Automatic cloud detection of sentinel-2 satellite images based on neural network

doi:10.13474/j.cnki.11-2246.2019.0248

Abstract

Abstract: This paper proposed a high accuracy segmentation method of remote sensing image cloud region based on deep learning to overcome the problem of misjudgment of information caused by cloud covering when using sentinel-2 satellite image to extract ground object information. This method constructs a deep neural network model through the preprocessed remote sensing sample data, and automatically extracts high-level image features. Then the image features are input into the classifier to realize the pixelwise classification of remote sensing image, and the cloud coverage matrix is segmented. Finally, the cloud coverage matrix is transformed into a cloud binary map, which is combined with the region of interest to accurately obtain the cloud detection results of the designated region. The method will provide a new idea for automatic cloud detection in irregular region of sentinel-2 satellite images without auxiliary information and human intervention.

Key words: cloud detection, deep learning algorithm, pixelwise, Sentinel-2 MSI, small region

CLC Number:

P237

YU Changhui, YU Haiwei, ZHANG Wen, MENG Lingkui. Automatic cloud detection of sentinel-2 satellite images based on neural network[J]. 测绘通报, 2019, 0(8): 39-43.

References

[1] ZHANG Y, ROSSOW W B, LACIS A A, et al. Calculation of radiative fluxes from the surface to top of atmosphere based on ISCCP and other global data sets:refinements of the radiative transfer model and the input data[J]. Journal of Geophysical Research:Atmospheres, 2004,109(D19):486-498.
[2] 刘紫涵, 吴艳兰. 遥感图像云检测方法研究进展[J]. 国土资源遥感, 2017,29(4):6-12.
[3] ZHU Z, WOODCOCK C E. Automated cloud, cloud shadow, and snow detection in multitemporal Landsat data:an algorithm designed specifically for monitoring land cover change[J]. Remote Sensing of Environment, 2014,152:217-234.
[4] ZHU Z, WANG S, WOODCOCK C E. Improvement and expansion of the Fmask algorithm:cloud, cloud shadow, and snow detection for Landsats 4-7, 8, and Sentinel 2 images[J]. Remote Sensing of Environment, 2015,159:269-277.
[5] HAN Y, KIM B, KIM Y, et al. Automatic cloud detection for high spatial resolution multi-temporal images[J]. Remote Sensing Letters, 2014,5(7):601-608.
[6] HOLLSTEIN A, SEGL K, GUANTER L, et al. Ready-to-use methods for the detection of clouds, cirrus, snow, shadow, water and clear sky pixels in sentinel-2 MSI images[J]. Remote Sensing, 2016,8(8):666.
[7] 康一飞, 潘励, 孙明伟, 等. 基于高斯混合模型法的国产高分辨率卫星影像云检测[J]. 武汉大学学报(信息科学版), 2017,42(6):782-788.
[8] 谭凯, 张永军, 童心, 等. 国产高分辨率遥感卫星影像自动云检测[J]. 测绘学报, 2016,45(5):581-591.
[9] YUAN Y, HU X. Bag-of-words and object-based classification for cloud extraction from satellite imagery[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2015,8(8):4197-4205.
[10] LI Z, SHEN H, LI H, et al. Multi-feature combined cloud and cloud shadow detection in GaoFen-1 wide field of view imagery[J]. Remote Sensing of Environment, 2017,191:342-358.
[11] 陈洋, 范荣双, 王竞雪, 等. 基于深度学习的资源三号卫星遥感影像云检测方法[J]. 光学学报, 2018,38(1):362-367.
[12] FAN Z, ZHANG W, ZHANG D, et al. An automatic accurate high-resolution satellite image retrieval method[J]. Remote Sensing, 2017,9(11):1092.
[13] AN Z, SHI Z. Scene learning for cloud detection on remote-sensing images[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2015,8(8):4206-4222.
[14] 李爱勤, 王环东, 王静怡, 等. 高分辨率遥感影像中云和似云目标的自动区分[J]. 测绘通报, 2017(6):31-35.
[15] KINGMA D P, BA J. Adam:a method for stochastic optimization[C]//The 3rd International Conference. San Diego:[s.n.], 2015.
[16] LONG J, SHELHAMER E, DARRELL T. Fully convolutional networks for semantic segmentation[J]. IEEE Thansactions onpattorn Analysis and Machine Intelligence, 2017, 29(4):640-654.
[17] SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for large-scale image recognition[C]//Computer Vision and Pattern Recognition.[S.l.]:Cornell University, 2014.
[18] HENSMAN P, MASKO D. The impact of imbalanced training data for convolutional neural networks[D].[S.l.]:KTH Royal Institute of Technology, 2015.
[19] KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks[C]//Advances in Neural Information Processing Systems 25.[S.l.]:Curran Associates, Inc., 2012:1097-1105.