Post-Disaster Building Damage Segmentation Using Convolutional Neural Networks
##plugins.themes.bootstrap3.article.main##
Abstract
Natural disasters are events caused by nature such as earthquakes, tornadoes, tsunamis, forest fires, and others. The impacts of natural disasters are significant and varied across various sectors, including the economy, health, and primarily, infrastructure. Effective and efficient actions are needed to assist in the recovery following natural disasters, one of which is aiding in the identification of building damage levels post-disaster. To address this issue, this research proposes a system capable of performing segmentation to determine the level of building damage post-natural disaster using convolutional neural network methods. The data utilized consists of aerial images sourced from xView2: Assess Building Damage, comprising 50 aerial images with 5 classes: no-damage, minor-damage, major-damage, destroyed, and unlabeled. The steps undertaken in this research include data preprocessing using patchify and data augmentation. Subsequently, feature extraction is performed using convolution, followed by the training process using a neural network with the proposed architecture. This study proposes an architecture with 27 hidden layers, with feature extraction utilizing average pooling. The model evaluation process will employ Mean Intersection over Union (MIoU) to assess how closely the segmentation prediction results resemble the original data. The proposed architecture demonstrates the best MIoU result with a value of 0.31 and an accuracy of 0.9577.
##plugins.themes.bootstrap3.article.details##
[2] J. Padli, M. Shah Habibullah, and A. H. Baharom, “Economic impact of natural disasters’ fatalities,” Int J Soc Econ, vol. 37, no. 6, pp. 429–441, May 2010, doi: 10.1108/03068291011042319.
[3] A. Pramono, I. W. S. Mahendra, I. B. A. Wijaya, I. A. Agustina, and R. T. Herman, “Diagnosis and Repair of the Cracking House Next to the River,” IOP Conf Ser Earth Environ Sci, vol. 998, no. 1, p. 012002, Feb. 2022, doi: 10.1088/1755-1315/998/1/012002.
[4] E. H. Zaryabi, B. Kalantar, L. Moradi, A. A. Halin, and N. Ueda, “MSBDA-Net: Multi-scale Siamese Building Damage Assessment Network,” in 2022 IEEE Asia-Pacific Conference on Computer Science and Data Engineering (CSDE), IEEE, Dec. 2022, pp. 1–6. doi: 10.1109/CSDE56538.2022.10089353.
[5] P. Pahlavani, F. Samadzadegan, and M. R. Delavar, “A GIS-Based Approach for Urban Multi-criteria Quasi Optimized Route Guidance by Considering Unspecified Site Satisfaction,” 2006, pp. 287–303. doi: 10.1007/11863939_19.
[6] D. S. Deswita Indriani, E. Juni Arta Sinaga, G. Oktavia, H. Syahputra, F. Ramadhani, and I. Komputer, “Identifikasi Tanda Tangan Dengan Menggunakan Metode Convolution Neural Network (CNN)”.
[7] A. T. W. Almais et al., “SASSD: A Smart Assessment System For Sector Damage Post-Natural Disaster Using Artificial Neural Networks,” in 2023 2nd International Conference on Computer System, Information Technology, and Electrical Engineering (COSITE), IEEE, Aug. 2023, pp. 96–101. doi: 10.1109/COSITE60233.2023.10249540.
[8] Moch. N. Annafii, O. V. Putra, T. Harmini, and N. Trisnaningrum, “Segmentasi Semantik pada Citra Hama Leafblast Menggunakan Unet dan Optimasi Hyperband,” Prosiding Sains Nasional dan Teknologi, vol. 12, no. 1, pp. 453–459, Nov. 2022, doi: 10.36499/psnst.v12i1.7230.
[9] I. Kotaridis and M. Lazaridou, “SEMANTIC SEGMENTATION USING A UNET ARCHITECTURE ON SENTINEL-2 DATA,” The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. XLIII-B3-2022, pp. 119–126, May 2022, doi: 10.5194/isprs-archives-XLIII-B3-2022-119-2022.
[10] A. Di Benedetto, M. Fiani, and L. M. Gujski, “U-Net-Based CNN Architecture for Road Crack Segmentation,” Infrastructures (Basel), vol. 8, no. 5, p. 90, May 2023, doi: 10.3390/infrastructures8050090.
[11] S. Gangurde, “Building and Road Segmentation Using EffUNet and Transfer Learning Approach,” Jul. 2023.
[12] W. Li, J. Wu, H. Chen, Y. Wang, Y. Jia, and G. Gui, “UNet Combined With Attention Mechanism Method for Extracting Flood Submerged Range,” IEEE J Sel Top Appl Earth Obs Remote Sens, vol. 15, pp. 6588–6597, 2022, doi: 10.1109/JSTARS.2022.3194375.
[13] Y. Li et al., “CTMU-Net: An Improved U-Net for Semantic Segmentation of Remote-Sensing Images Based on the Combined Attention Mechanism,” IEEE J Sel Top Appl Earth Obs Remote Sens, vol. 16, pp. 10148–10161, 2023, doi: 10.1109/JSTARS.2023.3326960.
[14] D. Varma, A. Nehansh, and P. Swathy, “Data Preprocessing Toolkit : An Approach to Automate Data Preprocessing,” INTERANTIONAL JOURNAL OF SCIENTIFIC RESEARCH IN ENGINEERING AND MANAGEMENT, vol. 07, no. 03, Mar. 2023, doi: 10.55041/IJSREM18270.
[15] Z. Chen et al., “DPT: Deformable Patch-based Transformer for Visual Recognition,” Jul. 2021, doi: 10.1145/3474085.3475467.
[16] S. Yang, W. Xiao, M. Zhang, S. Guo, J. Zhao, and F. Shen, “Image Data Augmentation for Deep Learning: A Survey,” Apr. 2022.
[17] W. Gao, “Investigation of multiple convolutional neural network models on emotion detection,” Applied and Computational Engineering, vol. 22, no. 1, pp. 35–41, Oct. 2023, doi: 10.54254/2755-2721/22/20231164.
[18] Y. Tzach et al., “The mechanism underlying successful deep learning,” May 2023.
[19] R. Scodellaro, A. Kulkarni, F. Alves, and M. Schröter, “Training Convolutional Neural Networks with the Forward-Forward algorithm,” Dec. 2023.
[20] M. S. Tuna and A. Kristianto, “Klasifikasi Cuaca Berbasis Citra dengan Model CNN LeNet-5 yang Dimodifikasi,” J-Intech : Journal of Information and Technology, no. 204, pp. 401–410, 2022.
[21] H. Jin, “Hyperparameter Importance for Machine Learning Algorithms,” Jan. 2022, [Online]. Available: http://arxiv.org/abs/2201.05132
[22] M. Z. Khan, M. K. Gajendran, Y. Lee, and M. A. Khan, “Deep Neural Architectures for Medical Image Semantic Segmentation: Review,” IEEE Access, vol. 9, pp. 83002–83024, 2021, doi: 10.1109/ACCESS.2021.3086530.