Fully convolutional neural networks on semantic segmentation of flooded areas
Resumo
Inundações causam grandes danos econômicos e perdas de vidas em todo o mundo. Assim, a detecção automática de imagens é valiosa para minimizar efetivamente o tempo de resposta a esses impactos. A tecnologia de Radar de Abertura Sintética (SAR) é crucial na gestão de inundações, sendo muito sensível à água. Este estudo usa Redes Neurais Totalmente Convolucionais (FCNN), especialmente as arquiteturas U-Net e U-Net++, para segmentar áreas afetadas por inundações em imagens do satélite Sentinel-1 do conjunto de dados Cloud to Street Microsoft floods. A arquitetura U-Net++ se destaca na identificação de áreas alagadas, com métricas de Interseção sobre União (IoU) de 0,8280, pontuação F1 de 0,9053 e sensibilidade de 0,9001.
Referências
Almeida, G. A. M. d., Bates, P., and Ozdemir, H. (2018). Modelling urban floods at submetre resolution: challenges or opportunities for flood risk management? Journal of Flood Risk Management, 11:S855–S865.
Bahrami, B. and Arbabkhah, H. (2024). Enhanced flood detection through precise water segmentation using advanced deep learning models. Journal of Civil Engineering Researchers, 6(1):1–8.
ESA (2015). The sentinel-1 toolbox. [link]. Accessed: 2023-11-02.
Guo, Z., Shengoku, H., Wu, G., Chen, Q., Yuan, W., Shi, X., Shao, X., Xu, Y., and Shibasaki, R. (2018). Semantic segmentation for urban planning maps based on u-net. In IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium, pages 6187–6190.
He, K., Zhang, X., Ren, S., and Sun, J. (2016). Deep residual learning for image recognition. In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 770–778.
Kumar, A. (2023). CNN basic architecture for classification & segmentation. [link]. Accessed: 2023-09-30.
Microsoft Open Source, McFarland, M., Emanuele, R., Morris, D., and Augspurger, T. (2022). Microsoft Planetary Computer Data Catalog. [link]. Accessed: 2024-07-02.
Montello, F., Arnaudo, E., and Rossi, C. (2022). Mmflood: A multimodal dataset for flood delineation from satellite imagery. IEEE Access, 10:96774–96787.
Newman, D. J. and ESDIS Standards Coordination Office (ESCO) (2023). Spatiotemporal asset catalogs (stac). [link]. Accessed: 2025-07-02.
Orfanidis, G., Ioannidis, K., Avgerinakis, K., Vrochidis, S., and Kompatsiaris, I. (2018). A deep neural network for oil spill semantic segmentation in sar images. In 2018 25th IEEE International Conference on Image Processing (ICIP), pages 3773–3777.
Rambour, C., Audebert, N., Koeniguer, E., Saux, B. L., Crucianu, M., and Datcu, M. (2020). Flood detection in time series of optical and sar images. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLIII-B2-2020:1343–1346.
Ronneberger, O., Fischer, P., and Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. In Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015, volume 9351, page 234–241.
Sausen, T. M. and Lacruz, M. S. P. (2015). Sensoriamento Remoto para Desastres. Oficina de Textos, São Paulo.
Shruti, J. (2020). A survey of loss functions for semantic segmentation. In 2020 IEEE Conference on Computational Intelligence in Bioinformatics and Computational Biology (CIBCB), pages 1–7.
Street, C. t., Microsoft, and Foundation, R. E. (2022). A global flood events and cloud cover dataset. [link]. Accessed: 2023-08-25.
Tanim, A. H., McRae, C. B., Tavakol-Davani, H., and Goharian, E. (2022). Flood detection in urban areas using satellite imagery and machine learning. Water, 14(7):1140.
Truckenbrodt, J., Cremer, F., Baris, I., and Eberle, J. (2019). Pyrosar: a framework for large-scale sar satellite data processing. In P. Soille, S. Loekken, and S. Albani, editors, Big Data from Space, page 197–200.
WMO (2023). Atlas of mortality and economic losses from weather, climate and water-related hazards (1970-2021). [link]. Accessed: 2024-06-22.
Zhang, A., Lipton, Z. C., Li, M., and Smola, A. J. (2023). Dive into Deep Learning. Cambridge University Press.
Zhou, Z., Siddiquee, M. M. R., Tajbakhsh, N., and Liang, J. (2018). Unet++: A nested u-net architecture for medical image segmentation. In Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support – DLMIA ML-CDS 2018, volume 11045, page 3–11.
Bahrami, B. and Arbabkhah, H. (2024). Enhanced flood detection through precise water segmentation using advanced deep learning models. Journal of Civil Engineering Researchers, 6(1):1–8.
ESA (2015). The sentinel-1 toolbox. [link]. Accessed: 2023-11-02.
Guo, Z., Shengoku, H., Wu, G., Chen, Q., Yuan, W., Shi, X., Shao, X., Xu, Y., and Shibasaki, R. (2018). Semantic segmentation for urban planning maps based on u-net. In IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium, pages 6187–6190.
He, K., Zhang, X., Ren, S., and Sun, J. (2016). Deep residual learning for image recognition. In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 770–778.
Kumar, A. (2023). CNN basic architecture for classification & segmentation. [link]. Accessed: 2023-09-30.
Microsoft Open Source, McFarland, M., Emanuele, R., Morris, D., and Augspurger, T. (2022). Microsoft Planetary Computer Data Catalog. [link]. Accessed: 2024-07-02.
Montello, F., Arnaudo, E., and Rossi, C. (2022). Mmflood: A multimodal dataset for flood delineation from satellite imagery. IEEE Access, 10:96774–96787.
Newman, D. J. and ESDIS Standards Coordination Office (ESCO) (2023). Spatiotemporal asset catalogs (stac). [link]. Accessed: 2025-07-02.
Orfanidis, G., Ioannidis, K., Avgerinakis, K., Vrochidis, S., and Kompatsiaris, I. (2018). A deep neural network for oil spill semantic segmentation in sar images. In 2018 25th IEEE International Conference on Image Processing (ICIP), pages 3773–3777.
Rambour, C., Audebert, N., Koeniguer, E., Saux, B. L., Crucianu, M., and Datcu, M. (2020). Flood detection in time series of optical and sar images. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLIII-B2-2020:1343–1346.
Ronneberger, O., Fischer, P., and Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. In Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015, volume 9351, page 234–241.
Sausen, T. M. and Lacruz, M. S. P. (2015). Sensoriamento Remoto para Desastres. Oficina de Textos, São Paulo.
Shruti, J. (2020). A survey of loss functions for semantic segmentation. In 2020 IEEE Conference on Computational Intelligence in Bioinformatics and Computational Biology (CIBCB), pages 1–7.
Street, C. t., Microsoft, and Foundation, R. E. (2022). A global flood events and cloud cover dataset. [link]. Accessed: 2023-08-25.
Tanim, A. H., McRae, C. B., Tavakol-Davani, H., and Goharian, E. (2022). Flood detection in urban areas using satellite imagery and machine learning. Water, 14(7):1140.
Truckenbrodt, J., Cremer, F., Baris, I., and Eberle, J. (2019). Pyrosar: a framework for large-scale sar satellite data processing. In P. Soille, S. Loekken, and S. Albani, editors, Big Data from Space, page 197–200.
WMO (2023). Atlas of mortality and economic losses from weather, climate and water-related hazards (1970-2021). [link]. Accessed: 2024-06-22.
Zhang, A., Lipton, Z. C., Li, M., and Smola, A. J. (2023). Dive into Deep Learning. Cambridge University Press.
Zhou, Z., Siddiquee, M. M. R., Tajbakhsh, N., and Liang, J. (2018). Unet++: A nested u-net architecture for medical image segmentation. In Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support – DLMIA ML-CDS 2018, volume 11045, page 3–11.
Publicado
11/09/2024
Como Citar
RIBEIRO, Samuel G.; SOUZA, Marcelo M. S. de.
Fully convolutional neural networks on semantic segmentation of flooded areas. In: ESCOLA REGIONAL DE COMPUTAÇÃO DO CEARÁ, MARANHÃO E PIAUÍ (ERCEMAPI), 12. , 2024, Parnaíba/PI.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2024
.
p. 71-79.
DOI: https://doi.org/10.5753/ercemapi.2024.243382.
