Machine learning algorithms for classifying flood areas from synthetic aperture radar images
DOI:
https://doi.org/10.24850/j-tyca-14-04-03Keywords:
Supervised learning, prediction models, satellite images, decision trees, watershed, remote sensing, ROC curvesAbstract
The use of synthetic aperture radar (SAR) images represents a valuable source of information to characterize geographic regions susceptible to flooding, such as southeastern Mexico, they are not sensitive to cloudy and / or dark conditions. This research presents a methodology to identify bodies of water in a region of southeastern Mexico. Three machine learning algorithms were implemented: Random forests (RF), Gradient Boosting (GB) and Support Vector Machines (SVM) to classify three target classes: Class A (water, flooded areas, and bodies of water); class I (urban infrastructure and / or bare soil), and class V (vegetation) from SAR images. The SAR image used covers a projected geographical area UTM Zona 15 Norte WGS84 located in the states of Tabasco and Chiapas; this was pre-processed to reduce errors in the image. The RF, GB and SVM models were implemented in Python language. These were trained and tested in prediction from a database of 12 000 samples with amplitude values of the SAR image. The RF model obtained an overall classification accuracy ( ) of 97.9 (+/- 0.003) %; GB obtained = 97.9 (+/- 0.003) %, and SVM = 97.4 (+/- 0.005). The three models obtained an value higher than 0.99 to predict class A; RF obtained = 1 for the three target classes. This study shows the potential use of SAR satellite images and the high performance of RF, GB and SVM machine learning models to classify and identify water bodies as well as highlighting its importance in studies of possible impacts of floods.
References
Abdurahman-Bayanudin, A., & Heru-Jatmiko, R. (2016). Orthorectification of Sentinel-1 SAR (Synthetic Aperture Radar) data in some parts of south-eastern Sulawesi using Sentinel-1 Toolbox. 2nd International Conference of Indonesian Society for Remote Sensing (ICOIRS), IOP Conference Series: Earth and Environmental Science, 47(012007). DOI: 10.1088/1755-1315/47/1/012007
ASF, Alaska Satellite Facility. (2020). Distributed active archive center. Recuperado de https://asf.alaska.edu/
Arreguín-Cortés, F. I., & Rubio-Gutiérrez, H. (2014). Análisis de las inundaciones en la planicie tabasqueña en el periodo 1995-2010. Tecnología y ciencias del agua, 5(3), 5-32.
Avendaño-Pérez, J., Parra-Plazas, J., & Fredy-Bayona, J. (2014). Segmentación y clasificación de imágenes SAR en zonas de inundación en Colombia, una herramienta computacional para prevención de desastres. Universidad Antonio Nariño, Revista Facultades de Ingeniería, 4(8), 24-38.
Barrero-Ortiz, G. (2019). Evaluación de la eficiencia de los modelos machine learning para la predicción de la calidad del software desarrollado en IBM RPG usando la matriz de confusión y las curvas ROC. Maestría en Gestión de Tecnologías de Información de la escuela de Postgrado de la Universidad Cesar Vallejo de Lima Perú. ResearchGate. DOI: 10.13140/RG.2.2.31760.87040
Bradley, A. P. (1997). The use of the area under the ROC curve in the evaluation of machine learning algorithms. Pattern Recognition, 30(7), 1145-1159. DOI: 10.1016/S0031-3203(96)00142-2
Breiman, L. (2001). Random forests. Machine Learning, 45, 5–32.
Chen, S., Huang, W., Chen, Y., & Feng, M. (2021). An adaptive thresholding approach toward rapid flood coverage extraction from Sentinel-1 SAR imagery. Remote Sensing, 13(23). DOI: https://doi.org/10.3390/rs13234899
Copernicus Sentinel Data. (2017). Open access to Sentinel-1 user products. Recuperado de https://scihub.copernicus.eu/
ESA & SEOM, European Spatial Agency & Scientific Exploitation of Operational Missions (2019). Sentinel Application Platform (SNAP 7.0). Programa para el análisis y procesamiento de imágenes satelitales (software). Recuperado de http://step.esa.int/main/download/snap-download/
Fawcett, T. (2006). An introduction to ROC analysis. Pattern Recognition Letters, 27, 861-874. DOI: 10.1016/j.patrec.2005.10.010
Fernández-Ordoñez, Y. M., Soria-Ruiz, J., Leblon, B., Macedo-Cruz, A., Ramírez-Guzmán, M. E., & Escalona-Maurice, M. (2020). Imágenes de radar para estudios territoriales, caso: inundaciones en Tabasco con el uso de imágenes SAR Sentinel-1A y Radarsat-2. Realidad, Datos y Espacio, Revista Internacional de Estadística y Geografía, 11(1).
Fernández-Ordoñez, Y. M., & Soria-Ruiz, J. (2015). Imágenes de radar de apertura sintética y conceptos básicos de polarimetría. En: Fernández-Ordoñez, Y., Escalona-Maurice, M. J., & Valdez-Lazalde, J. R. (eds.). Avances y perspectivas de geomática con aplicaciones ambientales, agrícolas y urbanas (pp. 37-66). Montecillo, México: Colegio de Postgraduados.
Friedman, J. (2001). Greedy function approximation: A gradient boosting machine. The Annals of Statistics, 29(5), 1189-1232.
Geurts, P., Ernst, D., & Wehenkel, L. (2006). Extremely randomized trees. Mach Learn, 63, 3-42. DOI: 10.1007/s10994-006-6226-1
Gini, C. (1912). Variabilità e mutabilità. Reprinted in: Pizetti, E., & Salvemini, T (eds.). Memorie di metodologica statistica. Rome, Italy: Libreria Eredi Virgilio Veschi.
Gomarasca, M. A., Tornato, A., Spizzichino, D., Valentini, E., Taramelli, A., Satalino, G., Vincini, M., Boschetti, M., Colombo, R., Rossi, L., Borgogno-Mondino, E., Perotti, L., Alberto, W., & Villa, F. (2019). Sentinel for applications in agriculture. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLII-3/W6, 91-98.
Hlaváčová, I., Kačmařík, M., Lazecký, M., Struhár, J., & Rapant, P. (2021). Automatic open water flood detection from Sentinel-1 multi-temporal imagery. Water, 13(23), 3392. DOI: https://doi.org/10.3390/w13233392
Lin, Y. N., Yun, S., Bhardwaj, A., & Hill, E. M. (2019). Urban flood detection with Sentinel-1 multi-temporal Synthetic Aperture Radar (SAR) observations in a Bayesian framework: A case study for hurricane Matthew. Remote Sensing, 11(15). DOI: 10.3390/rs11151778
MathWorks Inc. (2021). Documentación: cvpartition. Recuperado de https://la.mathworks.com/help/stats/cvpartition.html
MathWorks Inc. (2016). MATrix LABoratory (MATLAB R2016a). Plataforma matemática sumamente potente en la manipulación de matrices y análisis numérico (software). Natick, USA: MathWorks Inc.
Parikh, R., Mathai, A., Parikh, S., Chandra-Sekhar, G., & Thomas, R. (2008). Understanding and using sensitivity, specificity and predictive values. Indian Journal of Ophthalmology, 56(1), 45-50. DOI: 10.4103/0301-4738.37595
Patle, A., & Chouhan, D. S. (2013). SVM kernel functions for classification. 2013 International Conference on Advances in Technology and Engineering (ICATE), 2013, 1-9. DOI: 10.1109/ICAdTE.2013.6524743
Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., Vanderplas, J., Passos, A., Cournapeau, D., Brucher, M., Perrot, M., & Duchesnay, E. (2011). Scikit-learn: Machine learning in Python. Journal of Machine Learning Research, 12, 2825-2830.
Perevochtchikova, M., & Lezama-de-la-Torre, J. L. (2010). Causas de un desastre: inundaciones del 2007 en Tabasco, México. Journal of Latin American Geography, 9(2), 73-98.
Podest, E. (16 de abril, 2018). Imágenes Radar (SAR) Seminario NASA - Preprocesamiento, clasificación agua, tierra, Sentinel 1 (2/4) (video). Recuperado de https://youtu.be/pVeh9ChATwA
Pulella, A., Aragão-Santos, R., Sica, F., Posovszky, P., & Rizzoli, P. (2020). Multi-temporal Sentinel-1 backscatter and coherence for rainforest mapping. Remote Sensing, 12(847). DOI: 10.3390/rs12050847
QGIS.org. (2020). QuantumGis (QGIS versión 3.10.7-A Coruña). Sistema de Información Geográfica, proyecto de fundación geoespacial de código abierto (software). Recuperado de https://qgis.org/es/site/
Raschka, S. (2018). Model evaluation, model selection, and algorithm selection in machine learning. Madison, USA: University of Wisconsin.
Raschka, S., & Mirjalili, V. (2017). Python Machine Learning: Machine Learning and Deep Learning with Python, Scikit-learn, and TensorFlow (2nd ed.). Birmingham, UK: Packt Publishing Ltd.
Sami, A., & Abdulmunem, M. E. (2020). Synthetic aperture radar image classification: A survey. Iraqi Journal of Science, 61(5), 1223-1232. DOI: 10.24996/ijs.2020.61.5.29.
Sánchez, A. J., Salcedo, M. A., Florido, R., & Mendoza, J. D. (2015). Ciclos de inundación y conservación de servicios ambientales en la cuenca baja de los ríos Grijalva-Usumacinta. Contactos, 97, 5-14.
Shen, X., Wang, D., Mao, K., Anagnostou, E., & Hong, Y. (2019). Inundation extent mapping by synthetic aperture radar: A review. Remote Sensing, 11(879). DOI: 10.3390/rs11070879
Smola, A., & Schölkopf, B. (2004). A tutorial on support vector regression. Statistics and Computing, 14, 199-222.
Swamynathan, M. (2017). Mastering Machine Learning with Python in Six Steps. Apress. DOI: 10.1007/978-1-4842-2866-1
UN-SPIDER, United Nations Platform for Space-based Information for Disaster Management and Emergency Response. (2020). Step-by-step: Mudslides and associated flood detection using Sentinel-1 data. Recuperado de https://un-spider.org/advisory-support/recommended-practices/mudslides-flood-sentinel-1/step-by-step
Vapnik, V. (1995). The nature of statistical learning theory. New York, USA: Springer.
Zhang, B., Wdowinski, S., Oliver-Cabrera, T., Koirala, R., Jo, M. J., & Osmanoglu, B. (2018). Mapping the extent and magnitude of sever flooding induced by hurricane Irma with multi-temporal sentinel-1 SAR and InSAR observations. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLII-3(3), 2237-2244. DOI: 10.5194/isprs-archives-XLII-3-2237-2018
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2023 Tecnología y ciencias del agua
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
By Instituto Mexicano de Tecnología del Agua is distributed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. Based on a work at https://www.revistatyca.org.mx/. Permissions beyond what is covered by this license can be found in Editorial Policy.
