橋梁一直是人類重要的基礎設施建設，舉凡生活中的運輸、交通與經濟都佔著舉足輕重的角色。然而台灣的地理位置特殊，常有災害造成橋梁的損害。地理部分，台灣位於菲律賓海板塊與歐亞板塊的邊界，板塊之間的移動容易導致劇烈的地震與造山活動。而氣候部分，位於副熱帶與熱帶交界處，總是帶著強風與豪雨的颱風每年頻繁的入侵台灣。這些災害常造成橋樑結構與表面的毀損，一旦橋梁發生嚴重的損害，容易導致交通中斷與人員的傷亡，為了安全起見，橋梁檢測必須嚴格的執行，橋面的裂縫與剝落是反映橋梁安全狀況的重要指標。基本傳統的檢測裂縫方法主要是目視檢測，由檢測人員逐一目視橋梁外觀，依專業判斷與紀錄損壞的情況，然而通常橋梁的體積非常巨大且橋板距離地面高度很高，因此檢測的環境較為危險且檢測時間耗時，在裂縫的判斷方面難免有主觀上的盲點。在本次研究中，我們利用物件導向分析(Object-Based Image Analysis,OBIA)的方法進行自動化的裂縫檢測。首先在欲監測的強樑旁佈設控制點，接著將消費型相機安裝於多旋翼無人機上(Unmanned Aerial Vehicle,UAV)並對橋梁的表面進行拍照，利用多個影像前處理的技術來增加影像的反差與減少光線不均勻的影響。在物件導向分析階段，多重解析度切割將相片切割成多個具有相似性質的物件並可利用多個物件特徵來描述物件的型態，接著設定一個規則集(rule set)來區分出裂縫的物件。偵測出裂縫位於相片上的位置後，將原始相片轉換成核影像消除y視差以增加後續計算的效率，接著半全域匹配法(Semi-Global Matching,SGM)來尋找裂縫像點的共軛點，共軛點間即可用前方交會的方法來得到裂縫的三維空間坐標。除裂縫以外，剝落也是一個重要的劣化指標，透過人工於影像圈選出剝落大致的位置後，利用產製核影像、半全域匹配與前方交會製作出剝落區的密集點雲，將其與擬合的平面相減得到所有點雲與平面之高度差並製作表面高度模型，藉由高度差可偵測出剝落的位置，即可計算剝落區的三維空間資訊。後續透過前後期裂縫與剝落區的三維空間資訊比對以利協助橋梁的安全檢測。

Bridge is always an important infrastructure construction for the human being because of the benefit to transportation, economy and communication. However, the earthquake and typhoon are the main disasters in Taiwan. Once the bridge is damaged, it will result in the traffic interruption and casualties. For the purpose of the safety, the bridge monitoring must be strictly implemented. Therefore, crack and concrete delamination are important indicators reflecting the safety status of infrastructures. In conventional way, inspection data are obtained manually by using geotechnical field instrumentations and visual inspection that are time-consuming and dangerous for inspector. In this research, the consumer grade digital camera installed on a multi-rotary UAV to capture pictures of bridge surface that can prevent the risk of on-site inspection of cracks. Several image pre-processing techniques are used to enhance the image contrast and to prevent the effect of uneven illumination. Then we adopt object-based image analysis (OBIA) method for automatic crack detection. In addition to the crack, the concrete delamination is another important indicator reflecting the safety of bridge. The approximate location of concrete delamination is manual selected. The technique of photogrammetry will be used to generate the dense point cloud of this area. Then the digital elevation model can be obtained by using the dense point cloud to minus the fitting plane. By means of digital elevation model, we can calculate the three-dimensional information of concrete delamination. The comparison of the three-dimensional information of deteriorating area in different times can assist the safety inspection of bridges.

王澤雄, 2003. 高處橋面板下結構檢測維修之勞工安全研究. 國立台北科技大學土木與防災研究所, 184.
交通部, 2008. 公路鋼結構橋梁之檢測及補強規範.
何坤居, 2006. 熱紅外線影像檢測鋪面破壞之可行性評估. 國立高雄科技大學土木工程與防災科技研究所論文,75.
施旻汶, 2015. 自動化無人飛行載具輔助擷取橋梁影像之研究. 國立中央大學營建管理研究所學位論文, 119.
曾子峻, 2011. 多軸飛行器於橋梁檢測之應用研究. 國立高雄應用科技大學土木工程與防災科技研究所.
薛強, 張權, 翁健煌,徐偉誌,陳正忠,葉啟章, 2012. 台灣橋梁檢測有效實施建議. 中興工程.
Abdel-Qader, I., Abudayyeh, O., Kelly, M.E., 2003. Analysis of edge-detection techniques for crack identification in bridges. Journal of Computing in Civil Engineering 17, 255-263.
Baatz, M., 2000. Multiresolution segmentation: an optimization approach for high quality multi-scale image segmentation. Angewandte Geographische Informations verarbeitung XII, 12-23.
Blaschke, T., 2010. Object based image analysis for remote sensing. ISPRS Journal of Photogrammetry and Remote Sensing 65, 2-16.
Cho, W., Schenk, T., Madani, M., 1993. Resampling digital imagery to epipolar geometry. International Archives of Photogrammetry and Remote Sensing 29, 404-404.
Fraser, C.S., 1997. Digital camera self-calibration. ISPRS Journal of Photogrammetry and Remote sensing 52, 149-159.
Gavilán, M., Balcones, D., Marcos, O., Llorca, D.F., Sotelo, M.A., Parra, I., Ocaña, M., Aliseda, P., Yarza, P., Amírola, A., 2011. Adaptive road crack detection system by pavement classification. Sensors 11, 9628-9657.
Hirschmuller, H., 2008. Stereo processing by semiglobal matching and mutual information. IEEE Transactions on pattern analysis and machine intelligence 30, 328-341.
Lang, S., Albrecht, F., Blaschke, T., 2011. OBIA-Tutorial,Centre for Geoinformatics(Z_GIS). Paris-Lodron University Salzburg.
Lillesand, T., Kiefer, R.W., Chipman, J., 2014. Remote sensing and image interpretation. John Wiley & Sons.
Otsu, N., 1979. A threshold selection method from gray-level histograms. IEEE transactions on Systems, Man, and Cybernetics 9, 62-66.
Rashed, M., 2016. Rolling ball algorithm as a multitask filter for terrain conductivity measurements. Journal of Applied Geophysics 132, 17-24.
Sternberg, S.R., 1983. Biomedical image processing. Computer 16, 22-34.
Talab, A.M.A., Huang, Z., Xi, F., HaiMing, L., 2016. Detection crack in image using Otsu method and multiple filtering in image processing techniques. Optik-International Journal for Light and Electron Optics 127, 1030-1033.
Whiteside, T., Ahmad, W., 2005. A comparison of object-oriented and pixel-based classification methods for mapping land cover in northern Australia, Proceedings of SSC2005 Spatial intelligence, innovation and praxis: The national biennial Conference of the Spatial Sciences Institute, pp. 1225-1231.
Yamaguchi, T., Hashimoto, S., 2010. Fast crack detection method for large-size concrete surface images using percolation-based image processing. Machine Vision and Applications 21, 797-809.
Zhang, W., Zhang, Z., Qi, D., Liu, Y., 2014. Automatic crack detection and classification method for subway tunnel safety monitoring. Sensors 14, 19307-19328.