Aiming to support disaster recovery, we have developed a new method to extract damaged buildings by using machine learning that combines 10 explanatory variables obtained from analysis of aerial photographs and observation data. We used site amplification factors, seismic intensities of foreshock and mainshock, distance from faults, estimated building structures and ages, coverage by blue tarps, texture analysis, and digital surface model differences before and after the earthquake as explanatory variables, in addition to convolutional neural network prediction results based on post-earthquake aerial photographs. The random forest method resulted in an overall accuracy of about 81% and an average F-measure of three classes was about 70%, indicating that it can classify possible damage to buildings more accurately than our previous studies.

1. [1] S. Naito, H. Tomozawa, Y. Mori, T. Nagata, N. Monma, H. Nakamura, H. Fujiwara, and G. Shoji, “Building-damage detection method based on machine learning utilizing aerial photographs of the Kumamoto earthquake,” Earthquake Spectra, Vol.36, No.3, pp. 1166-1187, 2020. https://doi.org/10.1177/8755293019901309
2. [2] S. Naito, H. Tomozawa, Y. Mori, N. Monma, H. Nakamura, and H. Fujiwara, “Development of the deep learning based damage detection model for buildings utilizing aerial photographs of multiple earthquakes,” J. of Japan Association for Earthquake Engineering, Vol.21, Issue 3, pp. 72-118, 2021. https://doi.org/10.5610/jaee.21.3_72
3. [3] H. Miura, A. Tomohiro, and M. Matsuoka, “Deep learning-based identification of collapsed, non-collapsed and blue tarp-covered buildings from post-disaster aerial images,” Remote Sensing, Vol.12, No.12, Article No.1924, 2020. https://doi.org/10.3390/rs12121924
4. [4] B. Kalantar, N. Ueda, H. A. Al-Najjar, and A. A. Halin, “Assessment of convolutional neural network architectures for earthquake-induced building damage detection based on pre- and post-event orthophoto images,” Remote Sensing, Vol.12, No.21, Article No.3529, 2020. https://doi.org/10.3390/rs12213529
5. [5] S. Fujita and M. Hatayama, “Rapid and accurate detection of building damage investigation using automatic method to calculate roof damage rate,” IDRiM J., Vol.12, No.1, pp. 89-111, 2022. https://doi.org/10.5595/001c.35885
6. [6] O. R. Sodeinde, M. Koch, B. Moaveni, and L. G. Baise, “One versus all: Identifiability with a multi-hazard and multiclass building damage imagery dataset and a deep learning neural network,” Natural Hazards, Vol.120, pp. 8337-8366, 2024. https://doi.org/10.1007/s11069-024-06500-9
7. [7] H. Xiu, T. Shinohara, M. Matsuoka, M. Inoguchi, K. Kawabe, and K. Horie, “Collapsed building detection using 3D point clouds and deep learning,” Remote Sensing, Vol.12, No.24, Article No.4057, 2020. https://doi.org/10.3390/rs12244057
8. [8] F. Yamazaki, W. Liu, and K. Horie, “Use of multi-temporal LiDAR data to extract collapsed buildings and to monitor their removal process after the 2016 Kumamoto Earthquake,” Remote Sensing, Vol.14, No.23, Article No.5970, 2022. https://doi.org/10.3390/rs14235970
9. [9] Y. Zhan, W. Liu, and Y. Maruyama, “Damaged building extraction using modified mask R-CNN model using post-event aerial images of the 2016 Kumamoto Earthquake,” Remote Sensing, Vol.14, No.4, Article No.1002, 2022. https://doi.org/10.3390/rs14041002
10. [10] L. Deng and Y. Wang, “Post-disaster building damage assessment based on improved U-Net,” Scientific Reports, Vol.12, No.1, Article No.15862, 2022. https://doi.org/10.1038/s41598-022-20114-w
11. [11] C. Geiß, P. A. Pelizari, M. Marconcini, W. Sengara, M. Edwards, T. Lakes, and H. Taubenbock, “Estimation of seismic building structural types using multi-sensor remote sensing and machine learning techniques,” ISPRS J. of Photogrammetry and Remote Sensing, Vol.140, pp. 175-188, 2015. https://doi.org/10.1016/j.isprsjprs.2014.07.016
12. [12] M. Ji, L. Liu, R. Du, and M. F. Buchroithner, “A comparative study of texture and convolutional neural network features for detecting collapsed buildings after earthquakes using pre- and post-event satellite imagery,” Remote Sensing, Vol.11, No.11, Article No.1202, 2019. https://doi.org/10.3390/rs11101202
13. [13] L. Moya, E. Mas, S. Koshimura, F. Yamazaki, W. Liu, and C. Geiß, “Fusion of modeling and remote sensing for damage estimation,” 17th World Conf. on Earthquake Engineering (17WCEE), No.6a-0004, 2020.
14. [14] Disaster Management in Japan, Cabinet Office Japan (in Japanese). https://www.bousai.go.jp/updates/h280414jishin/pdf/h280414jishin_55.pdf [Accessed December 26, 2023]
15. [15] Geospatial Information Authority of Japan (in Japanese). https://fgd.gsi.go.jp/download/menu.php [Accessed December 26, 2023]
16. [16] S. Naito, N. Monma, H. Nakamura, H. Fujiwara, H. Shimomura, and T. Yamada, “Investigation of building damages caused by the 2016 Kumamoto Earthquake utilizing aerial photographic interpretation,” J. of Japan Society of Civil Engineers, Ser.A1 (Structural Engineering & Earthquake Engineering), Vol.74, No.4, pp. 464-480, 2018 (in Japanese). https://doi.org/10.2208/jscejseee.74.I_464
17. [17] S. Naito, K. X. Hao, S. Senna, T. Saeki, H. Nakamura, H. Fujiwara, and T. Azuma, “Investigation of damages in immediate vicinity of co-seismic faults during the 2016 Kumamoto Earthquake,” J. Disaster Res., Vol.12, No.5, pp. 899-915, 2017. https://doi.org/10.20965/jdr.2017.p0899
18. [18] National Institute for Land Infrastructure Management, “Quick report of the field survey on the building damage by the 2016 Kumamoto Earthquake,” Technical Note of National Institute for Land and Infrastructure Management, No.929, Building Research Data, No.173, 2016 (in Japanese).
19. [19] S. Okada and N. Takai, “Classification of structural types and damage patterns of buildings for earthquake field investigation,” Proc. of the 12th World Conf. on Earthquake Engineering (12WCEE), Article No.705, 2000.
20. [20] National Research Institute for Earth Science and Disaster Resilience, “J-SHIS.” https://www.j-shis.bosai.go.jp/en/ [Accessed December 26, 2023]
21. [21] K. Wakamatsu and M. Matsuoka, “Developing a 7.5-sec site-condition map for Japan based on geomorphologic classification,” Earthquake Resistant Engineering Structures, WIT Trans. on the Built Environment, Vol.120, pp. 101-112, 2011. https://doi.org/10.2495/ERES110091
22. [22] National Research Institute for Earth Science and Disaster Resilience, “J-RISQ.” https://www.j-risq.bosai.go.jp/report/en/ [Accessed December 26, 2023]
23. [23] Y. Kumahara and Research Group of Inter-University, “The characteristics of surface rupture associated by the 2016 Kumamoto Earthquake,” Programme and Abstracts, Japan Society for Active Fault Studies, Fall Meeting, pp. 76-77, 2016 (in Japanese).
24. [24] S. Naito, N. Monma, T. Yamada, H. Shimomura, K. Mochizuki, Y. Honda, H. Nakamura, H. Fujiwara, and G. Shoji, “Building damage detection of the Kumamoto Earthquake utilizing image analyzing methods with aerial photographs,” J. of Japan Society of Civil Engineers, Ser.A1 (Structural Engineering & Earthquake Engineering), Vol.75, Issue 4, pp. 218-237, 2019 (in Japanese). https://doi.org/10.2208/jscejseee.75.I_218
25. [25] L. Moya, E. Mas, F. Yamazaki, W. Liu, and S. Koshimura, “Statistical analysis of earthquake debris extent from wood-frame buildings and its use in road networks in Japan,” Earthquake Spectra, Vol.36, No.1, pp. 209-231, 2020. https://doi.org/10.1177/8755293019892423
26. [26] R. M. Hararic, K. Shanmugam, and I. Dinstein, “Textural features for image classification,” IEEE Trans. of Systems, Man and Cybernetics, Vol.SMC-3, Issue 6, pp. 610-621, 1973. https://doi.org/10.1109/TSMC.1973.4309314
27. [27] https://scikit-learn.org/stable/modules/tree.html [Accessed December 26, 2023]
28. [28] L. Breiman, “Random forests,” Machine Learning, Vol.45, No.1, pp. 5-32, 2001. https://doi.org/10.1023/A:1010933404324
29. [29] Y. Freund and R. E. Schapire, “Experiments with a new boosting algorithm,” Proc. of the 13th Int. Conf. on Int. Conf. on Machine Learning (ICML’96), pp. 148-156, 1996.
30. [30] G. Ke, Q. Meng, T. Finley, T. Wang, W. Chen, W. Ma, Q. Ye, and T. Y. Liu, “LightGBM: A highly efficient gradient boosting decision tree,” 31st Conf. on Neural Information Processing Systems (NIPS’2017), 2017.
31. [31] T. Chen and C. Guestrin, “XGBoost: A scalable tree boosting system,” Proc. of the 22nd ACM SIGKDD Int. Conf. on Knowledge Discovery and Data Mining, pp. 785-794, 2016. https://doi.org/10.1145/2939672.2939785
32. [32] https://scikit-learn.org/stable/modules/permutation_importance.html [Accessed January 16, 2024]
33. [33] C. Tomasi and T. Kanade, “Shape and motion from image streams under orthography: A factorization method,” Int. J. of Computer Vision, Vol.9, No.2, pp. 137-154, 1992. https://doi.org/10.1007/BF00129684