Article
Article name	Geoecological Mapping of Urbanized Territories of the Transbaikal Territory Using a Convolutional Neural Network
Authors	Kochev D.. ,
Bibliographic description	Kochev D. V. Geoecological Mapping of Urbanized Territories of the Transbaikal Territory Using a Convolutional Neural Network // Transbaikal State University Journal. 2024. Vol. 30, no. 3. P. 27–37. DOI: 10.21209/2227-9245-2024-30-3-27-37.
Category	Earth and Environmental Sciences
DOI	502/ 504; 528.88
DOI	10.21209/2227-9245-2024-30-3-27-37
Article type	Original article
Annotation	Floods pose a serious threat to economies and communities, but their danger is often underestimated. As a result, potentially dangerous areas of economic territories are subject to intensive development, which leads to damage when they are flooded. The use of modern methods of remote sensing of the Earth in combination with deep learning algorithms can significantly increase the efficiency and accuracy of mapping these territories and makes it possible to effectively manage them, optimizing the processes of planning and development of territories, reducing damage due to flooding. The object of the study is flood-prone areas of settlements in the Transbaikal Territory. The goal of the study is to improve geoecological mapping of the economic use of residential areas of the Transbaikal Territory that are prone to floods during the inter-flood period. The research objectives are as follows: image processing; visualization of objects belonging to different classes of economic use; analysis and assessment of changes within hazardous areas; checking the results obtained using a convolutional neural network; development of a software product that allows you to inform interested parties about the presence of dangerous territories. The methodology and methods are presented by research methods in the field of remote sensing of the Earth and processing of cartographic information. High-resolution remote sensing data, freely distributed by Google Earth services, and data from UAVs were obtained. The data is given in raster format and has a coordinate reference. Data processing has been carried out using the U-Net convolutional neural network and image processing by using a neural network and visualization of objects belonging to different classes of economic use of territories. A retrospective analysis and assessment of development changes in hazardous areas has also been peformed. The data obtained show active individual construction in the danger zone. The adequacy of the results obtained using a convolutional neural network was checked. A software product has been developed to determine the presence of hazardous areas within populated areas, which can significantly increase the efficiency and accuracy of mapping economic activities in populated areas.
Key words	flood-prone areas, floods, decryption, remote sensing of the Earth, neural networks, machine learning, computer program, Python, damage, mapping
Article information
References	1. Bogatyreva A. A., Vinogradova A. R., Tikhomirova S. A. Investigation of the Transfer Learning ability of convolutional neural networks trained on Imagenet. International Journal of Applied and Fundamental Research, no. 7, pp. 106–111, 2019. (In Rus.) 2. Galanov A. E., Selyukova G. P. Neural networks and neural technologies // Actual issues of science and economics: new challenges and solutions: sat. art. LIII International Student Scientific and Practical Conference. Tyumen: State Agrarian University of the Northern Urals, 2019. P. 399–405. (In Rus.) 3. Kochev D. V. Geoecological mapping of buildings in flood-prone areas of the cities of Shilka and Nerchinsk of the Trans-Baikal Territory using the NDBI spectral index and a neural network. Transbaikal State University Journal, vol. 30, no. 1, pp. 28–39, 2024. (In Rus.) 4. Kurganovich K. A., Shalikovsky A. V., Bosov M. A., Kochev D. V. Application of artificial intelligence algorithms for flood-prone areas control. Water Management of Russia: Problems, Technologies, Management, no. 3, pp. 6–24, 2021. (In Rus.) 5. Senichev A. V., Novikova A. I., Vasiliev P. V. Comparison of deep learning with traditional methods of computer vision in problems of defect identification. Young Researcher of the Don, no. 4, P. 64–67, 2020. (In Rus.) 6. Solodukhin A. A. Zabaikalsky Krai – flood-prone region. Technosphere safety of the Baikal region: collection of articles international scientific and practical conference. Chita: ZabGU, 2017. Pp. 24–32. (In Rus.) 7. Shalikovsky A. V. Fundamentals of rational use of flood-prone territories: abstract. ... Doctor of Geographical Sciences: 25.00.36. Chita, 2004. 40 p. (In Rus.) 8. Girshick R., Donahue J., Darrell T., Malik J. Region-based convolutional networks for accurate object detection and semantic segmentation. IEEE Trans Pattern Anal Mach Intell, no. 38, pp. 142–158, 2016. (In Eng.) 9. Goodfellow I., Bengio Y., Courville A. Deep learning. Cambridge (MA): MIT Press, 2016. (In Eng.) 10. Graves A., Liwicki M., Fernandez S., Bertolami R., Bunke H., Schmidhuber J. A. Novel connectionist system for improved unconstrained handwriting recognition. IEEE Trans Pattern Anal Mach Intell, no. 31, pp. 855–868, 2009. (In Eng.) 11. He K., Zhang X., Ren S., Sun J. Deep residual learning for image recognition. Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2016. (In Eng.) 12. Hinton G. E., Osindero S., Teh Y. W. A Fast-learning algorithm for deep belief nets. Neural Comput, no. 18, pp. 1527–1554, 2006. (In Eng.) 13. Hu W., Huang Y., Wei L., Zhang F., Li H. Deep convolutional neural networks for hyperspectral image classification. J Sens, no. 2, pp. 3–12, 2015. (In Eng.) 14. Krizhevsky A., Sutskever I., Hinton G. E. ImageNet classification with deep convolutional neural networks. Adv Neural Inf Process Syst 25. 2012. Web. 12.05.2024. https://papers.nips.cc/paper/4824-imagenet-classification-with-deep-convolutional-neuralnetworks.pdf. (In Eng.) 15. Simonyan K., Zisserman A. Very deep convolutional networks for large-scale image recognition. arXiv. 2015. Web. 12.05.2024. https://arxiv.org/pdf/1409.1556.pdf. (In Eng.)
Full article	Geoecological Mapping of Urbanized Territories of the Transbaikal Territory Using a Convolutional Neural Network