Cloud Fraction of Satellite Imagery Based On Convolutional Neural Networks

doi:10.16182/j.issn1004731x.joss.201805001

Abstract

Abstract: Cloud fraction is the basis for the application of meteorological satellite. Existing methods cannot use all the characteristics and optical parameters of the satellite cloud, which results in the inaccuracy of cloud detection and cloud fraction. In order to solve this problem, convolutional neural network is used for cloud detection. Based on the improved convolutional neural network, the satellite cloud image is divided into thin cloud, thick cloud and clear sky. Based on the cloud detection, an improved spatial correlation method is used for cloud fraction. The results for Chinese HJ-1A/B satellite imagery show that convolutional neural network can extract the features of cloud images effectively by optimizing the network structure and parameters, and the transition region between the thin cloud and thick cloud is clear for cloud classification. The simulation results show that the cloud classification and cloud fraction accuracy is better than traditional threshold, dynamic threshold method and extreme learning machine.

Key words: cloud fraction, convolutional neural network, cloud detection, spatial correlation

CLC Number:

TP183

Xia Min, Shen Maoyang, Wang Jianfeng, Wang Yangguang. Cloud Fraction of Satellite Imagery Based On Convolutional Neural Networks[J]. Journal of System Simulation, 2018, 30(5): 1623-1630.

References

[1] Rossow W B, Garder L C.Cloud Detection Using Satellite Measurements of Infrared and Visible Radiances for ISCCP[J]. Journal of Climate (S0894-8755), 1994, 6(12): 2341-2369.
[2] Stowe L L, Davis P A, Mcclain E P.Scientific Basis and Initial Evaluation of the CLAVR-1 Global Clear/Cloud Classification Algorithm for the Advanced Very High Resolution Radiometer[J]. Journal of Atmospheric & Oceanic Technology (S0739-0572), 2010, 16(6): 656-681.
[3] Thomas S M, Heidinger A K, Pavolonis M J.Comparison of NOAA's Operational AVHRR-Derived Cloud Amount to Other Satellite-Derived Cloud Climatologies[J]. Journal of Climate (S0894-8755), 2004, 17(24): 4805-4822.
[4] Usman M, Liedl R, Shahid M A, et al.Land use/land cover classification and its change detection using multi-temporal MODIS NDVI data[J]. Journal of Geographical Sciences (S1009-637X), 2015, 25(12): 1479-1506.
[5] Wylie, D P, Menzel, W P.Two Years of Cloud Cover Statistics Using VAS[J]. Journal of Climate (S0894-8755), 1989, 2(4): 380-392.
[6] R W SAUNDERS, K T KRIEBEL.An improved method for detecting clear sky and cloudy radiances from AVHRR data[J]. International Journal of Remote Sensing (S0143-1161), 1988, 9(1): 123-150.
[7] Mackie S, Embury O, Old C.Generalized Bayesian cloud detection for satellite imagery. Part 1: Technique and validation for night-time imagery over land and sea[J]. International Journal of Remote Sensing (S0143-1161), 2010, 31(10): 2573-2594.
[8] 郑晓辉, 徐国强, 魏荣庆. GRAPES新云量计算方案的引进和影响试验[J]. 气象, 2013, 39(1): 57-66.
Zheng X H, Xu G Q, Wei R Q.Introducing and Influence Testing of the New Cloud Fraction Scheme in the GRAPES[J]. Meteorological, 2013, 39(1): 57-66.
[9] Maddy E S, King T S, Sun H, et al.Using Met Op-AAVHRR Clear-Sky Measurements to Cloud-Clear Met Op-AIASI Column Radiances[J]. Journal of Atmospheric & Oceanic Technology (S0739-0572), 2011, 28(28): 1104-1116.
[10] 周文君, 牛生杰, 许潇锋. 全天空成像仪云量计算方法的改进[J]. 大气科学学报, 2014, 37(3): 289-296.
Zhou W J, Niu S J, Xu X F.Improvements of computational methods for cloud cover based on the total sky imager[J]. Transactions of Atmospheric Sciences, 2014, 37(3): 289-296.
[11] 蔡强, 薛子育, 毛典辉, 等. 具有目标偏见的全局对比度显著性区域检测[J]. 系统仿真学报, 2015, 27(10): 2489-2496.
Cai Q, Xue Z Y, Mao D H, et al.Salient Region Detection based on Object-Biased Gaussian Refinement and Global Contrast[J]. Journal of System Simulation, 2015, 27(10): 2489-2496.
[12] 马芳, 张强, 郭铌, 等. 多通道卫星云图云检测方法的研究[J]. 大气科学, 2007, 31(1): 119-128.
Ma F, Zhang Q, Guo N, et al.The Study of Cloud Detection with Multi-Channel Data of Satellite[J]. Chinese Journal of Atmospheric Sciences, 2007, 31(1): 119-128.
[13] Barnes B B, Hu C.A Hybrid Cloud Detection Algorithm to Improve MODIS Sea Surface Temperature Data Quality and Coverage Over the Eastern Gulf of Mexico[J]. IEEE Transactions on Geoscience & Remote Sensing (S0196-2892), 2013, 51(6): 3273-3285.
[14] Li X L, Tian Y C, Xing K Z, et al.Cloud Extraction of Satellite Image Using Fuzzy C-Means Clustering Approach[J]. Applied Mechanics & Materials (S1662-7482), 2015, 743: 289-292.
[15] Shi B Q, Liang J, Liu Q.Adaptive simplification of point cloud using k-means clustering[J]. Computer-Aided Design (S0010-4485), 2011, 43(8): 910-922.
[16] Long C N, Sabburg J M, Calbó J, et al.Retrieving Cloud Characteristics from Ground-Based Daytime Color All-Sky Images[J]. Journal of Atmospheric & Oceanic Technology (S0739-0572), 2006, 23(5): 633-652.
[17] Gomez-Chova L, Camps-Valls G, Bruzzone L, et al.Mean Map Kernel Methods for Semisupervised Cloud Classification[J]. IEEE Transactions on Geoscience & Remote Sensing (S0196-2892), 2010, 48(1): 207-220.
[18] Taravat A, Del Frate F, Cornaro C, et al.Neural Networks and Support Vector Machine Algorithms for Automatic Cloud Classification of Whole-Sky Ground-Based Images[J]. IEEE Geoscience & Remote Sensing Letters (S1545-598X), 2015, 12(3): 666-670.
[19] J Alonso-Montesinos, M Martínez-Durbán, J del Sagrado, et al. The application of Bayesian network classifiers to cloud classification in satellite images[J]. Renewable Energy (S0960-1481), 2016, 97: 155-161.
[20] Tian H, Zhuang B, Hua Y, et al.Depth inference with convolutional neural network[C]// Visual Communications and Image Processing Conference. USA: IEEE, 2014.