Vol. 37, issue 01, article # 6

Abramova I. A., Demchev D. M., Kharyutkina E. V., Savenkova E. N., Sudakov I. A. Use of the U-Net convolutional neural network and its modifications for segmentation of tundra lakes in satellite optical images. // Optika Atmosfery i Okeana. 2024. V. 37. No. 01. P. 48–53. DOI: 10.15372/AOO20240106 [in Russian].
Copy the reference to clipboard


Tundra lakes are an important indicator of climate change; therefore, the analysis of the dynamics of their size is of particular interest. This paper presents the results of using the U-Net convolutional neural network for tundra lakes segmentation in satellite optical images using Landsat data as an example. The comparative assessment of segmentation accuracy is performed for the original U-Net design and its modifications: U-Net++, Attention U-Net, and R2 U-Net, including with weights derived from a pre-trained VGG16 network. The segmentation accuracy is assessed based on the results of manual mapping of tundra lakes in northern Siberia. It is shown that more recent U-Net modifications do not provide a practically significant gain in segmentation accuracy, but increase the computational costs. A configuration based on the classic U-Net gives the best result in most cases (the average Soerens coefficient IoU = 0.88). The technique suggested and the resulting estimates can be used in analysis of modern climate trends.


tundra lakes, U-Net, Arctic, remote sensing, permafrost


1. Sudakov I., Essa A., Mander L., Gong M., Kariyawasam T. The geometry of large tundra lakes observed in historical maps and satellite images // Remote Sens. 2017. V. 9, N 10. P. 1072.
2. Sobiech J., Dierking W. Observing lake- and river-ice decay with SAR: advantages and limitations of the unsupervised k-means classification approach // Ann. Glaciol. 2013. V. 54, N 62. P. 65–72.
3. Demchev D., Sudakow I., Khodos A., Abramova I., Lyakhov D., Michels D. Recognizing the Shape and Size of Tundra Lakes in Synthetic Aperture Radar (SAR) Images Using Deep Learning Segmentation // Remote Sens. 2023. V. 15, N 5. P. 1298.
4. Ronneberger O., Fischer P., Brox T. U-net: Convolutional networks for biomedical image segmentation // Medical Image Computing and Computer-Assisted Intervention-MICCAI 2015: 18th International Conference, Munich, Germany, October 5–9, 2015. P. 234–241.
5. Sudakow I., Asari V.K., Liu R., Demchev D. MeltPondNet: A Swin Transformer U-Net for Detection of Melt Ponds on Arctic Sea Ice // IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. IEEE, 2022. V. 15. P. 8776–8784.
6. Wang Y.-R., Li X.-M. Arctic sea ice cover data from spaceborne synthetic aperture radar by deep learning // Earth Syst. Sci. Data. 2021. V. 13, N 6. P. 2723–2742.
7. Zhou Z., Rahman Siddiquee M.M., Tajbakhsh N., Liang J. U-Net++: A nested U-Net architecture for medical image segmentation // Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support: 4th International Workshop, DLMIA 2018, and 8th International Workshop, ML-CDS 2018, Held in Conjunction with MICCAI 2018, Granada, Spain, September 20, 2018. V. 11045. P. 3–11.
8. Oktay O., Schlemper J., Folgoc L. Le, Lee M., Heinrich M., Misawa K., Mori K., McDonagh S., Hammerla N.Y., Kainz B., Cilocker B., Rueckert D. Attention U-Net: Learning where to look for the pancreas // Arxiv: Computer Vision and Pattern Recognition. DOI: 10.48550/arXiv1804.03999.
9. Alom M.Z., Hasan M., Yakopcic C., Taha T.M., Asari V.K. Recurrent residual convolutional neural network based on U-Net (R2 U-Net) for medical image segmentation // Arxiv: Computer Vision and Pattern Recognition. DOI: 10.48550/arXiv1802.06955.
10. Deng J., Dong W., Socher R., Li L.-J., Li K., Fei-Fei L. Imagenet: A large-scale hierarchical image database // IEEE Conf. Comp. Vision and Patt. Recog. 2009. P. 248–255.
11. Kingma D.P., Ba J. Adam: A method for stochastic optimization // Arxivi Machine Learning. DOI: 10.48550/arXiv1412.6980.