請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/88482
完整後設資料紀錄
DC 欄位 | 值 | 語言 |
---|---|---|
dc.contributor.advisor | 陳柏翰 | zh_TW |
dc.contributor.advisor | Po-Han Chen | en |
dc.contributor.author | 盧湘凱 | zh_TW |
dc.contributor.author | Hsiang-Kai Lu | en |
dc.date.accessioned | 2023-08-15T16:30:07Z | - |
dc.date.available | 2023-11-09 | - |
dc.date.copyright | 2023-08-15 | - |
dc.date.issued | 2023 | - |
dc.date.submitted | 2023-08-04 | - |
dc.identifier.citation | [1]羅俊雄、饒正。(2007)。公共工程侵蝕及陰極防護現況探討。防蝕工程,第21 卷,第1 期,第1~20 頁。
[2]陳嘉彌。(1983)。金屬的侵蝕理論與防蝕設計。中學工藝教育,第16卷,第6期,第22-27頁。 [3]羅建明、溫志中、林鳳嬌、王俊欽、周偉龍、葉秀貞。(2019)。離岸風機基樁與金屬構件侵蝕防治之研究。交通部運輸研究所研究報告。 [4]Stephen Lower. (2021). Electrochemical Corrosion. Retrieved May 10, 2021, from https://chem.libretexts.org/@go/page/265 [5]Wang, S. G., Sun, M., Xu, Y. H., Long, K., & Zhang, Z. D. (2018). Enhanced localized and uniform corrosion resistances of bulk nanocrystalline 304 stainless steel in high-concentration hydrochloric acid solutions at room 盧湘凱論文(V)temperature. Journal of Materials Science & Technology, 34(12), 2498-2506. [6]Donatus, U., Terada, M., Ospina, C. R., Queiroz, F. M., Bugarin, A. F. S., & Costa, I. (2018). On the AA2198-T851 alloy microstructure and its correlation with localized corrosion behaviour. Corrosion Science, 131, 300-309. [7]Kolesnikov, A., Kolek, M., Dohmann, J. F., Horsthemke, F., Börner, M., Bieker, P., ... & Stan, M. C. (2020). Galvanic corrosion of lithium‐powder‐based electrodes. Advanced Energy Materials, 10(15), 2000017. [8]Zhang, X., Zhou, X., Hashimoto, T., & Liu, B. (2017). Localized corrosion in AA2024-T351 aluminium alloy: Transition from intergranular corrosion to crystallographic pitting. Materials Characterization, 130, 230-236. [9]Sieradzki, K., & Newman, R. C. (1987). Stress-corrosion cracking. Journal of physics and chemistry of solids, 48(11), 1101-1113. [10]Railkin, A. I. (2003). Marine biofouling: colonization processes and defenses. CRC press. [11]Okeniyi, J. O., Ogunlana, O. O., Ogunlana, O. E., Owoeye, T. F., & Okeniyi, E. T. (2015). Biochemical characterisation of the leaf of Morinda lucida: Prospects for environmentally-friendly steel-rebar corrosion-protection in aggressive medium. In TMS 2015 144th Annual Meeting & Exhibition (pp. 637-644), Springer, Cham. [12]羅建明、張道光。(2009)。高雄港碼頭鋼板樁侵蝕特性研究。交通部運輸研究所研究報告。 [13]吳興練。(2016)。離岸風機塔架塗裝系統奈蝕性評估。防蝕工程,第30卷,第1期,第24~37頁。 [14]呂傳盛。(2007)。船底防寄生先進奈米薄膜層及施工方法之研究。行政院農業委員會漁業署科技計畫研究報告。 [15]Lamastra, F. R., Leonardi, F., Montanari, R., Casadei, F., Valente, T., & Gusmano, G. (2006). X-ray residual stress analysis on CrN/Cr/CrN multilayer PVD coatings deposited on different steel substrates. Surface and Coatings technology, 200(22-23), 6172-6175. [16]Vetter, J. (2014). 60 years of DLC coatings: historical highlights and technical review of cathodic arc processes to synthesize various DLC types, and their evolution for industrial applications. Surface and Coatings Technology, 257, 213-240. [17]Grill, A. (1999). Diamond-like carbon: state of the art. Diamond and related materials, 8(2-5), 428-434. [18]Shibli, S. M. A., Meena, B. N., & Remya, R. (2015). A review on recent approaches in the field of hot dip zinc galvanizing process. Surface and Coatings Technology, 262, 210-215. [19]Pedeferri, P. (1996). Cathodic protection and cathodic prevention. Construction and building materials, 10(5), 391-402. [20]葉江榮、蔡茂雄、邱明義、廖財昌。(1986)。外加電流式自動控制電位陰極防蝕法。防蝕工程,第1卷,第6期,第29~41頁。 [21]羅俊雄、饒正。(2007)。公共工程侵蝕及陰極防蝕現況探討。防蝕工程,第21卷,第1期,第1~20頁。 [22]LeCun, Y., Bengio, Y., & Hinton, G. (2015). Deep learning. Nature, 521(7553), 436-444. [23]Chuang, M. C., Hwang, J. N., & Williams, K. (2016). A feature learning and object recognition framework for underwater fish images. IEEE Transactions on Image Processing, 25(4), 1862-1872. [24]Marini, S., Fanelli, E., Sbragaglia, V., Azzurro, E., Fernandez, J. D. R., & Aguzzi, J. (2018). Tracking fish abundance by underwater image recognition. Scientific reports, 8(1), 1-12. [25]Jin, L., & Liang, H. (2017). Deep learning for underwater image recognition in small sample size situations. In OCEANS 2017-Aberdeen (pp. 1-4). IEEE. [26]Li, C., Guo, C., Ren, W., Cong, R., Hou, J., Kwong, S., & Tao, D. (2019). An underwater image enhancement benchmark dataset and beyond. IEEE Transactions on Image Processing, 29, 4376-4389. [27]Sun, X., Shi, J., Liu, L., Dong, J., Plant, C., Wang, X., & Zhou, H. (2018). Transferring deep knowledge for object recognition in Low-quality underwater videos. Neurocomputing, 275, 897-908. [28]Lu, J., Li, N., Zhang, S., Yu, Z., Zheng, H., & Zheng, B. (2019). Multi-scale adversarial network for underwater image restoration. Optics & Laser Technology, 110, 105-113. [29]Jiang, Q., Chen, Y., Wang, G., & Ji, T. (2020). A novel deep neural network for noise removal from underwater image. Signal Processing: Image Communication, 87, 115921. [30]Shen, H. K., Chen, P. H., & Chang, L. M. (2018). Human-visual-perception-like intensity recognition for color rust images based on artificial neural network. Automation in Construction, 90, 178-187. [31]白博升。(2017)。結合擴增虛擬實境與即時影像辨識之工程應用---以鋼橋侵蝕辨識為例。臺灣大學土木工程學研究所碩士論文,台北。 [32]陳思愷。(2019)。應用全卷積神經網絡: U-Net 於侵蝕語義分割影像辨識。臺灣大學土木工程學研究所碩士論文,台北。 [33]Yuan, J., Xue, B., Zhang, W., Xu, L., Sun, H., & Zhou, J. (2019). RPN-FCN based Rust detection on power equipment. Procedia computer science, 147, 349-353. [34]Ren, S., He, K., Girshick, R., & Sun, J. (2015). Faster r-cnn: Towards real-time object detection with region proposal networks. arXiv preprint arXiv:1506.01497. [35]He, K., Gkioxari, G., Dollár, P., & Girshick, R. (2017). Mask r-cnn. In Proceedings of the IEEE international conference on computer vision (pp. 2961-2969). [36]Z. Li et al., "Corrosion detection and classification of metal surfaces using deep learning techniques," Materials and Corrosion, vol. 70, no. 6, pp. 773-781, 2019. [37]Y. Liu et al., "Deep learning-based corrosion detection and classification for metal surfaces," Applied Sciences, vol. 9, no. 10, pp. 2111, 2019. [38]S. Alharbi et al., "Deep learning approach for corrosion detection on metal surfaces," Measurement, vol. 156, pp. 256-265, 2020. [39]Y. Ma et al., "Corrosion detection and classification of metal surfaces based on deep learning and computer vision techniques," Materials, vol. 13, no. 8, pp. 1751, 2020. [40]M. J. Alharbi et al., "Deep learning techniques for corrosion detection and classification of metal surfaces," Engineering, vol. 6, no. 5, pp. 621-631, 2020. [41]影像辨識常見的IOU、AP、mAP是什麼意思? :http://yy-programer.blogspot.com/2020/06/iouapmap.html | - |
dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/88482 | - |
dc.description.abstract | 在港灣工程專案裡,金屬鋼材廣泛運用於結構支撐關鍵材料,然海洋生物附著與海水成分容易造成腐蝕,長時間浸泡於海水對鋼材之支撐力造成損害。舉例來說,鋼板於海中施放三個月後,在穿孔或裁切區域易產生間隙腐蝕現象,使工程進行中裸露於海水的鋼材,在長期使用上存在結構安全的疑慮,必須定時撤換新的鋼材增加施工或營運成本。因此,本研究將以大鵬灣港灣工程為例,進行海水中鋼材耐腐蝕與海洋生物附著之研究。深度學習已經被廣泛應用在各個領域中的深度學習方法,聚焦於海面下鋼材侵蝕之辨識。該學習方法已運用於陸地上鋼材侵蝕之辨識與海面下魚群之辨識,但尚未有研究著重在辨識海面下金屬侵蝕情況,原因為海底下侵蝕照片資料取得不易,再加上海水中生物與光線不足,會使得影像變得模糊不清,增加了辨識難度。由於資料收集不易,故設計一實驗來探討海面下金屬侵蝕發生原因與提供深度學習方法訓練之照片,本實驗使用不鏽鋼試片規格為:長度15公分、寬度7公分、厚度0.3公分、重量平均165公克。試片施放地點為大鵬灣國家風景管理處,因潮汐現象明顯,適合做於本實驗試放地點,本實驗並將試片分成三組:潮間帶組、水下清潔組、水下不動組,水下清潔組指每次實驗紀錄後會刮除試片附著物,反之水下不動組指任其恣意生長,目的為確認清潔是否會造成影響,潮間帶組也會做清潔動作。在深度學習方法中,本研究使用實驗試片水下之侵蝕照片,共蒐集了約300張侵蝕照片作為訓練資料、44張照片作為測試資料訓練,選用Mask R-CNN演算法來進行侵蝕影像辨識,Mask R-CNN為近期非常受歡迎的演算法,它不僅可以快速的偵測出辨識目標,還能準確的將辨識目標的形狀框架判別出來。因此,將上述實驗所獲得的海面下金屬侵蝕照片,用以訓練Mask R-CNN深度學習模型。本研究建立一種半自動檢測方式,隨著未來深度學習技術提升,更有機會將辨識技術用於調查海下結構物安全之用。結果表明,該模型可成功檢測不同表面處理之金屬侵蝕區域。 | zh_TW |
dc.description.abstract | In the harbor project, metal steel is widely used in the key materials of structural support, but the composition of Marine life is easy to cause corrosion, and soaking in seawater for a long time will cause damage to the support force of steel. For example, after the steel plate is applied in the sea for three months, the gap corrosion phenomenon may occur in the perforated or cutting area, which makes the steel exposed to the seawater have doubts about structural safety in the long-term use, and the new steel must be replaced regularly to increase the construction or operating costs. Therefore, this study will take the Dapeng Bay Marine Project as an example to study the corrosion resistance of steel and Marine life in seawater. This study uses deep learning methods, which have been widely used in various fields, focusing on the identification of steel erosion under the sea surface. The learning method has been applied to the identification of steel erosion on land and fish under the sea surface, but no research has been focused on the identification of metal erosion under the sea surface. The reason is that the photo erosion under the sea surface is not easy to obtain, coupled with the lack of biology and light in the sea, will make the image blurred and increase the difficulty of identification.In this study, an experiment was designed to investigate the causes of metal erosion under the sea surface and provide the training of deep learning methods. The stainless steel test pieces were 15 cm long, 7 cm wide, 0.3 cm thick, and the average weight of 165 grams. Trial casting place for dapeng bay national landscape management, because of the tidal phenomenon is obvious, suitable for this experiment test site, this experiment and will be divided into three groups: intertidal group, underwater cleaning group, underwater group, underwater cleaning group refers to scraping after each experiment record attached objects, while the underwater group arbitrary growth, purpose to confirm whether clean will affect, intertidal group will also do clean action.In deep learning method, this study using the experimental test underwater erosion photos, collected about 300 pieces of erosion photos as training data, 44 photos as test data training, choose Mask R-CNN algorithm to erosion image identification, Mask R-CNN recently very popular algorithm, it can not only quickly detect the target, also can accurately identify the shape of the target framework. Therefore, this study used the sub-surface metal erosion photos obtained from the above experiments to train the Mask R-CNN deep learning model. In this study, a semi-automatic detection method is established. With the improvement of deep learning technology in the future, there is more opportunity to use the identification technology to investigate the safety of offshore structures. The results show that the model can successfully detect the different surface treatments. | en |
dc.description.provenance | Submitted by admin ntu (admin@lib.ntu.edu.tw) on 2023-08-15T16:30:07Z No. of bitstreams: 0 | en |
dc.description.provenance | Made available in DSpace on 2023-08-15T16:30:07Z (GMT). No. of bitstreams: 0 | en |
dc.description.tableofcontents | 口試委員審定書............................I
致謝......................................II 摘要......................................III Abstract..................................V 目錄......................................VII 圖目錄.................................... X 表目錄....................................XIII 第一章 緒論...............................1 1.1. 研究背景..............................1 1.2. 研究動機..............................2 1.3. 研究目的..............................3 1.4. 研究限制..............................3 第二章 文獻回顧.............................5 2.1. 金屬侵蝕成因...........................5 2.1.1. 電化學侵蝕..........................5 2.1.2. 物理性侵蝕.........................7 2.1.3. 生化性侵蝕.........................7 2.2. 金屬防侵蝕方法.........................8 2.2.1. 塗層法(Coating)....................8 2.2.2. 熱浸鍍鋅(Hot-Dip Galvanization)....8 2.2.3. 陰極保護法(Cathodic Protection)....9 2.3. 深度學習.............................10 2.3.1. 水下影像處理與辨識...................10 2.3.2. 侵蝕影像處理與辨識...................12 2.4. 小結.................................13 第三章 研究方法............................14 3.1. 研究流程.............................14 3.2. 侵蝕試片取得..........................15 3.3. 深度學習辨識-卷積神經網路 CNN(Convolutional Neural Network)...........................16 3.3.1. 卷積層(Convolution Layer).........16 3.3.2. 池化層(Pooling Layer)............17 3.3.3. 全連接層(Fully Connected Layer)...18 3.3.4. 神經網路(Neural Network).......... 18 第四章 實驗設計.............................21 4.1. 試片處理辦法..........................21 4.1.1. 材料備製..........................21 4.1.2. 代號介紹..........................22 4.1.3. 試片架設及記錄設計..................22 4.1.4. 研究紀錄工具.......................23 4.1.4.1. 重量量測.......................23 4.1.4.2. 試片清潔.......................23 4.1.4.3. 影像紀錄.......................23 4.1.4.4. 試片記錄方式....................24 4.1.5. 施放環境..........................25 4.2. 深度學習前處理-標註....................26 第五章 實驗結果與分析.......................29 5.1. 海洋生物附著情形......................29 5.2. 侵蝕與穿孔情形........................29 5.3. 拋光之影響............................30 5.4. 拉伸實驗探討..........................31 5.4.1. 機械性質-拉伸試驗..................32 5.4.2. 耐侵蝕性分析.......................34 5.5. 放置環境之探討........................34 5.5.1. A 系列-潮間帶.....................34 5.5.2. B 系列-水下清潔組..................39 5.5.3. C 系列-水下不動組..................43 第六章 影像辨識應用.........................47 6.1. 海底影像收集..........................47 6.2. 訓練方法及流程........................47 6.3. 辨識結果討論...........................48 6.3.1. 辨識成果...........................49 6.3.2. 辨識成果評估.......................51 6.3.3. 辨識影像討論.......................52 第七章 結論與建議...........................55 7.1. 發現與貢獻............................55 7.2. 未來方向及建議........................56 參考文獻..................................58 | - |
dc.language.iso | zh_TW | - |
dc.title | 應用深度學習於海面下金屬侵蝕辨識–以大鵬灣海域為例 | zh_TW |
dc.title | Application of Deep Learning Method to Underwater Rust Image Recognition–In Case of Dapeng Bay | en |
dc.type | Thesis | - |
dc.date.schoolyear | 111-2 | - |
dc.description.degree | 碩士 | - |
dc.contributor.coadvisor | 林之謙 | zh_TW |
dc.contributor.coadvisor | Jacob J. Lin | en |
dc.contributor.oralexamcommittee | 曾惠斌;張陸滿;紀乃文 | zh_TW |
dc.contributor.oralexamcommittee | Hui-Ping Tserng;Luh-Maan Chang;CHI NAI-WEN | en |
dc.subject.keyword | 港灣工程,深度學習,Mask R-CNN,海面下鋼材侵蝕之辨識,半自動檢測方式, | zh_TW |
dc.subject.keyword | harbor engineering,deep learning,Mask R-CNN,identification of steel erosion under the sea surface,semi-automatic detection method, | en |
dc.relation.page | 61 | - |
dc.identifier.doi | 10.6342/NTU202302452 | - |
dc.rights.note | 未授權 | - |
dc.date.accepted | 2023-08-04 | - |
dc.contributor.author-college | 工學院 | - |
dc.contributor.author-dept | 土木工程學系 | - |
顯示於系所單位: | 土木工程學系 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-111-2.pdf 目前未授權公開取用 | 5.68 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。