Raqamli Transformatsiya va Sun’iy Intellekt
CHUQUR O‘RGANISH ASOSIDA TASVIRLARDAN PM2.5 KONSENTRATSIYASINI BASHORATLOVCHI CNN–FNN REGRESSIYA MODELI
1 Toshkent irrigatsiya va qishloq xoʻjaligini mexanizatsiyalash muhandislari instituti milliy tadqiqot universiteti
Keywords
Abstract
Atmosfera havosidagi mayda dispers zarrachalar, xususan PM2.5, aholi salomatligi va ekologik xavfsizlikka bevosita ta’sir ko‘rsatadigan muhim ifloslantiruvchi omillardan biridir. Mazkur maqolada havo muhitining vizual belgilariga tayangan holda PM2.5 konsentratsiyasini avtomatik baholashga mo‘ljallangan CNN–FNN regressiya modeli taklif etiladi. Tadqiqotda Hindiston va Nepal hududlarida olingan 224×224 o‘lchamli 12240 ta tasvir hamda ularga mos PM2.5 qiymatlaridan foydalanildi. Tasvirlardan chuqur xususiyatlarni ajratish uchun ImageNet ma’lumotlar to‘plamida oldindan o‘qitilgan EfficientNetV2B0 arxitekturasi qo‘llanildi, hosil qilingan 1280 o‘lchamli belgilar vektori esa to‘liq bog‘langan neyron tarmoq orqali regressiya masalasi sifatida qayta ishlanadi. Tajriba natijalarida normallashtirilgan birliklarda Test MAE = 0.072 va Test MSE = 0.011 ko‘rsatkichlari olindi. Natijalar tasvirga asoslangan yondashuv havo sifatini kontaktlarsiz monitoring qilish, sensor tarmoqlari yetarli bo‘lmagan hududlarda PM2.5 darajasini tezkor baholash va kelgusida multimodal ekologik monitoring tizimlariga integratsiya qilish uchun istiqbolli ekanligini ko‘rsatadi
References
- Qin, Y., Zhou, Y., & Zhang, Y. (2021).
- https://doi.org/10.1016/j.envpol.2019.05.143
- Zhou, T., & Feng, Y. (2023). Hybrid deep learning approach for PM2.5 forecasting using image and numerical data fusion. Environmental Modelling & Software, 164, 105602. https://doi.org/10.1016/j.envsoft.2023.105602
- https://doi.org/10.1109/ACCESS.2024.3350045
- Liu, Q., Zhang, C., Li, J., & Zhao, B. (2022). Integrating CNN and LSTM networks for dynamic air quality prediction based on urban images and meteorological factors. Environmental Pollution, 299, 118884.
- Zhang, C., & Li, Y. (2021). Multimodal deep learning for air quality prediction using images and environmental data. IEEE Access, 9, 43225–43237. https://doi.org/10.1109/ACCESS.2021.3065994
- Chakraborty, T., Banerjee, S., & Basu, S. (2023). Fine particulate matter (PM2.5) prediction using transfer learning-based CNN models from satellite imagery. Environmental Science and Pollution Research, 30(18), 53986–53999.
- Zhu, Y., Liu, Y., & Qin, K. (2018). Prediction of PM2.5 concentration based on convolutional neural network and meteorological data. Atmospheric Pollution Research, 9(5), 840–848. https://doi.org/10.1016/j.apr.2018.02.003
- Li, T., Shen, H., Yuan, Q., Zhang, X., & Zhang, L. (2017). Estimating ground-level PM2.5 by fusing satellite and station observations: A geo-intelligent deep learning approach. Geophysical Research Letters, 44(23), 11,985–11,993. https://doi.org/10.1002/2017GL075710
- A machine learning method to estimate PM2.5 concentrations across Europe using Sentinel-3 satellite data. Remote Sensing of Environment, 222, 417–426. https://doi.org/10.1016/j.rse.2018.12.034
- Remote Sensing, 13(6), 1138. https://doi.org/10.3390/rs13061138
- Zheng, Y., Liu, F., & Hsieh, H. P. (2013). U-Air: When urban air quality inference meets big data. Proceedings of the 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD '13), 1436–1444. https://doi.org/10.1145/2487575.2488188
- https://doi.org/10.1007/s11356-023-27657-0
- Ma, X., Shao, Y., Ma, J., & Chen, S. (2020). A convolutional neural network model for predicting PM2.5 concentrations with satellite remote sensing data. Atmospheric Environment, 229, 117484. https://doi.org/10.1016/j.atmosenv.2020.117484
- Sharma, M., Patel, R., & Joshi, P. (2024). Deep CNN-based estimation of urban air pollution (PM2.5) from ground images. IEEE Access, 12, 15246–15260.
- Hong, C., Zhang, Q., Zhang, Y., Davis, S. J., Zhang, X., & Zheng, Y. (2020). Impacts of climate change on future air quality and human health in China. Proceedings of the National Academy of Sciences, 117(29), 17193–17200. https://doi.org/10.1073/pnas.1915935117
- Wang, Y., & Liu, Y. (2021). Deep learning-based retrieval of PM2.5 from Sentinel-2 images with spatial and temporal data fusion.
- Yang, Y., & Jiang, B. (2021).
- Air quality prediction using multimodal deep learning with meteorological and image data. Atmospheric Research, 260, 105685. https://doi.org/10.1016/j.atmosres.2021.105685
- Zhang, H., Xu, H., & Gong, W. (2020). PM2.5 concentration estimation using deep CNN with multisource data fusion. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 13, 538–549. https://doi.org/10.1109/JSTARS.2020.2968519
- Chen, J., de Hoogh, K., Gulliver, J., Hoffmann, B., & Hertel, O. (2019).
- Lary, D. J., Lary, T., & Sattler, B. (2015). Using machine learning to estimate global PM2.5 for environmental health studies. Environmental Health Insights, 9(S1), 41–52. https://doi.org/10.4137/EHI.S15664
- Li, X., Peng, L., Yao, X., Cui, S., & Hu, Y. (2022). Long short-term memory neural network for air pollutant concentration predictions: Method and evaluation. Environmental Pollution, 231, 997–1004. https://doi.org/10.1016/j.envpol.2017.08.114
- Gupta, P., & Christopher, S. A. (2009). Particulate matter air quality assessment using integrated surface, satellite, and meteorological products: Multiple regression approach. Journal of Geophysical Research: Atmospheres, 114(D14). https://doi.org/10.1029/2008JD011496
- Ma, J., Ding, Y., Cheng, J. C. P., Jiang, F., & Wan, Z. (2019). A spatiotemporal data-driven approach for ground-level PM2.5 estimation using satellite remote sensing data. Environmental Pollution, 252, 575–585.
- https://doi.org/10.1016/j.envpol.2022.118884
- Transfer learning for air pollution estimation using satellite imagery. Environmental Modelling & Software, 145, 105188. https://doi.org/10.1016/j.envsoft.2021.105188
- Hu, X., Belle, J. H., Meng, X., Wildani, A., Waller, L. A., & Liu, Y. (2017). Estimating PM2.5 concentrations in the conterminous United States using the random forest approach. Environmental Science & Technology, 51(12), 6936–6944. https://doi.org/10.1021/acs.est.7b01210