Deep Learning-Based Detection of Damaged Buildings in Satellite Imagery

Fixation, Monitoring & Assessment of War Consequences and Post-War Reconstruction (NEW)

Authors

First and Last Name Academic degree E-mail Affiliation
Tetiana Kozlova No tanko220703 [at] gmail.com National Aviation University
Kyiv, Ukraine
Yuri Velikodsky Ph.D. yuri.velikodsky [at] gmail.com National Aviation University
Kyiv, Ukraine

I and my co-authors (if any) authorize the use of the Paper in accordance with the Creative Commons CC BY license

First published on this website: 25.08.2024 - 17:25
Abstract 

A deep learning model based on the ResNet50 architecture was developed to detect damaged buildings on satellite images of Mariupol acquired in 2024, with an estimated accuracy of 98.7% on test data and 95.7% on real data. The model was applied to classify damaged buildings in the Left Bank district of Mariupol, where significant destruction was found, especially in the western part of the district, where numerous buildings were destroyed or severely damaged. To facilitate the analysis of the results, a thematic map was created to illustrate the location and condition of the classified buildings. This map was used to visualise the data and improve understanding of the extent of the damage.

Keywords 
Deep Learning
building
damage
satellite imagery
GIS
ArcCIS
CNN
Python
TensorFlow
Keras
References 
  1. Alem, A., & Kumar, S. (2020). Deep learning methods for land cover and land use classification in remote sensing: A review. In 2020 8th International Conference on Reliability, Infocom Technologies and Optimization (Trends and Future Directions, ICRITO), 903-908. IEEE.
  2. Chicho, B. T., & Sallow, A. B. (2021). A comprehensive survey of deep learning models based on Keras framework. Journal of Soft Computing and Data Mining, 2(2), 49-62.
  3. Deng, J., Dong, W., Socher. R., Li, L.-J., Li, K., & Fei-Fei, L. (2009). ImageNet: A Large-Scale Hierarchical Image Database. IEEE Computer Vision and Pattern Recognition (CVPR).
  4. Duckham, M., Sun, Q. C., & Worboys, M. F. (2023). GIS: a computing perspective. CRC press.
  5. Lalitha, V., & Latha, B. (2022). A review on remote sensing imagery augmentation using deep learning. Materials Today: Proceedings, 62, 4772-4778.
  6. Mascarenhas, S., & Agarwal, M. (2021). A comparison between VGG16, VGG19 and ResNet50 architecture frameworks for Image Classification. In 2021 International conference on disruptive technologies for multi-disciplinary research and applications (CENTCON), 1, 96-99. IEEE.
  7. Matin, S. S., & Pradhan, B. (2022). Challenges and limitations of earthquake-induced building damage mapping techniques using remote sensing images-A systematic review. Geocarto International, 37(21), 6186-6212.
  8. Mohan, S., & Giridhar, M. V. S. S. (2022). A brief review of recent developments in the integration of deep learning with GIS. Geomatics and Environmental Engineering, 16(2), 21-38.
  9. Neupane, B., Horanont, T., & Aryal, J. (2021). Deep learning-based semantic segmentation of urban features in satellite images: A review and meta-analysis. Remote Sensing, 13(4), 808.
  10. Pang, B., Nijkamp, E., & Wu, Y. N. (2020). Deep learning with tensorflow: A review. Journal of Educational and Behavioral Statistics, 45(2), 227-248.
  11. Raihan, A. (2023). A comprehensive review of the recent advancement in integrating deep learning with geographic information systems. Research Briefs on Information and Communication Technology Evolution, 9, 98-115.
  12. Sharifani, K., & Amini, M. (2023). Machine learning and deep learning: A review of methods and applications. World Information Technology and Engineering Journal, 10(07), 3897-3904.