Russian Federation
The relevance of this study is driven by the rapid growth of graphic data coming from unmanned aerial vehicles, video surveillance, satellite reconnaissance, and other sources used in emergency monitoring and response. This paper presents a comprehensive overview of modern methods for processing graphic data based on artificial intelligence technologies, including convolutional neural networks, generative adversarial networks, and variational autoencoders. It examines the main applications of intelligent models for image analysis and generation, as well as their role in classification, segmentation, object detection, image restoration, modeling possible emergency scenarios, and generating synthetic training data. The study's results demonstrate that the use of deep neural network models can significantly improve the accuracy, speed, and reliability of graphic data processing, thereby expanding the functionality of emergency monitoring, warning, and response systems.
artificial intelligence, model of artificial intelligence, deep learning, image recognition, image generation, neural network, variable autocoder
1. Zhang X., Dahu W. Application of artificial intelligence algorithms in image processing // Journal of Visual Communication and Image Representation. 2019. Vol. 61. P. 42–49.
2. Labinskij A.Yu. Osobennosti ispol'zovaniya nejronnoj seti dlya raspoznavaniya izobrazhenij // Problemy upravleniya riskami v tekhnosfere. 2016. № 3 (39). S. 67–73.
3. SHkurat D.E., Matveev A.V. Problemy obnaruzheniya pozhara po videoizobrazheniyu: obzor issledovanij // Prirodnye i tekhnogennye riski (fiziko-matematicheskie i prikladnye aspekty). 2025. № 2 (54). S. 59–73.
4. Labinskij A.Yu. Osobennosti nejronnyh setej, ispol'zuyushchih metody glubokogo obucheniya // Nauchno-analiticheskij zhurnal «Vestnik Sankt-Peterburgskogo universiteta Gosudarstvennoj protivopozharnoj sluzhby MCHS Rossii». 2021. № 4. S. 58–66.
5. Comprehensive review of artificial neural network applications to pattern recognition / O.I. Abiodun [et al.] // IEEE access. 2019. Vol. 7. P. 158820–158846.
6. Mual' M.N.B., Kozyrev D.V. Primenenie svertochnyh nejronnyh setej dlya obnaruzheniya i raspoznavaniya izobrazhenij na osnove samopisnogo generatora // Sovremennye informacionnye tekhnologii i IT-obrazovanie. 2022. T. 18. № 3. S. 507–515.
7. Loktev D.A., Proletarskij A.V., Loktev A.A. Modelirovanie sistemy monitoringa ob"ektov s ispol'zovaniem svertochnoj nejronnoj seti // Promyshlennye ASU i kontrollery. 2020. № 9. S. 39–46.
8. Tarasov A.S., Nikiforov M.B., Bakambis N.I. Primenenie svertochnyh segmentacionnyh nejronnyh setej dlya ekologicheskogo monitoringa zemnoj poverhnosti // Izvestiya Tul'skogo gosudarstvennogo universiteta. Tekhnicheskie nauki. 2021. № 6. S. 3–10.
9. Malahov Yu.A., Androsov A.A., Averchenkov A.V. Analiz i primenenie generativno-sostyazatel'nyh setej dlya polucheniya izobrazhenij vysokogo kachestva // Ergodizajn. 2020. № 4 (10). S. 167–176.
10. Synthetic Data generation using DCGAN for improved traffic sign recognition / C. Dewi [et al.] // Neural Computing and Applications. 2022. Vol. 34. №. 24. P. 21465–21480.
11. Modeling and simulating spatial extremes by combining extreme value theory with generative adversarial networks / Y. Boulaguiem [et al.] // Environmental Data Science. 2022. Vol. 1. P. e5.
12. GANsight Innovations: Generative Adversarial Networks for Innovative Medical Imaging / L. Chandana [et al.] // 2024 4th International Conference on Mobile Networks and Wireless Communications (ICMNWC). IEEE, 2024. P. 1–7.
13. Stackgan++: Realistic image synthesis with stacked generative adversarial networks / H. Zhang [et al.] // IEEE transactions on pattern analysis and machine intelligence. 2018. Vol. 41. № 8. P. 1947–1962.
14. Text-to-traffic generative adversarial network for traffic situation generation / Huo G. [et al.] // IEEE Transactions on Intelligent Transportation Systems. 2021. Vol. 23. № 3. P. 2623–2636.
15. Variational autoencoder anomaly-detection of avalanche deposits in satellite SAR imagery / S. Sinha [et al.] // Proceedings of the 10th International Conference on Climate Informatics. 2020. P. 113–119.
16. Wangiyana S. Flood Detection Using Variational Autoencoder in SAR Images // 2023 Signal Processing Symposium (SPSympo). IEEE, 2023. P. 195–197.
17. Avtomatizaciya procedur obnaruzheniya lesnyh pozharov po rezul'tatam obrabotki video / D.V. Vasil'eva [i dr.] // Nauchno-analiticheskij zhurnal Vestnik Sankt-Peterburgskogo universiteta gosudarstvennoj protivopozharnoj sluzhby MCHS Rossii". 2023. № 4. S. 47–58.
