鄭富書臺灣大學：土木工程學研究所曹峻瑋Chao, Chun-WeiChun-WeiChao2007-11-252018-07-092007-11-252018-07-092006http://ntur.lib.ntu.edu.tw//handle/246246/50103傳統隧道剖面監測方法，常需花費許多人力、財力、時間，且針對單一隧道剖面而言，僅能得到少數監測點之資訊。鑑於現今光學技術及電腦運算能力之進步，若能善用此一特性，發展一套易於操作、快速、經濟且適用於隧道工程之掃瞄新技術，透過長時間之定期監測工作，以獲得更完整之隧道剖面三維座標資訊，如此將有助於工程師了解整體隧道長時間之變形趨勢，並做為未來維修補強之參考及反算周圍岩體塑性區發展之依據。 循此，本研究主要目的係發展一套非接觸性量測之新技術，利用二維雷射影像，經程式處理後，建構出隧道剖面三維座標資訊。有別於傳統監測方法，本研究技術可有效的利用少數已知三維座標值之校正點，得知完整且詳盡之隧道剖面三維資訊。 研究過程中，除施作小型模擬隧道掃瞄試驗外，亦製作一尺寸接近真實隧道之大型模擬隧道施作掃瞄試驗。由小型模擬隧道試驗結果中發現，當隧道頂拱處產生整體變形時，利用本研究掃瞄技術，可有效觀察出不同時期之隧道變形趨勢；而於大型模擬隧道疊合點雷射試驗結果發現，利用本研究掃瞄技術分析出之隧道頂拱剖面，其RMSE值約可控制在4 mm內。 此外，由於本研究係利用影像處理之方式，建構出待測物體三維資訊，在研究過程中發現，主要影響分析結果之誤差來源，多來自於影像部份，未來若欲提升此掃瞄技術之準確度，應於影像處理部份多做考量。Traditional tunnel’s profile monitoring usually costs a lot of manpower, finance and time, and it utilizes only few, selected and fixed observation points in a single tunnel’s profile to get the desired information. In the advancement of optics’ technique and of the performance of computer recently, the development of a new monitoring technique, which is easy to operate, rapid and inexpensive, for tunnel engineering might be helpful for us to equally understand the deformation tendency in a tunnel. From this, derived results can serve as a reference for future repairing of the support system, or the parameters of back analyzing about the development of plastic zone considering long-term and/or regular monitoring works, and for developing a more complete three-dimensional (3-D) information about tunnel’s profile. This study was conducted to develop a new non-contact measurement technique which can be applied to tunnel engineering using a two-dimensional laser image and a program to construct the 3-D information in a certain profile. Different from traditional monitoring, this technique can provide a more complete 3-D profile information by first establishing several points whose 3-D information were already known for calibration. In addition to a small simulative tunnel, a bigger one was constructed in the laboratory to simulate the full-scale tunnel in this study. Based from the results in small tunnel, it was observed that the measurement of deformation behavior from the two successive analyses performed is effective. Also, profiling results from this technique and a total-station-based method were compared and showed that RMSE value can be controlled in about 4 mm in this new technique. Moreover, this research uses image processing to construct a 3-D tunnel’s profile information, thus, the influence are mostly from images. Therefore, improvement in the image processing method is advised to achieve a more precise result.摘要.......................................................I Abstract..................................................II 目錄.....................................................III 表目錄....................................................VI 圖目錄..................................................VIII 第一章 緒論.............................................1-1 1.1 研究動機.........................................1-1 1.2 研究目的.........................................1-2 1.3 研究內容與方法...................................1-3 1.4 本文內容.........................................1-4 第二章 文獻回顧.........................................2-1 2.1 隧道變形機制及危險分級...........................2-1 2.2 隧道剖面掃瞄系統比較.............................2-1 2.2.1 TS360隧道全斷面掃瞄系統....................2-1 2.2.2 BIPS 掃瞄系統..............................2-2 2.2.3 DIBIT 掃瞄系統.............................2-2 2.3 三維雷射掃瞄儀器原理介紹.........................2-3 2.3.1 時間差法(TOF, Time of Flight)..............2-3 2.3.2 三角法(Triangulation)......................2-3 2.4 三維雷射掃瞄儀器應用現況.........................2-4 2.5 影像立體成像原理.................................2-6 2.5.1 主動式立體視覺法...........................2-6 2.5.2 被動式立體視覺法...........................2-6 2.5.3 本研究之原理...............................2-7 2.6 綜合討論........................................2-12 第三章 研究方法.........................................3-1 3.1 試驗儀器.........................................3-1 3.1.1 試驗模型...................................3-1 3.1.2 攝像機具...................................3-3 3.1.3 雷射光源...................................3-3 3.1.4 校正盒.....................................3-4 3.1.5 全測站(Leica TCR705).......................3-4 3.1.6 RIEGL(LMS-Z360i)三維雷射掃瞄儀器簡介.......3-4 3.1.7 實驗配置...................................3-5 3.2 雷射掃瞄技術改良.................................3-5 3.2.1 攝像機具...................................3-5 3.2.2 雷射光源...................................3-8 3.2.3 影像疊合技術...............................3-9 3.2.4 雙相機改善遮蔽問題........................3-10 3.2.5 掃瞄型式..................................3-12 3.3 實驗方法........................................3-14 3.3.1 精度試驗..................................3-14 3.3.2 平面式掃瞄試驗............................3-14 3.3.3 空間式(隧道)掃瞄試驗......................3-14 3.3.4 RIEGL三維雷射掃瞄儀器試驗.................3-15 3.4 影像分析方法介紹................................3-15 3.4.1 分析軟體..................................3-16 3.4.2 分析硬體設備..............................3-16 3.4.3 建構三維數值模型..........................3-17 3.4.4 掃瞄結果疊合..............................3-17 第四章 試驗分析結果討論.................................4-1 4.1 三維雷射掃瞄試驗精度分析討論.....................4-1 4.2 平面式三維雷射掃瞄試驗結果討論...................4-4 4.2.1 平面式線雷射掃瞄試驗.......................4-4 4.2.2 平面式疊合點雷射掃瞄試驗...................4-5 4.3 小型模擬隧道掃瞄試驗結果討論.....................4-6 4.3.1 單一線雷射掃瞄試驗.........................4-7 4.3.2 疊合線雷射掃瞄試驗.........................4-9 4.3.3 疊合點雷射掃瞄試驗........................4-10 4.4 大型模擬隧道三維雷射掃瞄試驗結果討論............4-12 4.4.1 疊合線雷射掃瞄試驗........................4-13 4.4.2 疊合點雷射掃瞄試驗........................4-15 4.4.3 存在環境光源疊合點雷射掃瞄試驗............4-17 4.5 RIEGL三維雷射掃瞄儀分析結果討論.................4-18 4.5.1 RIEGL三維雷射掃瞄儀器自我誤差試驗.........4-18 4.5.2 大型模擬隧道RIEGL三維雷射掃瞄儀試驗.......4-19 4.6 綜合討論........................................4-21 4.6.1 小型模擬隧道掃瞄試驗......................4-21 4.6.2 大型模擬隧道掃瞄試驗......................4-22 第五章 誤差來源討論.....................................5-1 5.1 可能誤差來源定性討論.............................5-1 5.1.1 傳統量測部份...............................5-1 5.1.2 影像處理階段...............................5-3 5.2 可能改善誤差來源方式.............................5-5 5.2.1 傳統量測部份...............................5-5 5.2.2 影像處理階段...............................5-5 5.3 大型模擬隧道試驗誤差來源定量討論.................5-7 5.3.1 校正點輔以經緯儀量測前後分析結果討論.......5-7 5.3.2 不同光圈曝光時間分析結果討論...............5-8 第六章 結論與建議.......................................6-1 6.1 結論.............................................6-1 6.2 建議與未來改善...................................6-4 參考文獻.................................................R-1 附錄A DIBIT隧道掃瞄系統.................................A-1 附錄B 三維雷射掃瞄技術分析流程...........................B-1 附錄C 現地掃瞄試驗.......................................C-1 附錄D 相機鏡頭扭曲量計算.................................D-148651409 bytesapplication/pdfen-US三維雷射掃瞄隧道影像3DLaserScanTunnelImagethesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/50103/1/ntu-95-R93521104-1.pdf