• Tam Thanh Nguyen Faculty of Information Technology, Ho Chi Minh City University of Technology (HUTECH), Ho Chi Minh City, Vietnam
  • Toan Thanh Nguyen Faculty of Information Technology, Ho Chi Minh City University of Technology (HUTECH), Ho Chi Minh City, Vietnam
  • Cong Thanh Phan Faculty of Information Technology, Ho Chi Minh City University of Technology (HUTECH), Ho Chi Minh City, Vietnam
  • Quoc Viet Hung Nguyen




Deep Learning, Satellite Imagery Mining, Spatio-temporal Change Prediction, WaterSurface Extraction


Changes in surface water might result in natural disasters such as floods, water shortages, landslides, waterborne diseases, which lead to loss of lives. Timely extracting for surface water and predicting its movement is essential for planning activities and decision-making processes. Most existing works on extracting water surface using satellite images focus on static spectral images and ignore the temporal evolution of data in streams, leading to less accuracy and lack of prediction power. Although some works realize that modeling temporal information of satellite signals could boost the forecasting capability on environmental changes, most of them only focus on prediction tasks independently and separately from the extraction task. In this paper, we propose a unified framework for water extraction and change prediction (WECP) built on top of imagery data streams, which are free to access from orbiting satellites, to locate water surface and predict its changes over time. Our framework is evaluated on Landsat 8 data due to its high spatial resolution. Empirical evaluations on real imagery datasets of different landscapes reveal that our framework is robust in extracting and capturing spatio-temporal changes in the water surface.


[1] T. D. Acharya, A. Subedi, and D. H. Lee, “Evaluation of machine learning algorithms for surfacewater extraction in a landsat 8 scene of nepal,”Sensors, vol. 19, no. 12, p. 2769, 2019.

[2] J. C. Aerts, W. J. Botzen, K. C. Clarke, S. L. Cutter, J. W. Hall, B. Merz, E. Michel-Kerjan,J. Mysiak, S. Surminski, and H. Kunreuther, “Integrating human behaviour dynamics into flooddisaster risk assessment,”Nature Climate Change, vol. 8, no. 3, pp. 193–199, 2018.

[3] R. S. Andersen, A. Peimankar, and S. Puthusserypady, “A deep learning approach for real-timedetection of atrial fibrillation,”Expert Systems with Applications, vol. 115, pp. 465–473, 2019.

[4] T. Ara ́ujo, G. Aresta, E. Castro, J. Rouco, P. Aguiar, C. Eloy, A. Pol ́onia, and A. Campilho,“Classification of breast cancer histology images using convolutional neural networks,”PloS one,vol. 12, no. 6, p. e0177544, 2017.

[5] W. Byeon, M. Liwicki, and T. M. Breuel, “Texture classification using 2d lstm networks,” in2014 22nd international conference on pattern recognition. IEEE, 2014, pp. 1144–1149.

[6] Y. Chen, Z. Lin, X. Zhao, G. Wang, and Y. Gu, “Deep learning-based classification of hyperspec-tral data,”IEEE Journal of Selected topics in applied earth observations and remote sensing,vol. 7, no. 6, pp. 2094–2107, 2014.

[7] R. G. Congalton and K. Green,Assessing the accuracy of remotely sensed data: principles andpractices. CRC press, 2019.

[8] G. L. Feyisa, H. Meilby, R. Fensholt, and S. R. Proud, “Automated water extraction index: Anew technique for surface water mapping using landsat imagery,”Remote Sensing of Environ-ment, vol. 140, pp. 23–35, 2014.

[9] C. Giardino, M. Bresciani, P. Villa, and A. Martinelli, “Application of remote sensing in waterresource management: the case study of lake trasimeno, italy,”Water resources management,vol. 24, no. 14, pp. 3885–3899, 2010.

[10] I. Goodfellow, Y. Bengio, and A. Courville,Deep learning. MIT press, 2016.

[11] A. Graves and J. Schmidhuber, “Framewise phoneme classification with bidirectional lstm andother neural network architectures,”Neural Networks, vol. 18, no. 5-6, pp. 602–610, 2005.

[12] R. Gupta, S. J. Nanda, and U. P. Shukla, “Cloud detection in satellite images using multi-objective social spider optimization,”Applied Soft Computing, vol. 79, pp. 203–226, 2019.

[13] S. K. Jain, A. K. Saraf, A. Goswami, and T. Ahmad, “Flood inundation mapping using noaaavhrr data,”Water Resources Management, vol. 20, no. 6, pp. 949–959, 2006.

[14] S. K. Jain, R. Singh, M. Jain, and A. Lohani, “Delineation of flood-prone areas using remotesensing techniques,”Water Resources Management, vol. 19, no. 4, pp. 333–347, 2005.

[15] L. Ji, L. Zhang, and B. Wylie, “Analysis of dynamic thresholds for the normalized differencewater index,”Photogrammetric Engineering & Remote Sensing, vol. 75, no. 11, 2009.

[16] Z. Jiang, J. Qi, S. Su, Z. Zhang, and J. Wu, “Water body delineation using index compositionand his transformation,”International Journal of Remote Sensing, vol. 33, no. 11, 2012.

[17] C. Jing-bo, L. Shun-xi, W. Cheng-yi, Y. Shu-cheng, and W. Zhong-wu, “Research on urban waterbody extraction using knowledge-based decision tree,”Remote sensing information, vol. 1, 2013.

[18] G. Kallis and D. Butler, “The eu water framework directive: measures and implications,”Waterpolicy, vol. 3, no. 2, pp. 125–142, 2001.

[19] L. Li, Z. Yan, Q. Shen, G. Cheng, L. Gao, and B. Zhang, “Water body extraction from very highspatial resolution remote sensing data based on fully convolutional networks,”Remote Sensing,vol. 11, no. 10, p. 1162, 2019.

[20] P. Liu, H. Zhang, and K. B. Eom, “Active deep learning for classification of hyperspectralimages,”IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,vol. 10, no. 2, pp. 712–724, 2017.

[21] E. Maggiori, Y. Tarabalka, G. Charpiat, and P. Alliez, “Convolutional neural networks forlarge-scale remote-sensing image classification,”IEEE Transactions on Geoscience and RemoteSensing, vol. 55, no. 2, pp. 645–657, 2016.

[22] S. K. McFeeters, “The use of the normalized difference water index (ndwi) in the delineation ofopen water features,”International journal of remote sensing, vol. 17, no. 7, 1996.

[23] A. Mitchell, G. H. Romano, B. Groisman, A. Yona, E. Dekel, M. Kupiec, O. Dahan, and Y. Pilpel,“Adaptive prediction of environmental changes by microorganisms,”Nature, vol. 460, 2009.

[24] L. Mou, L. Bruzzone, and X. X. Zhu, “Learning spectral-spatial-temporal features via a recurrentconvolutional neural network for change detection in multispectral imagery,”IEEE Transactionson Geoscience and Remote Sensing, vol. 57, no. 2, pp. 924–935, 2018.

[25] NASA. (2019) Landsat 8. [Online]. Available: https://landsat.gsfc.nasa.gov/landsat-8/

[26] S. C. Palmer, T. Kutser, and P. D. Hunter, “Remote sensing of inland waters: Challenges,progress and future directions,” 2015.

[27] J. Park, H. Kim, Y.-W. Tai, M. S. Brown, and I. Kweon, “High quality depth map upsamplingfor 3d-tof cameras,” inICCV, 2011, pp. 1623–1630.

[28] L. Qi, D. Yong, N. Xin, X. Jiaqing, and X. Fei, “Remote sensing image classification based ondbn model,”Journal of computer research and development, vol. 51, no. 9, p. 1911, 2014.

[29] G. Sarp and M. Ozcelik, “Water body extraction and change detection using time series: A casestudy of lake burdur, turkey,”Journal of Taibah University for Science, vol. 11, 2017.

[30] P. F. Scheelbeek, F. A. Bird, H. L. Tuomisto, R. Green, F. B. Harris, E. J. Joy, Z. Chalabi,E. Allen, A. Haines, and A. D. Dangour, “Effect of environmental changes on vegetable andlegume yields and nutritional quality,”Proceedings of the National Academy of Sciences, vol.115, no. 26, pp. 6804–6809, 2018.

[31] V. Slavkovikj, S. Verstockt, W. De Neve, S. Van Hoecke, and R. Van de Walle, “Hyperspectralimage classification with convolutional neural networks,” inMM, 2015, pp. 1159–1162.

[32] F. Sun, W. Sun, J. Chen, and P. Gong, “Comparison and improvement of methods for identifyingwaterbodies in remotely sensed imagery,”International journal of remote sensing, vol. 33, no. 21,pp. 6854–6875, 2012.

[33] S. Thirumuruganathan, N. Tang, and M. Ouzzani, “Data curation with deep learning [vision]:Towards self driving data curation,”arXiv preprint arXiv:1803.01384, 2018.

[34] S. Van Tran, W. B. Boyd, P. Slavich, and T. M. Van, “Agriculture and climate change: percep-tions of provincial officials in vietnam,”Journal of Basic and Applied Sciences, vol. 11, 2015.

[35] Y. Wang, Z. Li, C. Zeng, G.-S. Xia, and H. Shen, “An urban water extraction method combin-ing deep learning and google earth engine,”IEEE Journal of Selected Topics in Applied EarthObservations and Remote Sensing, vol. 13, pp. 768–781, 2020.

[36] G. Xu, P. Li, K. Lu, Z. Tantai, J. Zhang, Z. Ren, X. Wang, K. Yu, P. Shi, and Y. Cheng,“Seasonal changes in water quality and its main influencing factors in the dan river basin,”Catena, vol. 173, pp. 131–140, 2019.

[37] H. Xu, “Modification of normalised difference water index (ndwi) to enhance open water featuresin remotely sensed imagery,”International journal of remote sensing, vol. 27, 2006.

[38] S. Yang, C. Xue, T. Liu, and Y. Li, “A method of small water information automatic extractionfrom tm remote sensing images,”Acta Geodaetica et Cartographica Sinica, vol. 39, 2010.

[39] Y. Yang, Y. Liu, M. Zhou, S. Zhang, W. Zhan, C. Sun, and Y. Duan, “Landsat 8 oli image basedterrestrial water extraction from heterogeneous backgrounds using a reflectance homogenizationapproach,”Remote Sensing of Environment, vol. 171, pp. 14–32, 2015.

[40] L. Yu, Z. Wang, S. Tian, F. Ye, J. Ding, and J. Kong, “Convolutional neural networks for waterbody extraction from landsat imagery,”International Journal of Computational Intelligence andApplications, vol. 16, no. 01, p. 1750001, 2017.

[41] C. Zhu, X. Zhang, J. Luo, W. Li, and J. Yang, “Automatic extraction of coastline by remotesensing technology based on svm and auto-selection of training samples,”Remote Sensing forLand and Resources, vol. 25, no. 2, pp. 69–74, 2013.

[42] Z. Zhu, S. Wang, and C. E. Woodcock, “Improvement and expansion of the fmask algorithm:Cloud, cloud shadow, and snow detection for landsats 4–7, 8, and sentinel 2 images,”RemoteSensing of Environment, vol. 159, pp. 269–277, 2015.