在每次地震後，建築物的殘餘耐震能力是否仍足夠是民眾最為關心的。在目前國內震後評估中，欲瞭解建築物的殘餘強度，僅有以詳細評估經由側推分析等方式去估算建物的殘餘耐震能力，然而其成本龐大，因此開發一個較經濟性的評估方法推估其殘餘耐震能力，已成為近期地震工程研究的重要議題。本研究根據用現有的RC震損構件相關研究成果，對既有的初步評估方法提出構件的殘餘強度折減係數，應用於初步評估方法推估建築物震後殘餘耐震能力。近年來，電腦硬體設備的普及和進步，使得深度學習的應用大幅成長，其中卷積類神經網路的出現，使影像識別技術有了更多的發展性。過去已有研究透過卷積類神經網路萃取損傷構件的影像特徵，以無損傷、髮絲裂紋以及中至重度損壞的三種區分，或對構件表面剝落、裂縫和鋼筋裸露或挫曲之四種損傷特徵，然而其識別之損傷型式皆不易與構件的殘餘強度對應，或者其識別之損傷程度不易應用於殘餘耐震能力評估。本研究利用既有研究對受損構件各強度下所歸納的損傷特徵為依據，並與營建署目前分級制度對應，透過卷積類神經網路(GoogLeNet)和區域型卷積類神經網路(Faster R-CNN)進行訓練，以識別模型判斷其損傷程度並應用於建築物殘餘的耐震能力初步評估。最後將本研究提出的殘餘耐震能力初步評估法實際應用於國家地震工程研究中心震損資料庫之震損建物，探討此評估方法特性與識別誤差，了解本研究所提出方法之可行性。

Most citizens care about whether the residual seismic capacity of the building is still sufficient after each earthquake. At present, the only way to estimate the residual strength of the building is through the detailed assessment of seismic performance by push over analysis. However, the cost is huge, hence developing a more economical assessment method has become an important topic in recent earthquake engineering research. Based on the research results of earthquake-damaged RC members, this study proposes the residual strength reduction factor and applies it to estimate the residual earthquake resistance of the building after the earthquake using existing preliminary assessment method. Recently, benefit from the progress of computer hardware, deep learning techniques improve substantially. The advent of the Convolution Neural Network (CNN) makes image recognition more feasible. Previous studies had already tried to classify damage levels or patterns of damaged members. However, it is not easy to use the classified results to evaluate the residual seismic capacity. In this study, the damage levels defined by the Construction and Planning Agency are employed and the residual strength factors of the damaged elements of different damage levels estimated based on these damage levels are proposed. And then, CNN (GoogLeNet) and Regional Convolutional Neural Network (Faster R-CNN) are trained to identify damage levels of the element photos taken during earthquake reconnaissance. Preliminary seismic evaluation of damaged RC buildings during several destructive earthquakes are performed to demonstrate the feasibility of the proposed approach, and results using the damage levels identified by the trained neural netorks are compared to the one identified by experts.

1.內政部營建署, 「災害後危險建築物緊急評估辦法」. 2009.
2.日本建築防災協会, 震災建築物の被災度区分判定基準及び復旧技術指針. 2001.
3.Chiu, Chien-Kuo, Hsin-Fang Sung, and Tsung-Chih Chiou, Quantification of the reduction factors of seismic capacity for damaged RC column members using the experiment database. Earthquake Engineering & Structural Dynamics, 2021. 50(3): p. 756-776.
4.姚亭君, 「界面滑移對於震損RC填充牆耐震容量之影響」. 國立臺灣科技大學營建工程學系，碩士論文，台北, 2021.
5.Rou-Ling Yeh, Chien-Chuang Tseng and Hwang Shyh-Jiann, Shear Strength of Reinforced Concrete Vertical Wall Segments under Seismic Loading. ACI Structural Journal. 115(5).
6.Chiou, Tc and Shyh‐Jiann Hwang, Tests on cyclic behavior of reinforced concrete frames with brick infill. Earthquake Engineering & Structural Dynamics, 2015. 44.
7.Chiou, T. C., C. T. Huang, and L. L. Chung, Verification on Seismic Assessment Models of Brick Infill by Tests of RC Frame With Brick infill. Journal of the Chinese Institute of Civil and Hydraulic Engineering, 2017. 29: p. 257-268.
8.Ahmed, F. Hassan and A. Sozen Mete, Seismic Vulnerability Assessment of Low-Rise Buildings in Regions with Infrequent Earthquakes. ACI Structural Journal. 94(1).
9.Maeda, Masaki and Dae Eon Kang, Post-Earthquake Damage Evaluation of Reinforced Concrete Buildings. Journal of Advanced Concrete Technology, 2009. 7(3): p. 327-335.
10.Chiou, T.-C., Hwang, S.-J., Chung, L.-L., Tu, Y.-S., Shen, W.-C. and Weng, P.-W., Preliminary Seismic Assessment of Low-Rise Reinforced Concrete Buildings in Taiwan. 16th World Conference on Earthquake Engineering, 16WCEE 2017, Santiago, Chile, Paper No. 2977, 2017.
11.宋嘉誠, 邱聰智, 黃世建,, 臺灣中小學校舍結構耐震安全柱量比之研究. 國家地震工程研究中心,NCREE-13-031, 2013.
12.林昱衡、邱聰智、黃世建, 鋼筋混凝土住宅建築之耐震能力初步評估法. 中華民國第14屆結構工程及第4屆地震工程研討會 The Fourteenth National Conference on Structural Engineering/ The Fourth National Conference on Earthquake Engineering Taichung, Taiwan, 6-8 November, 2018 Paper No.13013.
13.陳建忠、葉勇凱、翁樸文、邱聰智、沈文成., 震災後危險建築物緊急評估技術及制度之修訂研究. 內政部建築研究所協同報告, 2014.
14.ASCE/SEI-41, Seismic rehabilitation of existing buildings supplement no. 1. American Society of Civil Engineers: Reston.
15.Chuang, Ming-Chieh, Edward Liao, Vicki Pei Lai, Yi-Jer Yu, and Keh-Chyuan Tsai, Development of PISA4SB for Applications in the Taiwan School Building Seismic Retrofit Program. Procedia Engineering, 2011. 14: p. 965-973.
16.ASCE/SEI41-13, seismic evaluation and retrofi t of existing buildings. American Society of Civil Engineers, 2013.
17.Yamaguchi, Tomoyuki and Shuji Hashimoto, Fast crack detection method for large-size concrete surface images using percolation-based image processing. Machine Vision and Applications, 2010. 21(5): p. 797-809.
18.Shan, Baohua, Shijie Zheng, and Jinping Ou, A stereovision-based crack width detection approach for concrete surface assessment. KSCE Journal of Civil Engineering, 2016. 20(2): p. 803-812.
19.Yang, Yuan-Sen, Chung-Ming Yang, and Chang-Wei Huang, Thin crack observation in a reinforced concrete bridge pier test using image processing and analysis. Advances in Engineering Software, 2015. 83.
20.German, Stephanie, Ioannis Brilakis, and Reginald DesRoches, Rapid entropy-based detection and properties measurement of concrete spalling with machine vision for post-earthquake safety assessments. Advanced Engineering Informatics, 2012. 26(4): p. 846-858.
21.Gao, Yuqing and Khalid M. Mosalam, Deep Transfer Learning for Image-Based Structural Damage Recognition. Computer-Aided Civil and Infrastructure Engineering, 2018. 33(9): p. 748-768.
22.Cha, Young-Jin, Wooram Choi, Gahyun Suh, Sadegh Mahmoudkhani, and Oral Buyukozturk, Autonomous Structural Visual Inspection Using Region-Based Deep Learning for Detecting Multiple Damage Types. Computer-Aided Civil and Infrastructure Engineering, 2017. 00: p. 1-17.
23.Hoskere V, Narazaki Y, Hoang T, Spencer Jr B., Vision-based structural inspection using multiscale deep convolutional neural networks. arXiv preprint arXiv:1805.01055, 2018.
24.Ghosh Mondal, Tarutal, Mohammad R. Jahanshahi, Rih-Teng Wu, and Zheng Yi Wu, Deep learning-based multi-class damage detection for autonomous post-disaster reconnaissance. Structural Control and Health Monitoring, 2020. 27(4).
25.Wang, Niannian, Xuefeng Zhao, Peng Zhao, Yang Zhang, Zheng Zou, and Jinping Ou, Automatic damage detection of historic masonry buildings based on mobile deep learning. Automation in Construction, 2019. 103: p. 53-66.
26.Perez, Husein and Joseph H. M. Tah, Deep learning smartphone application for real-time detection of defects in buildings. Structural Control and Health Monitoring, 2021. n/a(n/a): p. e2751.
27.內政部營建署, 建築物耐震設計規範與解說. 內政部, 2011.
28.邱聰智、翁樸文、沈文成、葉勇凱、鍾立來、黃世建, 「 台灣典型街屋建築結構資料庫」. 國家地震工程研究中心，研究報告 No.NCREE-14-006, 台北(2014).
29.Chiou, T. C., Hwang, S. J., Chung, L. L., Tu, Y. S., Shen, W. C., and Weng, P. W., “Preliminary Seismic Assessment of Low-Rise Reinforced Concrete Buildings in Taiwan”. 16WCEE 2017, Santiago, Chile, Paper No. 2977 2017.
30.邱聰智、何郁姍、沈文成、盧亦羽, 「0206 花蓮地震震損建物資料分析與探討」. 中華民國第14屆結構工程及第4屆地震工程研討會The Fourteenth National Conference on Structural Engineering/ The Fourth National Conference on Earthquake Engineering Taichung, Taiwan, 6-8 November, 2018 Paper No. 24007.
31.國家地震工程研究中心、普渡大學, 「20160206美濃地震震損建物資料庫」. https://www.ncree.org/recce/20160206/，臺北, 2018.
32.邱聰智、何郁姍、翁樸文、沈文成, 「20180206花蓮地震震損建物資料庫」. https://www.ncree.org/recce/20180206/，臺北, 2018.
33.普渡大學, Digital Environment for Enabling Data-Driven Science. https://datacenterhub.org/.
34.Werbos, Paul and Paul John, Beyond regression : new tools for prediction and analysis in the behavioral sciences. 1974.
35.Srivastava, Nitish, Geoffrey Hinton, Alex Krizhevsky, Ilya Sutskever, and Ruslan Salakhutdinov, Dropout: A Simple Way to Prevent Neural Networks from Overfitting. Journal of Machine Learning Research, 2014. 15: p. 1929-1958.
36.Christian Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed S., Anguelov, D., Erhan1, D., Vanhoucke V. and Rabinovich, A., Going Deeper with Convolutions. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 1-9. 2015.
37.Girshick, R., J. Donahue, T. Darrell, and J. Malik. Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation. in 2014 IEEE Conference on Computer Vision and Pattern Recognition. 2014.
38.Uijlings, Jasper, K. Sande, T. Gevers, and A. W. M. Smeulders, Selective Search for Object Recognition. International Journal of Computer Vision, 2013. 104: p. 154-171.
39.Ren, Shaoqing, Kaiming He, Ross Girshick, and Jian Sun, Faster R-CNN: towards real-time object detection with region proposal networks, in Proceedings of the 28th International Conference on Neural Information Processing Systems - Volume 1. 2015, MIT Press: Montreal, Canada. p. 91–99.
40.Lin, Tsung-Yi, Piotr Dollar, Ross Girshick, Kaiming He, Bharath Hariharan, and Serge Belongie, Feature Pyramid Networks for Object Detection. 2017. 936-944.
41.He, Kaiming, Xiangyu Zhang, Shaoqing Ren, and Jian Sun, Deep Residual Learning for Image Recognition. 2016. 770-778.
42.He, K., G. Gkioxari, P. Dollár, and R. Girshick. Mask R-CNN. in 2017 IEEE International Conference on Computer Vision (ICCV). 2017.
43.Pan, S. J., Yang, Q, A survey on transfer learning. IEEE Transactions on knowledge and data engineering. 22(10), pp. 1345-1359., 2010.
44.Github.Available, https://github.com/wkentaro/labelme.
45.Lin, Tsung-Yi, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Dollár, and C. Lawrence Zitnick. Microsoft COCO: Common Objects in Context. in Computer Vision – ECCV 2014. 2014. Cham: Springer International Publishing.
46.蘇耕立, 台灣中小學校舍結構耐震能力初步評估方法之探討. 國立臺灣大學土木工程學系碩士論文, 2008.