector image to generate a a second line detector image; a noise cancellation processing unit that removes noise from the image of the second line detector to generate a noise that has eliminated the image of the line detector; an overhead contact wire A portion of the edge detecting portion that detects that the noise has eliminated the edge of the worn portion of the overhead contact wire in the image of the second line detector; and an empty contact wire loss portion calculating portion The width of the worn portion of the overhead contact wire is calculated by detecting the edge of the worn portion of the contact wire and the height of the overhead contact wire. The overhead contact wire loss portion width calculation unit is configured to perform the following steps: selecting an edge point as a reference edge point on a scan line in the line detector image; and a predetermined range adjacent to the reference edge point Determining an approximate straight line of the edge points; determining a second line including the reference edge point and perpendicular to the approximate line; determining an edge intersection, the second line is in contact with the overhead of the reference edge point The other edge of the worn portion of the wire intersects the edge intersection; and the loss portion of the overhead contact wire is calculated based on the distance between the reference edge point and the edge intersection and the height of the overhead contact wire The width of the share. The overhead contact wire loss portion width calculating portion is configured to perform a step of determining a first group of edge points within a predetermined range adjacent to a first edge point on a scan line of the line detector image a first approximate straight line; a second Ί 306814 approximate line of a second set of edge points is determined within a predetermined range adjacent to a second edge point on the scan line; determining the relationship associated with the first set of edge points a degree of error of the first approximate straight line; determining a degree of error of the second approximate straight line associated with the second set of edge points; selecting, at the first edge point and the second edge point, a lesser degree of correlation error One of the values is a reference edge point; determining a third line including the reference edge point and perpendicular to the associated approximate line; determining an edge intersection, the third line is in contact with the overhead of the reference edge point The other edge of the worn portion of the wire intersects the edge intersection; and the distance between the reference edge point and the edge intersection and the height of the overhead contact wire Calculating the consumption of the overhead contact wire ® loss of part of the width. The overhead contact wire loss measuring device further comprises: a second line detector for scanning a top end of a pantograph; and a 'second data processing unit connected to the second line detecting The second data processing unit includes: a line detector image generating unit, wherein the second line detector combines the data to generate a line detector image; and a conductive bow tip height calculating unit, according to the line detecting The image calculates the height of the overhead contact wire. According to another aspect of the present invention, there is provided an overhead contact wire wear® inspection vehicle comprising: a body mounted with a wheel; a line detector mounted on a roof of the body and pointing upward; and a data processing unit Connected to the line detector, the data processing unit includes: a line detector image generating unit, the line detector combines the data to generate a line detector image; and an image binarization processing unit The line detector image is used to generate a binary line detector image; a noise cancellation processing unit that removes noise from the second line detector image to generate a noise to eliminate the binary line Detector image; an empty contact wire consumes part of the edge detection portion to detect the noise Ί 306814 to eliminate the edge of the overhead contact wire in the image of the dioxin detector; and an empty contact wire wear The edge of the lossy portion of the overhead contact wire detected by the portion of the width calculating portion and the height of the contact wire are used to calculate the loss-width of the overhead contact wire. According to still another aspect of the present invention, an overhead contact electrical measurement method is provided, comprising the steps of: combining a data line detector image by a line detector; and multiplexing the line detector image to generate a line detector image; Detecting the image of the Detector of the Detector from the image of the Detector of the Detector; detecting the edge of the lossy portion of one of the contact wires in the image of the Detector The width of the portion of the overhead contact wire is calculated based on the detected height of the edge overhead contact wire of the worn portion of the overhead contact wire. [Embodiment] Please refer to Figs. 1 to 5, Fig. 10, and Fig. 1 1 for the overhead contact wire loss measuring apparatus of the first embodiment. The overhead contact wire loss measuring apparatus of the first embodiment of the first invention is as shown in Fig. 1, and the first line detector 1 1 , the line detector 12 , a 13 , and a conductive bow 14 are mounted on the The top of a 10 body of a wheel body. The illumination lamp 13 is an ordinary white light source for white light to illuminate an overhead contact wire 16 . A first data processing unit connection is used for signal communication to the line detector 11. The connection of a second asset unit 2 is used for signal communication to the line detector 12. The material processing unit 1 and the second data processing unit 2 are connected to each other, according to the line loss of the overhead portion to generate a binary signal to produce a second defect; and the loss portion shows the present figure view. Illumination lamp Inspection vehicle Transmit one unit 1 material processing First capital -10- 1306814 The line detector 11 is installed on the top of the inspection vehicle 1 ,, and the direction is upwards perpendicular to the surface of the inspection vehicle 10. The line detector u is configured in such a manner that the scanning direction 18 is the same as the sleeper and the scanning line intersects the overhead contact wire 16. On the other hand, the line detector 12 is mounted on the top of the inspection vehicle 10 with the direction facing upward and inclined to the surface of the top of the inspection vehicle 10. The line detector 12 is arranged in such a manner that the scanning direction 19 is extended along the vertical direction of the inspection vehicle 1〇, and the sweeping cat line intersects the conductive bow 14. Fig. 5 is a schematic system block diagram of the overhead contact wire wear measuring apparatus of the first embodiment. The first data processing unit 1 includes components 50 to 57. As shown in FIG. 5, the line detector 11 obtains a line detector image signal F50 from each scan line and transmits a line detector image signal F50 to a line detection. The image generating unit 50. The line detector image generating unit 50 is configured to configure the line detector image signal F50 according to the timing to generate a line detector image F5 1. The line detector image F5 1 is stored in a memory 5 1 . . The line detector image F5 1 stored in the memory 51 is transferred via a data transmission line 57 and stored in a memory 52. An image binarization processing unit 53 is configured to apply an image binarization process to the line detector image F5 1 stored in the memory 52 to generate a binary line detector image F52, and Set to the second line detector image F5 2 and store it in the memory 52. A noise cancellation processing unit 54 is configured to apply a noise cancellation process to the binary line detector image F52 stored in the memory 52 to generate a noise cancellation binary line detector image F5 3 And configured to store the noise cancellation binary line detector image F53 in the memory 52. 1306814 An overhead contact wire loss portion edge detecting portion 55 is installed to apply an empty contact wire loss portion edge detection process to the noise canceling diode line detector image F 5 3 stored in the memory 52 The overhead contact wire loss portion edge data F54 is generated, and the overhead contact wire loss portion edge edge information F 5 4 is stored in the memory 52. An empty contact wire loss portion width calculating portion 56 is disposed to apply an empty contact wire loss portion width calculation process to the overhead contact wire loss portion edge data Ρ 54 stored in the memory 52 and the overhead contact wire 0 height data F55. In order to generate the overhead contact wire, the partial width data F5 8 and the overhead contact wire loss portion width data F58 are stored in the memory 52. The pan height height data F116 is used as the overhead contact wire height data F5 5 as described below. The pantograph tip height data F1 16 stored in a memory 112 is transmitted via a data transmission line 57 and a data transmission line 1 18 connected to the data transmission line 57, and is stored in the cell 5 2, as shown in FIG. Shown. The measurement processing of the width of the worn portion of the overhead contact wire 16 of the first embodiment will be described below with reference to the flowchart of Fig. 4. As shown in FIG. 4, first, in step S1 1, the line detector image signal F50 of each scan line (along the X axis) is obtained by the line detector 1 1 and is in accordance with the timing (along the t axis) The line detector image F5U is generated in the line detector image generating unit 50, and the line detector image F5 1 is stored in the memory 51 as an input image. Following the step S1 1 , in step S12, the line detector image signal F5 0 stored in the memory 51 is transferred via the data transmission line 57 and stored in the memory 52. Then, in order to set a threshold for the second generation, and Ί 306814, an image of the overhead portion of the overhead contact wire in the line detector image F5 1 is formed in the strip detector and the other background portion, based on the image The image processing unit 53 generates the image binarization processing of the threshold 値* 値 of the binary line detector image F52 and is separated from each other. Since the lossy portion of the contact wire 16 - is formed by being consumed by the pantograph 14, the worn portion of the overhead contact wire 16 has a more robust overall mass than the other undepleted portions. Therefore, in the line detector image F 5 1 , the lossy portion of the overhead contact wire 16 is revealed to be a band portion of a light-emitting device different from the background portion. This image processing produces a binary line detector image ® F52, as shown in Figure 2. In the binary line detector image 26, the lossy portion of the overhead contact wire 16 is shown as a white area 20, and the background portion is represented as a black area 21. After step S12, in step S13, the noise cancellation processing unit 54 removes the binary line detection generated by the binary line detector image F5 1 by performing an expansion process and a shrink process. The noise of the image F5 2 . This process is called noise cancellation processing. This processing is used to remove the scattering point of the noise caused by the lossy portion of the overhead contact wire 16 and the condition of the background portion from the second line detector image F5 2 . After step S1 3, in step S14, the contact wire loss portion edge detecting portion 55 detects the overhead contact wire marked as the white region 20 in the noise canceling binarization line detector image f5 3 . The left and right edges 22 of 16. This process is called overhead touch wire loss part edge detection processing. When searching along a scan line 23 in the noise cancellation binary line detector image F5 3 'a point is detected and treated as a left edge 24 (upper edge), where the color is The black area 2 1 marked as the background portion is changed to the white area 20 of the loss portion labeled -13- Ί 306814; and along the scan line 23 in the noise canceling line detector image F53 When searching, another point is detected and treated as a right edge 25 (lower edge), where the color is changed from the white area 20 marked as the loss portion to the background portion. Black area 21. This processing is performed from top to bottom on each scan line to detect the edge 3 of the loss portion of the overhead contact wire 16 in the noise canceling line detector image F5 3 . After step S14, in step S15, the overhead contact wire of the edge of the edge portion of the lossy portion of the contact wire 16 generated by the noise elimination line detector image F53 is consumed. F54, calculating the edge-to-edge distance 31 between the left and right edges 30 of each scan line of the line detector 1 as a distance on the image. This process is called the overhead contact wire loss part width calculation process. Based on the height from the line detector n to the overhead contact wire 16, a lens focal length, a detector scan width, and the number of detector pixels' can be calculated as each pixel on the image [pix The image resolution [mm/pix] of an actual size [mm]. The width of the lossy portion of the overhead contact wire can be calculated by doubling the width of the image of the overhead portion of the overhead contact wire and the image B resolution. The calculated edge data, the calculated width of the overhead contact wire of the calculation, the data indicating the line detector image used to calculate the edge data, and the data of the corresponding scan line used for confirming It is recorded on a recording area (not shown). The height F55 of the line detector 11 to the overhead contact wire 16 used for the above-described metering is calculated by measuring the height of the top end of the pantograph 5 as the height of the line detector 1 1 to the overhead contact wire 16. The height of the top end 15 of the pantograph is transmitted via the data transmission line 57 and stored in the memory 52 by 1306814. Referring to Fig. 11, a configuration of the overhead contact wire wear measuring apparatus for measuring the height of the conductive bow 1-4 is explained below. The line detector 1 2 is mounted on the front side of the conductive bow 14 on the top of the inspection vehicle 10 to measure the height of the conductive bow 14. In this manner, the scanning direction extends along the vertical direction of the conductive bow 14. . The second data processing 2 includes elements 11 〇 to 1 18 . As shown in Fig. 1 'the line detector 1 2 obtains the line detector image signal F 1 1 用于 for each scan line and sends the line detector image signal F〗i to the first line detection The image generating unit 1 1 〇. The line detector image generating unit 1 is configured to process the line detector image signal F丨丨〇 according to the sequence to generate a line detector image F 1 1 1 , and the device is also used for storing line detection. The detector image F 1 1 1 is in a memory 1 1 1 . The line detector image F 1 1 1 stored in the memory η1 is transferred via a data transmission line 丨 18 and stored in the memory 1 1 2 . The memory 112 is mounted to store and maintain a series of line detector images for a series of areas of the overhead contact wire, which can be used to address an overhead contact wire having a problem after the operation of the overhead contact wire wear measurement The part confirmed by the loss part is checked. An image binarization processing unit 1 1 3 is configured to apply an image binarization process to the line detector image F 1 1 1 stored in the memory 1 1 2 to generate a binary line detector image. F 1 1 2, and is installed to store the binary line detector image F 1 1 2 in the memory 1 1 2 . A pan width filter processing unit 114 is configured to apply a pan width filtering process to the binarized line detector image F112 stored in the memory 1 1 2 to generate a conductive bow image Fii3, and is mounted Store the image of the conductive 106814 bow F 1 1 3 in the memory 1 1 2 . A trajectory filter processing unit 115 is disposed to apply a trajectory filtering process to the shadow image of the conductive bow stored in the memory 1 1 2 to generate a trajectory filtering process. Conductive . The image of the bow is located at F 1 1 4 , and is installed to store the image F 1 14 of the conductive bow processed by the trajectory of the conductive bow at the memory 1 1 2 . A conductive bow odd characteristic filter processing unit 1 16 is installed to apply a conductive bow characteristic filtering process to the image of the conductive bow of the memory 1 1 2 to filter the conductive bow image F114 to generate a conductive bow odd characteristic Filtering ^ processes the image F115 where the conductive bow is located, and is installed to store the image of the conductive bow to filter the image of the conductive bow F 1 15 5 to the memory 1 1 2 . A pantograph tip height calculating portion 117 is configured to apply a pantograph tip height calculation process to the image of the pantograph of the conductive bow stored in the memory 1 1 2 by a conductive bow, and to filter The bow top height data F116 is installed to store the pantograph tip height data F116 in the memory 1 1 2 . The measurement process of the height of the tip end of the conductive bow 14 of the first embodiment will be described below with reference to the flowchart of Fig. 10. As shown in Fig. 10, 'first', in step S21, the line detector image signal F 1 1 0 of each scan line (along the y-axis) is obtained by the line detector 12 and is in accordance with the timing (along t The line detector image F1 11 (two-dimensional image)' and the line detector image F 1 1 1 are stored in the memory detector 1 1 1 as an input. image. After step S 2 1 , in step S 2 2, the line detector image F 1 1 stored in the memory 1 1 1 is transmitted via the data transmission line 11 8 and stored in the memory 16- 1306814. Body 1 1 2 . Then, a threshold is set for the second pass, and a pantograph 14 image of the strip and other background portions in the line detector image F 1 1 1 appears, and is used in the image binarization processing unit 113. * This image of the threshold of the diffracted line detector image FI 1 2 is generated. Deuterated and separated from each other. Since the pantograph 14 has a different overall from the background portion, the image of the pantograph 14 is revealed to be one of the illuminating bands different from the background portion of the line detector image F 1 14 Shaped part. The image processing produces a binary line detector image F 1 1 2 . In the binary line detector image F112, the image of the conductive bow 14 is represented as a ® white area 20, and the background portion is represented as a black area. Following the step S22, in step S23, the pantograph width filtering process is implemented. In particular, each scan line of the binary line detector image F 1 1 2 is inspected in the scanning direction to detect an upper edge and a lower edge. When the image of the pantograph 14 appears as a band having a constant width corresponding to the thickness of the pantograph 14 in the binarized line detector image F112, by detecting the image having the conductive bow 14 The conductive bow 14 can be confirmed by a white portion of the same width. When the interval between the upper edge and the lower edge of a scan line is within a predetermined range of width, it can be detected and confirmed as the conductive bow 14. The lower edge is kept white to indicate the top end of the conductive bow 14 while the other portions are changed to black. By repeating the above process, the position of the top end 15 of the pantograph can be confirmed. This process is called a pan width filter process. The image showing the position of the top end of the pantograph produced at this process is referred to as the image F113 where the pantograph is located. However, it is possible that when the location of the top end 15 of the pantograph is confirmed by the pan width filtering process, the 1306814 portion that does not correspond to the position of the top end of the pantograph 15 can in fact be detected as the position of the top end of the pantograph.丄5 will occur in objects such as overhead contact wires 16, insulators, and wires that use empty contact wires, such as a conductive bow, to have a width of the width of the conductive bow 14 in the wide-twist line detector image F 1 1 2 • In the case of a mode. In order to eliminate the erroneously detected portions of the noise in the conductive image F 1 1 3, it is necessary to perform the procedure. Following the step S23, in step S24, the guide track filtering process is executed. The pantograph 14 appears to have a strip shape having a constant width in the bismuth line detector F1 12, and therefore, the pantograph is a line formed by one of the top ends 15 of the pantographs in the image F113 where the pantograph is located. In this step, a series of points which are compared to a predetermined threshold value are eliminated as noise, and this processing is referred to as trajectory filtering processing. However, the trajectory filtering of the pantograph eliminates a noise caused by a short string of points, but does not eliminate the noise portion of the condition that satisfies the length of the top end of the pantograph. The following steps of the noise cancellation process are based on the fact that • 1 4 appears as a single one on the line detector image, so that the guide 1 5 also appears as a line on the scan line of the line detector 1 2 Single - causes most of the points detected by the bow width filtering process to indicate the point 15 of the conductive bow in the image of the conductive bow. Following the step S24, in step S25, the guide point filtering process is executed. When two or more points are detected by a scan line of the detector 1 2 of the conductive bow, the detected point closest to the position of the string is confirmed as - true, this The image of the bow that is close to an image can be raised in the following steps. The image of the detector is displayed as a series of points. All the shorter long conductive bows can be used to ensure that the other one is included. The top of the electric bow-point, as well as the point correctly in the mid-point of the electric bow fancy image, eliminates the point where it is detected by -18-1306814. This process is called a conductive bow odd characteristic filtering process. After step S25, in step S26, the pantograph height calculation process is performed. The image of the noise canceling the conductive bow is converted to calculate a true three-dimensional height of the top of the pantograph. The conversion includes line-based detection. A projection conversion of the mounting position and type of the device 12. Like Dijon! The connection data transmission line 118 of FIG. 75 and the data transmission line 57' shown in FIG. 5 are used as the overhead height of the overhead contact wire in the overhead contact wire loss portion width calculating portion 57 as determined above. F55. In the case where the height of an empty contact wire is pre-measured for maintenance and inspection, this measurement is used as overhead contact wire information F55. In particular, the 'memory 112' is mounted to store information about the height of the overhead contact wires 16 prior to the measurement of the overhead contact wire loss. Δ靑· FIGS. 6 and 8 show an overhead contact wire loss measuring apparatus according to a second embodiment of the present invention. Fig. 6 is a system block diagram of the overhead contact wire wear measuring apparatus of the second embodiment. In the overhead contact electric wire loss measuring apparatus shown in Fig. 6, in the step S15 shown in the flowchart of Fig. 4 of the first embodiment, the overhead contact electric wire loss portion width calculating portion 6 uses the inclination of an overhead contact electric wire. Execution of compensation - overhead contact wire wear part width calculation processing. In Fig. 8, a scan line 87 as a target includes a left edge point 80. First, 'the 'edge' within a predetermined range 81 near the left edge point 8〇. . ί is approximated to become an approximate straight line 8 2 '. A right edge point on a line 83 including a left edge point 8 〇 and a vertical approximation line 82 is defined as an edge intersection 85. The width of the contact consuming portion on the left edge point 80 is defined as the gg separation 86 between the left edge point 80 and the edge intersection 85. -19- 1306814 Referring to Figures 7 and 9, there is shown an overhead contact wire loss measuring apparatus according to a third embodiment of the present invention. Fig. 7 is a system block diagram of the rack-contact electric wire loss measuring apparatus of the third embodiment. In the step S15 shown in the flow contact diagram of the air contact electric wire loss measuring apparatus shown in Fig. 7, the overhead contact electric wire loss portion width calculating portion uses the inclination of an overhead contact electric wire. Compensating and performing another overhead contact line consumption and loss part width calculation processing. In Fig. 9, a scan line 1 〇 2 as a target includes a left φ edge 90 and a right edge 91. First, edge points within a predetermined range of the vicinity 92 of the left and right edge points 90 91 are approximated to approximate lines 93 and 94, respectively. When the sum of the approximation errors 95 of the approximate line 93 is the sum of the approximation errors 96 of the approximation line 94, an edge intersection point® is calculated as a right edge on the f spring 98 including the left edge point 90 and the vertical approximation line point. The contact wear portion width on the scan line 1〇2 is defined as a distance between the left edge point 90 and the edge intersection 99. On the other hand, when the sum of the approximation errors 96 is less than the sum of the approximation errors, an edge intersection is calculated as a left edge point on the straight line 101 including the right edge point W and the vertical approximation line 94. The contact loss portion width on the sweep line 102 is defined as a distance between the left edge intersection and the right edge point 91. The overhead contact wire loss device using the image processing of the illustrated embodiment is applied to supply power to an electric train. Or the loss measurement of the overhead contact wires of the vehicle. The advantages and effects of the overhead contact wire wear measurement arrangement of the illustrated embodiment will be described below. Empty frame 70 electric side and small 99 straight distance 95 91 cat point test -20 - 1306814 (1) Since the operation of the overhead contact wire wear measuring device does not need to directly contact the overhead contact wire, the overhead contact wire The wear measuring device can perform a high speed operation of the wear measurement so that the 'long distance of the overhead contact wire can be measured in a short time. • (2) Since the overhead contact wire wear measurement device includes a detector that is located away from an existing structure such as a point, an air area, and an anchor, the overhead contact wire wear measurement device does not need to be considered The conflict of the existing structure makes the overhead contact wire wear measuring device applicable to the place with the existing structure, and forms a method of directly measuring the thickness of an overhead contact wire by using an alternating roller and a beta optical detector. Compared. (3) Since the overhead contact wire wear measuring device generates the line detector image in all required areas, it is possible to obtain overhead contact wires in the vicinity of the adjacent areas of the overhead contact wires in all of the required areas The image data and the existing structure. (4) The overhead contact wire wear measuring device does not require the use of special light, such as a sodium lamp for a light source. ® (5) Since the overhead contact wire wear measuring device does not use laser light, the overhead contact wire wear measuring device does not need to consider the influence of a laser light on the human body, so that the device can be handled more easily, and a laser is used. The method of beaming is contrasted. (6) since the overhead contact wire wear measuring device does not use specular reflection to receive the reflected light of one of the light irradiated by the worn portion of the overhead contact wire, the overhead contact wire wear measuring device does not need to accurately target a light source and A light transmitting device performs positioning. -21- 1306814 (7) Since the image data in the measurement area is stored, it is possible to check the lossy portion of an empty contact wire that has been confirmed to be a problem by using the stored image. (8) An advantage of the overhead contact wire wear measurement apparatus based on the second embodiment is that 'when the overhead contact wire is tilted with respect to the direction of movement of an inspection vehicle, the inclination is the overhead Compensation by contact with the wires 'This is an advantage other than the first embodiment. (9) An advantage of the overhead contact wire wear measuring apparatus based on the third embodiment is that the inclination is compensated to be more correct when the contact wire is tilted with respect to a direction in which the inspection vehicle moves. The ground decides the overhead connection. The width of the worn portion of the electric wire is an advantage other than the first embodiment. In particular, 'even if there is wavy wear at the edge of one of the lossy parts of the overhead contact wire or a series of notch shape loss', it is possible to more accurately determine the width of the lossy portion of the overhead contact wire. The present application is based on Japanese Patent Application No. 2005-68796, filed on Jan. 11, 2005, which is hereby incorporated by reference. Although the present invention has been described above with reference to certain embodiments of the present invention, the present invention is not limited to the above-described embodiments, and the above-described adjustments and variations of the embodiments may be familiar to the technology. The scope of the present invention is defined by the scope of the following claims. -22- 1306814 BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a perspective view showing an overhead contact wire loss measuring apparatus according to first to third embodiments of the present invention. Fig. 2 is a schematic view of the image of the dichroic line detector of the first to third embodiments. Fig. 3 is a schematic view showing the processing of edge detection of the first to third embodiments. Fig. 4 is a flow chart showing the operational processing of the overhead contact wire loss measuring apparatus of the first embodiment. B Fig. 5 is a system block diagram of the overhead contact wire loss measuring apparatus of the first embodiment. Fig. 6 is a system block diagram of the overhead contact wire wear measuring apparatus of the second embodiment. Fig. 7 is a system block diagram of the overhead contact wire loss measuring apparatus of the third embodiment. Fig. 8 is a view showing the compensation processing applied to the inclination of the overhead contact electric wire of the second embodiment. β Fig. 9 is a view showing the compensation processing of the slope applied to the overhead contact electric wire of the third embodiment. Fig. 10 is a flow chart for processing the height of the pantograph in the first to third embodiments. Fig. 11 is a system block diagram of an overhead contact wire loss measuring apparatus for measuring the height of the pantograph in the first to third embodiments. -23 - 1306814 [Component Symbol Description]
1 第一資料處理單元 2 第二資料處理單元 10 檢查車輛 11 線偵測器 12 線偵測器 13 照射燈 14 導電弓 15 導電弓頂端 16 架空接觸電線 18 掃猫方向 19 掃瞄方向 20 白區域 21 黑區域 22 左及右邊緣 23 掃猫線 24 左邊線 25 右邊線 26 二値化線偵測器影 30 邊緣 3 1 邊緣至邊緣距離 80 左邊緣點 81 預定範圍 82 近似直線 -24 - 近似直線 邊緣交叉點 距離 掃瞄線 左邊緣點 右邊緣點 附近 近似直線 近似直線 近似誤差 近似誤差 直線 邊緣交叉點 距離 掃猫線 掃瞄線 -25 -1 First data processing unit 2 Second data processing unit 10 Inspection vehicle 11 Line detector 12 Line detector 13 Illumination lamp 14 Conductor bow 15 Conductor tip 16 Overhead contact wire 18 Sweeping direction 19 Scanning direction 20 White area 21 Black area 22 Left and right edge 23 Sweeping cat line 24 Left line 25 Right side line 26 Dividing line detector shadow 30 Edge 3 1 Edge to edge distance 80 Left edge point 81 Predetermined range 82 Approximate line - 24 - Approximate straight line Edge intersection point scan line left edge point right edge point near approximate line approximation straight line approximation error approximation error line edge intersection point distance sweeping cat line scan line -25 -