Детектирование новых антропогенных объектов на разновременных спутниковых изображениях Белорусской космической системы дистанционного зондирования Земли

Леонид Владимирович Семененко; Евгений Николаевич Кочик; Александр Михайлович Сорока

doi:10.33581/2520-2243-2025-3-%p

Authors

Leanid V. Semianenka A. N. Sevchenko Institute of Applied Physics Problems, Belarusian State University, 7 Kurchatava Street, Minsk 220045, Belarus
Eugene N. Kochyk A. N. Sevchenko Institute of Applied Physics Problems, Belarusian State University, 7 Kurchatava Street, Minsk 220045, Belarus
Aliaksandr M. Saroka Independent researcher, Zaslawl, Belarus

Keywords:

detection, environmental monitoring, multi-temporal satellite images, Belarusian space system of Earth remote sensing, neural network, deep learning, semantic segmentation

Abstract

The article considers the solution to the problem of detecting new anthropogenic objects during environmental monitoring in the territory of Belarus using multi-temporal panchromatic satellite images of the Belarusian space system of Earth remote sensing and a deep learning neural network that ensures the most reliable detection of changes in ground objects while minimising false positives. To solve this problem, an analysis of deep learning neural network models for detecting new anthropogenic objects based on semantic segmentation of multi-temporal satellite images was carried out. Based on this results, a deep learning neural network model was selected, the hypothesis about the need for training on our own formed dataset of anthropogenic objects was tested, training and validation datasets were formed. The training of the selected deep learning neural network model, including the optimisation of settings and selection of hyperparameters, was carried out to implement the task of detecting new anthropogenic objects based on multi-temporal panchromatic satellite images. The obtained results of this research showcased the practical feasibility of automating the detection of new anthropogenic objects during environmental monitoring.

Author Biographies

Leanid V. Semianenka, A. N. Sevchenko Institute of Applied Physics Problems, Belarusian State University, 7 Kurchatava Street, Minsk 220045, Belarus

PhD (engineering); head of the laboratory of information technology
Eugene N. Kochyk, A. N. Sevchenko Institute of Applied Physics Problems, Belarusian State University, 7 Kurchatava Street, Minsk 220045, Belarus

senior researcher at the laboratory of information technology
Aliaksandr M. Saroka, Independent researcher, Zaslawl, Belarus

independent researcher

References

Белозерский ЛА, Мурашко НИ, Орешкина ЛВ. Проблемы использования разновременных спутниковых снимков в решении задач автоматического обнаружения изменений на местности. В: Абламейко СВ, Казаченок ВВ, Красногир ЕГ, Буза МК, Воротницкий ЮИ, Журавков МА и др., редакторы. Международный конгресс по информатике: информационные системы и технологии. Материалы Международного научного конгресса; 24–27 октября 2016 г.; Минск, Беларусь. Минск: БГУ; 2016. c. 4–8.
Макриденко ЛА, Волков СН, Горбунов АВ, Салихов РС, Ходненко ВП. КА «Канопус-В» № 1 – первый российский малый космический аппарат высокодетального дистанционного зондирования Земли нового поколения. Вопросы электромеханики. Труды ВНИИЭМ. 2017;156:10–20. EDN: YWELEH.
Bandara WGC, Patel VM. A transformer-based Siamese network for change detection. In: Institute of Electrical and Electronics Engineers. IGARSS-2022. Proceedings of the 2022 IEEE International geoscience and remote sensing symposium; 2022 July 17–22; Kuala Lumpur, Malaysia. [S. l.]: Institute of Electrical and Electronics Engineers; 2022. p. 207–210. DOI: 10.1109/IGARSS46834.2022.9883686.
Daudt RC, Le Saux B, Boulch A. Fully convolutional Siamese networks for change detection. In: Institute of Electrical and Electronics Engineers. 2018 25th IEEE International conference on image processing (ICIP); 2018 October 7–10; Athens, Greece. [S. l.]: Institute of Electrical and Electronics Engineers; 2018. p. 4063–4067. DOI: 10.1109/ICIP.2018.8451652.
Liu Y, Pang C, Zhan Z, Zhang X, Xue Yang X. Building change detection for remote sensing images using a dual-task constrained deep Siamese convolutional network model. IEEE Geoscience and Remote Sensing Letters. 2021;18(5):811–815. DOI: 10.1109/LGRS.2020.2988032.
Chen H, Shi Z. A spatial-temporal attention-based method and a new dataset for remote sensing image change detection. Remote Sensing. 2020;12(10):1662. DOI: 10.3390/rs12101662.
Zhang C, Yue P, Tapete D, Jiang L, Shangguan B, Huang L, et al. A deeply supervised image fusion network for change detection in high resolution bi-temporal remote sensing images. ISPRS Journal of Photogrammetry and Remote Sensing. 2020;166:183–200. DOI: 10.1016/j.isprsjprs.2020.06.003.
Fang S, Li K, Shao J, Li Z. SNUNet-CD: a densely connected Siamese network for change detection of VHR images. IEEE Geoscience and Remote Sensing Letters. 2021;19:8007805. DOI: 10.1109/LGRS.2021.3056416.
Chen H, Qi Z, Shi Z. Remote sensing image change detection with transformers. IEEE Transactions on Geoscience and Remote Sensing. 2022;60:5607514. DOI: 10.1109/TGRS.2021.3095166.
van Rijsbergen CJ. Information retrieval. 2nd edition. London: Butterworths; 1979. 208 p.
Loshchilov I, Hutter F. Decoupled weight decay regularization. In: Proceedings of the 7th International conference on learning representations (ICLR-2019); 2019 May 6–9; New Orleans, USA. Volume 6. Red Hook: Curran Associates; 2019. p. 4061–4078.
Ronneberger O, Fischer P, Brox T. U-Net: convolutional networks for biomedical image segmentation. In: Navab N, Hornegger J, Wells WM, Frangi AF, editors. Medical image computing and computer-assisted intervention – MICCAI-2015. Proceedings of the 18th International conference; 2015 October 5–9; Munich, Germany. Part 3. Cham: Springer; 2015. p. 234–241 (Lecture notes in computer science; volume 9351). DOI: 10.1007/978-3-319-24574-4_28.