耐震初步評估為一可快速診斷建物耐震性能之方法。過去我國研發的耐震初步評估方法，需透過人工讀取工程藍圖中重要資訊，如相關構件之截面積，但該步驟相當耗費時間與人力成本。另外，雖數位化圖檔可配合影像辨識技術，直接提取重要構件之截面積，但這些技術應用於工程藍圖仍會面臨各種不同的挑戰。因此，本研究研發一套基於電腦視覺與深度學習方法之影像辨識方法，由工程藍圖中提取建物重要構件之截面積，結合過去研發之耐震初步評估方法，降低人工判讀圖資之時間並減少人為失誤之可能性。本研究所提出的方法，首先，透過影像辨識的方式，找出結構圖或建築圖中之結構柱並相互套疊，於建築圖中的結構柱位置，自動化框選出圖面中之牆體，再透過人工智慧深度學習，分出不同類型的牆體，之後由本研究應用影像處理技術，提出牆體長度及厚度計算方法，可擷取牆截面積的關鍵參數，最後藉由柱、牆之類型與截面積，可套入過去研發之耐震初步評估計算方法，即可得出耐震初步評估分數。目前已對72棟建物進行此方法之評估，正確預測的準確率達87%，符合工程應用的信心程度可達95%。

The preliminary seismic assessment is a method for rapidly diagnosing the seismic performance of buildings. In the past, seismic preliminary assessment methods developed in our country required the manual extraction of important information from engineering blueprints, such as the cross-sectional areas of relevant components. However, this step is quite time-consuming and labor-intensive. Furthermore, while digital image files can be processed with image recognition technology to directly extract the cross-sectional areas of important components, applying these techniques to engineering blueprints still presents various challenges. Therefore, this study has developed a computer vision and deep learning-based image recognition method to extract the cross-sectional areas of critical building components from engineering blueprints. This method, when combined with previously developed seismic preliminary assessment methods, reduces the time required for manual interpretation of drawings and minimizes the potential for human errors. The method proposed in this study first identifies structural columns in structural or architectural drawings through image recognition, automatically outlines the walls in the architectural drawings by overlaying the positions of structural columns, uses artificial intelligence and deep learning to classify different types of walls, and then applies image processing techniques to propose methods for calculating wall lengths and thicknesses. These methods extract key parameters for wall cross-sectional areas. Finally, by incorporating the types and cross-sectional areas of columns and walls, the previously developed seismic preliminary assessment calculation method can be applied to obtain a seismic preliminary assessment score. This method has already been applied to assess 72 buildings, with a confidence level suitable for engineering applications of up to 95%.

1.Lowe, D. G. (1966). The recognition of speech using template techniques. Proceedings of the IEEE, 54(4), 555-560.
2.許丁友等，「國民中小學典型校舍耐震能力初步評估法」，國家地震工程研究中心，研究報告No. NCREE-03-049，臺北(2003)。
3.Chiou TC, Hwang SJ, Chung LL et al (2017) Preliminary seismic assessment of low-rise reinforced concrete buildings in Taiwan.In: 16th World conference on earthquake engineering, Santiago, Chile. Paper No. 2977.
4.林煜衡，「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 ，臺北(2018)。
5.F. Hassan and M. A. Sozen, Seismic Vulnerability Assessment of Low-Rise Buildings in Regions with Infrequent Earthquakes, ACI Structural J., Vol.94, Issue 1, pp. 31-39, 1997.
6.O’Brien, P.; Eberhard, M.; Haraldsson, O.; Irfanoglu, A.; Lattanzi, D.; Lauer, S.; Pujol, S. Measures of the Seismic Vulnerability of Reinforced Concrete Buildings in Haiti. Earthquake. Spectra 2011, 27, S373–S386.
7.S. Pujol, L. Laughery, A. Puranam, P. Hesam, L. Cheng, A. Lund, and A. Irfanoglu, Evaluation of Seismic Vulnerability Indices for Low-Rise Reinforced Concrete Buildings Including Data from the 6 February 2016 Taiwan Earthquake, Journal of Disaster Research, Vol.15, No.1, pp. 9-19, 2020.
8.Shiga T, Shibata A, Takahashi T (1968) Earthquake damage and wall index of reinforced concrete buildings. In: Proceedings, Tohuku district symposium, Architectural Institute of Japan, no. 12, Dec. 1968, pp 29–32 (in Japanese)
9.Islam MS, Alwashali H, Torihata Y, Jin K, Maeda M (2017) Rapid seismic capacity evaluation method of RC buildings with masonry infll. In: Architectural Institute of Japan-Academic Lecture Collection Summary, August 2017
10.Puranam AY, Irfanoglu A, Pujol S, Chiou T, Hwang S (2019) Evaluation of seismic vulnerability screening indices using data from the Taiwan Earthquake of 6 February 2016. Bulletin of Earthquake Engineering.
11.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年。
12.內政部營建署，「建築物耐震設計規範與解說」，內政部，臺北 (2011)。
13.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，臺北，2013。
14.Harirchian E, Kumari V, Jadhav K, Rasulzade S, Lahmer T, Raj Das R (2021) A synthesized study based on machine learning approaches for rapid classifying earthquake damage grades to RC buildings. Applied Sciences 11(16): 7540.
15.de Lautour, O. R., and P. Omenzetter. 2009.Prediction of seismic-induced structural damage using artificial neural networks. Engineering structural.31 (2).
16.S Tesfamaraim, M Saatcioglu.Seismic risk assessment of RC buildings using fuzzy synthetic evaluationJournal of Earthquake Engineering, 12 (7) (2008).
17.Khaloo, A.; Lattanzi, D. Extracting structural models through computer vision. In Proceedings of the 2015 Structures Congress, Portland, OR, USA, 23–25 April 2015.
18.Aoki T, Ceravolo R, De Stefano A, Genovese C and Sabia D (2002) Seismic vulnerability assessment of chemical plants through probabilistic neural networks. Reliability Engineering and System Safety 77(3): 263–268.
19.Agatonovic-Kustrin, S. & Beresford, R. Basic concepts of artificial neural network (ANN) modeling and its application in pharmaceutical research. J. Pharm. Biomed. Anal. 22, 717–727 (2000).
20.Hirasawa, K.; Ohbayashi, M.; Koga, M.; Harada, M. Forward propagation universal learning network. In Proceedings of the International Conference on Neural Networks (ICNN’96), Washington, DC, USA, 3–6 June 1996; IEEE: Piscataway, NJ, USA, 1996; Volume 1, pp. 353–358.
21.Hochreiter, S. The vanishing gradient problem during learning recurrent neural nets and problem solutions. Int. J. Uncertain., Fuzziness Knowl.-Based Syst. 6, 107–116 (1998).
22.Maxwell, J. C. (1861). On the Theory of Three Primary Colours. Philosophical Transactions of the Royal Society of London, 151.
23.H. M. Bui, M. Lech, E. Cheng, K. Neville, and I. S. Burnett, Using grayscale images for object recognition with convolutional-recursive neural network, in Proc. IEEE 6th Int. Conf. Commun. Electron. (ICCE), Jul. 2016, pp. 321–325.
24.A. M. Naser, Color to grayscale image conversion based dimensionality reduction with stationary wavelet transform, in Proc. Al-Sadeq Int. Conf. Multidisciplinary IT Commun. Sci. Appl. (AIC-MITCSA), 2016.
25.Acharya and Ray, Image Processing: Principles and Applications, Wiley-Interscience 2005 ISBN 0471719986.
26.Pizer, S. M., E. P. Amburn, J. D. Austin, R. Cromartie, A. Geselowitz, T. Greer, B. T. H. Romeny, J. B. Zimmerman, and K. Zuiderveld (1987), Adaptive histogram equalization and its variations, Comput. Vision Graphics Image Process., 39(3), 355– 368, doi:10.1016/S0734-189X(87)80186-X.
27.Brunelli R (ed) (2009) Template matching techniques in computer vision: theory and practice, 1st edn. Wiley, New York.
28.A.W. Lohmann, (1993), Image rotation, wigner rotation, and the fractional fourier transform, J Opt Soc Am A.
29.P.F. Felzenszwalb, R.B. Girshick, D. McAllester, D. Ramanan, Object detection
with discriminatively trained part-based models, IEEE Trans. Pattern Anal.
Mach. Intell. 32 (9) (2010) 1627–1645, https://doi.org/10.1109/TPAMI.2009.
167
30.E. Rezende, G. Ruppert, T. Carvalho, F. Ramos, and P. De Geus, Malicious software classification using transfer learning of ResNet50 deep neural network, in Proc. 16th IEEE Int. Conf. Mach. Learn. Appl. (ICMLA), 2017.
31.Z. Xu, K. Sun, and J. Mao, Research on ResNet101 network chemical reagent label image classification based on transfer learning, in Proc. IEEE 2nd Int. Conf. Civil Aviation Saf. Inf. Technol. (ICCASIT, Oct. 2020, pp. 354–358, doi: 10.1109/ICCASIT50869.2020.9368658.
32.Liu S, Tian G, Xu Y (2019) A novel scene classification model combining ResNet based transfer learning and data augmentation with a filter. Neurocomputing 338:191–206
33.Krizhevsky A, Sutskever I, Hinton GE. Imagenet classification with deep convolutional neural networks. In: Advances in neural information processing systems. 2012; p. 1097–105.
34.Singarimbun, R.N.; Nababan, E.B.; Sitompul, O.S. Adaptive Moment Estimation To Minimize Square Error In Backpropagation Algorithm. In Proceedings of the 2019 International Conference of Computer Science and Information Technology (ICoSNIKOM), Medan, Indonesia, 28–29 November 2019.
35.吳青峰，基於深度學習之「RC建築震後初步耐震能力評估」，國立臺灣科技大學營建工程學系碩士論文，2020。