ABSTRACT
After any disaster, the Government rehabilitates the victims based on the severity of the damage caused to their properties. Since a huge number of rehabilitation claims flow in after the disaster, it takes up a lot of manual labor in inspecting and validating the claims along with deciding the amount of rehabilitation to be granted. Moreover, such manual inspection leads to a lack of transparency. In recent years, social media posts, text, and images have become a rich source of post-disaster situational information that may be useful in assessing damage at a low cost. Most of the existing research explores the use of social media text for extracting situational information useful for disaster response. The usage of social media images to assess disaster damage is limited. In this paper, we propose a convolutional neural network-based approach to locate damage in a disaster image and to quantify the degree of the damage. The proposed damage assessment system categorizes images of earthquake-affected buildings and decides the severity of the damage caused by the earthquake. Our proposed approach enables the use of social media images for post-disaster damage assessment and provides an inexpensive and feasible alternative to the more expensive GIS approach. Our approach exhibits high accuracy in classifying earthquake-affected buildings and determining the severity of damage at a negligible loss.
- S. Basu, S. Roy, S. Bandyopadhyay and S. Das Bit, "A Utility Driven Post Disaster Emergency Resource Allocation System Using DTN," in IEEE Transactions on Systems, Man, and Cybernetics: Systems. doi: 10.1109/TSMC.2018.2813008Google Scholar
- Y. Fan, Q. Wen, W. Wang, P. Wang, L. Li, and P. Zhang, "Quantifying disaster physical damage using remote sensing data technical work flow and case study of the 2014 ludian earthquake in china," International Journal of Disaster Risk Science, vol. 8, no. 4, pp. 471--488, 2017.Google ScholarCross Ref
- A. Das, N. Mallik, S. Bandyopadhyay, S. Das Bit and J. Basak, "Interactive information crowdsourcing for disaster management using SMS and Twitter: A research prototype," 2016 IEEE International Conference on Pervasive Computing and Communication Workshops (PerCom Workshops), Sydney, NSW, 2016, pp. 1--6. doi: 10.1109/PERCOMW.2016.7457101Google Scholar
- A.A. Hamed and A. Hossein, "CNN-based estimation of pre- and post-earthquake height models from single optical images for identification of collapsed buildings", Remote Sensing Letters, vol 10, no 7, pp. 679--688, Jan 2019.Google ScholarCross Ref
- M. Ji, L. Liu, and M. Buchrithner, "Identifying Collapsed Buildings Using Post-Earthquake Satellite Imagery and Convolutional Neural Networks: A Case Study of the 2010 Haiti Earthquake", Remote Sensing, Oct 2018Google Scholar
- M. Ji, L. Liu, and M. Buchrithner, "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, May 2019Google Scholar
- D. Duarte, F. Nex, N. Kerle, and G. Vosselman, "Multi-Resolution Feature Fusion for Image Classification of Building Damages with Convolutional Neural Networks", Remote Sensings, Oct 2018Google ScholarCross Ref
- M. Mateen, J. Wen, Dr. Nasrullah, S. Song, and Z. Huang "Fundus image classification using VGG-19 architecture with PCA and SVD", Symmetry, Dec 2018.Google Scholar
- P. Marcelino, 2018, Transfer Learning from pre-trained model, Towards Data Science, DOI:https://towardsdatascience.com/transfer-learning-from-pre-trained-models-f2393f124751Google Scholar
- L. Gueguen and R. Hamid, "Large-scale damage detection using satellite imagery," in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 1321--1328, 2015.Google Scholar
- K. Simonyan and A. Zisserman, "Very deep convolutional networks for large-scale image recognition", arXiv preprint arXiv:1409.1556, 2014.Google Scholar
- K. R. Nia and G. Mori, "Building damage assessment using deep learning and ground-level image data", in 14th Conference on Computer and Robot Vision (CRV). IEEE, 2017, pp. 95--102.Google ScholarCross Ref
- K. Gopalakrishnan, S.K. Khaitan, A. Choudhary, A. Agarwal, "Deep Convolutional Neural Networks with transfer learning for computer vision-based data-driven pavement distress detection", Elsevier, Sept 2017Google Scholar
- Facebook, 2019, Facebook, Retrieved from: https://www.facebook.com/search/str/earthquake/photos-keyword?f=AboGO9ioyXWyLH3sDGxCf5_KeijqtmcdccuNUqhLuXFBzS5folEFOO0BkOl8HQcDdctaYdBlQBl6hiMKoFGSftGr2drGCxoaD71TufYIl0p5YX-M1ShwRmGexQOOeM5Rm0c&epa=SEE_MORE.Google Scholar
- Srijan Nag, 2019, Google Drive, Retrieved from: https://drive.google.com/drive/folders/1yVLScW6cd4zzJzLZZ6VlY4A-JToc9vT-?usp=sharingGoogle Scholar
- A. Chattopadhyay, A. Sarkar, P. Howlader, and V. N. Balasubramanian, "Grad-cam++: Generalized gradient-based visual explanations for deep convolutional networks," CoRR, vol. abs/1710.11063, 2017.Google Scholar
- Jacob Gildenblat, 2017, Class activation maps in Keras for visualizing where deep learning networks pay attention, DOI:https://jacobgil.github.io/deeplearning/class-activation-mapsGoogle Scholar
- R. Banerjee, S. Chatterjee and S. DasBit, "An Energy Saving Audio Compression Scheme for WMSN using Spatio-temporal Partial DWT", Journal of Computer and Electrical Engineering, Elsevier Science, vol 48, pp 389--404, 2015.Google ScholarDigital Library
Index Terms
- CNN Based Approach for Post Disaster Damage Assessment
Recommendations
Social media crowdsourcing for rapid damage assessment following a sudden-onset natural hazard event
Highlights- Develop a crowdsourcing approach using Twitter data for rapid earthquake damage assessment.
AbstractRapid appraisal of damages related to hazard events is of importance to first responders, government agencies, insurance industries, and other private and public organizations. While satellite monitoring, ground-based sensor systems, ...
Disaster damage assessment based on fine-grained topics in social media
AbstractSocial media data have been widely used to enrich human-centric information for situational awareness and disaster assessment. Owing to the granularity of topics detected from disaster-related contents, the effectiveness of social ...
Highlights- A new semantic method is used to represent topics at a fine granularity level.
- ...
Natural Disaster Building Damage Assessment Using a Two-Encoder U-Net
Advances in Visual ComputingAbstractWhen a natural disaster occurs, damaged regions rely on timely damage assessments to receive relief. Currently, this is a slow and laborious process, during which emergency response groups conduct on-the-ground evaluations to form fiscal ...
Comments