Vol. 37, issue 02, article # 3

Abstract:

The paper presents an algorithm based on a convolutional neural network with a modified U-Net architecture for detecting cloud formations in satellite images. Multispectral satellite images obtained from the MSU-GS instrument installed at Arktika-M No 1 satellite are used as input data. The accuracy of the algorithm was evaluated using machine learning metrics and comparing the results with reference masks compiled by manual decryption of the satellite images by an experienced decoder specialist. In addition, a comparison with similar products based on data of the SEVIRI and VIIRS instruments was conducted. For areas illuminated by the sun, the cloud mask obtained by the proposed algorithm has an accuracy of 92% compared to the reference mask, and for areas not illuminated by the sun, the accuracy is 89%.

Keywords:

MSU-GS, Arktika-M, cloud mask, cloud detection, neural network classifier, U-Net

References:

1. Miller S., Lee T., Fennimore R. Satellite-based imagery techniques for daytime cloud/snow delineation from MODIS // J. Appl. Meteorol. 2005. V. 44. P. 987–997.
2. Hawotte F., Radoux J., Chomé G., Defourny P. Assessment of automated snow cover detection at high solar zenith angles with PROBA-V // Remote Sens. 2016. V. 8, N 9. P. 699.
3. Zhu Z., Woodcock C.E. Object-based cloud and cloud shadow detection in Landsat imagery // Remote Sens. Environ. 2012. V. 118. P. 83–94.
4. Jedlovec G. Automated detection of clouds in satellite imagery // Adv. Geosci. Remote Sens. 2009. P. 303–316.
5. Mahajan S., Fataniya B. Cloud detection methodologies: Variants and development – a review // Complex Intelligent Syst. 2020. V. 6. P. 251–261.
6. Chen Y., Fan R., Bilal M., Yang X., Wang J., Li W. Multilevel cloud detection for high-resolution remote sensing imagery using multiple convolutional neural networks // ISPRS Int. J. Geo-Inf. 2018. V. 7, N 5. P. 181.
7. Wen-Jia C., Jiang-Yong D., Juan M. Cloud detection via convolutional neural network in visible light remote sensing image // 2nd International Conference on Artificial Intelligence: Techniques and Applications (AITA_2017). China: DEStech Publications, 2017. P. 38–43.
8. Zhaoxiang Z., Iwasaki A., Xu G., Song J. Cloud detection on small satellites based on lightweight U-net and image compression // J. Appl. Remote Sens. 2019. V. 13, N 2. P. 026502.
9. Bloshchinskiy V.D., Kuchma M.O., Andreev A.I., Sorokin A.A. Snow and cloud detection using a convolutional neural network and low-resolution data from the Electro-L No. 2 Satellite // J. Appl. Remote Sens. 2020. V. 14, N 3. P. 034506.
10. Cao K., Zhang X. An Improved Res-UNet model for tree species classification using airborne high-resolution images // Remote Sens. 2020. V. 12. P. 1128.
11. Soni A., Koner R., Villuri V.G.K. M-UNet: Modified U-Net segmentation framework with satellite imagery / J. Mandal, S. Mukhopadhyay (eds.) // Proc. of the Global AI Congress 2019. Advances in Intelligent Systems and Computing. Singapore: Springer, 2020. V. 1112. P. 47–59.
12. Ronneberger O., Fischer P., Brox T. U-Net: Convolutional networks for biomedical image segmentation / N. Navab, J. Hornegger, W. Wells, A. Frangi (eds.). Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015. MICCAI 2015. Lecture Notes in Computer Science. Cham: Springer, 2015. V. 9351.
13. Guo Y., Cao X., Liu B., Gao M. Cloud detection for satellite imagery using attention-based U-Net convolutional neural network // Symmetry. 2020. V. 12, N 6. P. 1056.
14. Filei A.A. Vosstanovlenie opticheskoi tolshchiny i effektivnogo radiusa chastits oblachnosti po dannym dnevnykh izmerenii sputnikovogo radiometra MSU-MR // Optika atmosf. i okeana. 2019. V. 32, N 8. P. 650–656.