[1]高军,荆益国.基于全卷积神经网络的卫星遥感图像云检测方法[J].红外技术,2019,41(7):607-615.[doi:10.11846/j.issn.1001_8891.201907003]
 GAO Jun,JING Yiguo.A Fully Convoluted Neural Network-based Cloud Detection Method for Satellite Remote Sensing Images[J].Infrared Technology,2019,41(7):607-615.[doi:10.11846/j.issn.1001_8891.201907003]
点击复制

基于全卷积神经网络的卫星遥感图像云检测方法
分享到:

《红外技术》[ISSN:1001-8891/CN:CN 53-1053/TN]

卷:
41卷
期数:
2019年第7期
页码:
607-615
栏目:
出版日期:
2019-07-20

文章信息/Info

Title:
A Fully Convoluted Neural Network-based Cloud Detection Method for Satellite Remote Sensing Images

文章编号:
1001-8891(2019)07-0607-09
作者:
高军荆益国
上海海事大学 信息工程学院
Author(s):
GAO JunJING Yiguo
College of Information Engineering, Shanghai Maritime University
关键词:
云检测遥感影像风云卫星全卷积神经网络
Keywords:
cloud detectionremote imageFengyun satellitefully convolutional network
分类号:
TP389.1
DOI:
10.11846/j.issn.1001_8891.201907003
文献标志码:
A
摘要:
云检测作为遥感影像数据处理中的重要组成部分,在气候分析等各个方面起到了重要的作用。在云检测研究中,无论是应用广泛的阈值法或是基于模式识别的方法,以及在二者基础上的综合分析法。这些方法大多都依赖于单一类型的遥感数据来源,且在特征提取方面十分依赖先验知识,受主观影响较大。本文利用两种不同类型“风云”系列气象遥感卫星的可见光红外扫描辐射计(Visible and Infrared Radiometer,VIRR)以及多通道扫描成像辐射计(Advanced Geosynchronous Radiation Imager,AGRI)数据,以全卷积神经网络为基础进行云检测,利用其自动提取深层隐含特征等特性,极大保留特征信息。最后结合全连接条件随机场模型进行云系边缘优化。实验结果表明,该算法分别应用于以上两种不同类型遥感影像数据,都较好地完成了云像元和非云像元的分离。
Abstract:
Cloud detection is an important aspect of remote sensing image data processing, which plays a key role in climate analysis and other relevant aspects. In the research of cloud detection, researchers often use the threshold method or the method based on pattern recognition; furthermore, both may be combined in a comprehensive analysis method. Most of these methods rely on a single type of remote sensing data source. In terms of feature extraction, they rely heavily on prior knowledge of the researcher and are subjectively influenced. In this paper, the VIRR and AGRI data of two different types of “Fengyun” series meteorological remote sensing satellites has been used to perform cloud detection based on the fully convoluted neural network. It can automatically extract deeply hidden features, while retaining feature information. Finally, cloud edge optimization was carried out with a fully connected conditional random field model. Cloud detection results from our experiments employing the proposed method on two different types of remote sensing image data achieved separation of cloud and non-cloud pixels.

参考文献/References:

[1] 林晔. 大气探测学教程[M]. 北京: 气象出版社, 1995.
LIN Ye. Course of Atmospheric Exploration[M]. Beijing: Meteorological Publishing House, 1995.
[2] Rossow W B. Cloud detection using satellite measurements of infrared and visible radiances for ISCCP[J]. Journal of Climate, 1993, 12(12): 2341-2369.
[3] 冯春. 基于辐射传输模型的遥感定量化关键关键问题研究[D]. 北京: 中国地质大学, 2005.
FENG Chun. Research on Key Key Issues of Remote Sensing Quantification Based on Radiation Transfer Model[D]. Beijing: China University of Geoscience, 2005.
[4] WYLIE D P, MENZEL W P, WOOLF H M, et al. Four years of global cirrus cloud statistics using HIRS[J]. J. Climat, 1994, 7(12): 1972-1986.
[5] Kegelmeyer W P Jr. Extraction of Cloud Statistics from Whole Sky Imaging Caemeras[R]. Livermore, CA, United States: Sandia National Labs, 1994.
[6] Ackerman S, Strabala K, Menzel W P, et a1. Discriminatingclear-sky from clouds with MODIS[J]. Journal of Geophysical Research, 1998, 103: 141-157.
[7] Solvsteen A C. Correlation based cloud-detection and an examination of the split-window method[J]. Proceedings of SPIE - The International Society for Optical Engineering, 1995: 86-97.
[8] Baum B A, Trepte Q. A Grouped Threshold Approach for Scene Identification in AVHRR Imagery[J]. Journal of Atmospheric & Oceanic Technology, 1999, 16(6): 793-800.
[9] 刘希, 许健民, 杜秉玉. 用双通道动态阈值对GMS-5图像进行自动云检测[J]. 应用气象学报, 2005(4): 434-444.
LIU Xi, XU Jianmin, DU Bingyu. ABI-Channel Dynamic Threshold Algorithm Usedin Automatically Identifying Clouds on GMS-5 Imagery[J]. Journal of Applied Meteorological Science, 2005(4): 434-444.
[10] LI W, LI D. The universal cloud detection algorithm of MODIS data[C]//Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, 2006: 64190F-64190F-6.
[11] Azimi-Sadjadi M R, Shaikh M A, Tian B, et al. Neural network-based cloud detection/classification using textural and spectral features[C]//Geoscience and Remote Sensing Symposium, IEEE, 1996: 1105-1107.
[12] Vasquez R, Manian V. Texture based cloud detection inMODIS images[C]//The International Society for Optical Engineering, Proceedings of SPIE, 2002, 4882: 259-267.
[13] 陈鹏, 张荣, 刘政凯. 遥感图像云图识别中的特征提取[J]. 中国科学技术大学学报, 2009, 39(5):484-488.
CHEN Peng, ZHANG Rong, LIU Zhengkai. Feature extraction in cloud image recognition from remote sensing images[J]. Journal of University of Science and Technology of China, 2009, 39(5): 484-488.
[14] 王丹凤, 张记龙, 王志斌,等. 基于MODIS云产品的AIRS像素云检测[J]. 国土资源遥感, 2013, 25(1): 13-17.
WANG Danfeng, ZHANG Jilong, WANG Zhibin, et al. AIRS pixel cloud detection based on MODIS cloud products[J]. Remote Sensing For Land & Resources, 2013, 25(1): 13-17.
[15] 陈振炜, 张过, 宁津生, 等. 资源三号测绘卫星自动云检测[J]. 测绘学报, 2015, 44(3): 292-300.
CHEN Zhenwei, ZHANG Guo, NING Jinsheng, et al. Automated Cloud Detection of Resource 3 Surveying and Mapping Satellite[J]. Acta Geodaetica et Cartographica Sinica, 2015, 44(3): 292-300.
[16] 曹琼, 郑红, 李行善. 一种基于纹理特征的卫星遥感图像云探测方法[J]. 航空学报, 2007, 28(3): 661-666.
CAO Qiong, ZHENG Hong, LI Xingshan. A Cloud Detection Method for Satellite Remote Sensing Images Based on Texture Features[J]. Acta Aeronautica ET Astronautica Sinica, 2007, 28(3): 661-666.
[17] Jedlovec G J, Haines S L, Lafontaine F J. Spatial and Temporal Varying Thresholds for Cloud Detection in GOES Imagery[J]. IEEE Transactions on Geoscience & Remote Sensing, 2008, 46(6): 1705-1717.
[18] 单娜, 郑天垚, 王贞松. 快速高准确度云检测算法及其应用[J]. 遥感学报, 2009, 13(6): 1138-1155.?
SHAN Na, ZHENG Tianbiao, WANG Zhensong. Fast and high accuracy cloud detection algorithm and its application[J]. Journal of Remote Sensing, 2009, 13(6): 1138-1155.
[19] Amato U, Antoniadis A, Cuomo V, et al. Statistical cloud detection from SEVIRI multispectral images[J]. Remote Sensing of Environment, 2008, 112(3): 750-766.
[20] Merchant C J, Harris A R, Maturi E, et al. Probabilistic physically based cloud screening of satellite infrared imagery for operational sea surface temperature retrieval[J]. Quarterly Journal of the Royal Meteorological Society, 2010, 131(611): 2735-2755.
[21] 盛夏, 孙龙祥, 郑庆梅. 利用MODIS数据进行云检测[J]. 解放军理工大学自然科学版, 2004, 5(4): 98-102.
SHENG Xia, SUN Longxiang, ZHENG Qingmei. Cloud detection using MODIS data[J]. Natural Science Edition of PLA University of Technology, 2004, 5(4): 98-102.
[22] GomezChova L, CampsValls G, AmorosLopez J, et al. New Cloud Detection Algorithm for Multispectral and Hyperspectral Images: Application to ENVISAT/MERIS and PROBA/CHRIS Sensors[C]// IEEE International Conference on Geoscience and Remote Sensing Symposium, 2007: 2757-2760.
[23] 潘聪, 夏斌, 陈彧, 等. 基于模糊聚类的MODIS云检测算法研究[J]. 微计算机信息, 2009(4): 124-125.
PAN Cong, XIA Bin, CHEN Yu, et al. Research on MODIS Cloud Detection Algorithms Based on Fuzzy Clustering[J]. Microcomputer Information, 2009(4): 124-125.
[24] ZHANG W D, HE M X, Mak M W. Cloud detection using probabilistic neural networks[C]//Geoscience and Remote Sensing Symposium, IEEE, 2001: 2373-2375.
[25] 夏旻, 申茂阳, 王舰锋, et al. 基于卷积神经网络的卫星云图云量计算[J]. 系统仿真学报, 2018, 30(5): 7-14.
XIA Min, SHEN Maoyang, WANG JianFeng, et al. Cloud Fraction of Satellite Imagery Based On Convolutional Neural Networks[J]. Journal of System Simulation, 2018, 30(5): 7-14.
[26] Ricciardelli E, Romano F, Cuomo V. Physical and statistical approaches for cloud identification using Meteosat Second Generation-Spinning Enhanced Visible and Infrared Imager Data[J]. Remote Sensing of Environment, 2008, 112(6): 2741-2760.
[27] 康晓光, 孙龙祥. 基于人工神经网络的云自动检测算法[J]. 解放军理工大学学报:自然科学版, 2005, 6(5): 506-510.
KANG Xiaoguang, SUN Longxiang. Cloud automatic detection algorithm based on artificial neural network[J]. Journal of PLA University of Technology: Natural Science Edition, 2005, 6(5): 506-510.
[28] 靳泽群, 张玲, 刘神聪,等. 基于BP神经网络的云检测和云相态识别[J]. 光学与光电技术, 2016, 14(5):74-77.
JIN Zequn, ZHANG Ling, LIU Shencong, et al. Cloud Detection and Cloud Phase Recognition Based on BP Neural Network[J]. Optical & Optical Technology, 2016, 14(5): 74-77.
[29] LIU Y, XIA J, SHI C X, et al. An Improved Cloud Classification Algorithm for China’s FY-2C Multi-Channel Images Using Artificial Neural Network[J]. Sensors, 2009, 9(7): 5558-5579.
[30] Christodoulou C I, Michaelides S C, Pattichis C S. Multifeature texture analysis for the classification of clouds in satellite imagery[J]. IEEE Transactions on Geoscience & Remote Sensing, 2003, 41(11): 2662-2668.
[31] Tomun I S, Kwiatkowska E. Neural network system forcloud classification from satellite images[C]//Proceedings of the International Joint Conference on Neural Networks, 1999, 6: 3785-3790.
[32] LI P, DONG L, XIAO H, et al. A cloud image detection method based on SVM vector machine[J]. Neurocomputing, 2015, 169: 34-42.
[33] SHAO Z, DENG J, WANG L, et al. Fuzzy AutoEncode Based Cloud Detection for Remote Sensing Imagery[J]. Remote Sensing, 2017, 9(4): 311.
[34] Reguiegue M, Chouireb F. Automatic day time cloud detection over land and sea from MSG SEVIRI images using three features and two artificial intelligence approaches[J]. Signal Image & Video Processing, 2017, 12(3): 1-8.
[35] Choi H, Bindschadler R. Cloud detection in Landsat imagery of ice sheets using shadow matching technique and automatic normalized difference snow index threshold value decision[J]. Remote Sensing of Environment, 2004, 91(2): 237-242.
[36] 刘志刚, 李元祥, 黄峰. 基于动态聚类的MODIS云检测算法[J]. 遥感信息, 2007(4): 33-35.
LIU Zhigang, LI Yuanxiang, HUANG Feng. MODIS cloud detection algorithm based on dynamic clustering[J]. Remote Sensing Information, 2007(4): 33-35.
[37] 王伟, 宋卫国, 刘士兴, 等. Kmeans聚类与多光谱阈值相结合的MODIS云检测算法[J]. 光谱学与光谱分析, 2011, 31(4): 1061-1064.
WANG Wei, SONG Weiguo, LIU Shixing, et al. A Cloud Detection Algorithm for MODIS Images Combining Kmeans Clustering and Multi-Spectral Threshold Method[J]. Spectroscopy and spectral analysis, 2011, 31(4): 1061-1064.
[38] HE K, ZHANG X, REN S, et al. Identity Mappings in Deep Residual Networks[J]. Computer Vision – ECCV, 2016, 9908: 630-645.
[39] HUANG G, LIU Z, Laurens V D M, et al. Densely Connected Convolutional Networks[J]. IEEE Conference on Computer Vision & Pattern Recognition, 2017: 2261-2269.
[40] Jégou, Simon, Drozdzal M , Vazquez D , et al. The One Hundred Layers Tiramisu: Fully Convolutional DenseNets for Semantic Segmentation[J]. IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2017: 1175-1183.
[41] Kr?henbühl, Philipp, Koltun V . Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials[J]. NIPS’11 Proceedings of the 24th International Conference on Neural Information Processing Systems, 2012: 109-117.
[42] 高军, 王恺, 田晓宇, 等. 基于BP神经网络的风云四号遥感图像云检测算法[J]. 红外与毫米波学报, 2018, 37(4): 477-485.
GAO Jun, WANG Kai, TIAN Xiaoyu, et al. A BP-NN based cloud detection method for FY-4 remote sensing images[J]. Journal of Infrared and Millimeter Wave, 2018, 37(4): 477-485.
[43] Griggin M, Burke H H, Mandl D, et al. Cloud cover detection algorithm for EO-1 Hyperion imagery[C]//Geoscience and Remote Sensing Symposium, IEEE, 2003: 86-89.

备注/Memo

备注/Memo:
收稿日期:2019-03-08;修订日期:2019-07-04.
作者简介:高军(1979-),男,浙江嘉兴人,博士,主要从事遥感信息处理、网络通信方面的研究。E-mail:jungao@shmtu.edu.cn。
基金项目:国家自然科学基金项目(61602296)。

更新日期/Last Update: 2019-07-12