多层感知机结合辐射传输模型的复杂陆地表面云检测

doi:10.12263/DZXB.20210636

电子学报 ›› 2022, Vol. 50 ›› Issue (4): 932-942.DOI: 10.12263/DZXB.20210636

所属专题：机器学习交叉融合创新

• 机器学习交叉融合创新 • 上一篇下一篇

多层感知机结合辐射传输模型的复杂陆地表面云检测

邓梦娇¹, 徐新¹^,², 马盈盈³, 龚威³, 金适宽³, 胡瑞敏⁴

^1.武汉科技大学计算机科学与技术学院，湖北武汉 430065
^2.武汉科技大学智能信息处理与实时工业系统重点实验室，湖北武汉 430065
^3.武汉大学测绘遥感信息工程国家重点实验室，湖北武汉 430072
^4.武汉大学计算机学院，湖北武汉 430072

收稿日期:2021-07-17 修回日期:2022-03-07 出版日期:2022-04-25
- 作者简介:
- 邓梦娇女，1998年生，湖北咸宁人.现为武汉科技大学硕士研究生.主要研究方向为大气遥感云检测. E-mail: dengmjiao@qq.com
  徐新（通讯作者）男，1982年生，湖北武汉人.博士，武汉科技大学计算机科学与技术学院教授、博士生导师.主要研究方向为计算机视觉、人工智能、大气遥感. E-mail: xuxin@wust.edu.cn
- 基金资助:
- 国家自然科学基金 (U1803262)

Multi-layer Perceptron Combined with Radiative Transfer Model for Complex Land Surface Cloud Detection

DENG Meng-jiao¹, XU Xin¹^,², MA Ying-ying³, GONG Wei³, JIN Shi-kuan³, HU Rui-min⁴

^1.School of Computer Science and Technology, Wuhan University of Science and Technology, Wuhan, Hubei 430065, China
^2.Hubei Province Key Laboratory of Intelligent Information Processing and Real-time Industrial System, Wuhan University of Science and Technology, Wuhan, Hubei 430065, China
^3.State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan, Hubei 430072, China
^4.School of Computer Science and Technology, Wuhan University, Wuhan, Hubei 430072, China

Received:2021-07-17 Revised:2022-03-07 Online:2022-04-25 Published:2022-04-25
- Supported by:
- National Natural Science Foundation of China (U1803262)

摘要/Abstract

摘要：

云检测是卫星遥感数据预处理中至关重要的工作.本文将多层感知机和辐射传输模型相结合，利用可见光和近红外波段反射率信息从卫星影像中识别出云像元.该方法利用SBDART辐射传输模型，模拟获得了各种复杂陆地地表的反射率值数据集，为多层感知机提供训练样本.随后，用训练好的多层感知机模型区分FY-3D卫星MERSI II影像中的云像元和非云像元，利用CALIPSO垂直特性掩膜产品（Vertical Feature Mask，VFM）逐像元进行验证，并与MODIS云掩膜产品（MYD35）进行横向对比.结果表明，以VFM数据集为标准的情况下，多层感知机识别云的总正确率为76.25%，其中在夏季和低纬度地区效果最好，如赤道附近地表识别的准确率可达到91.74%，而在城市、农田和裸地等复杂地表类型条件下的云检测识别正确率分别为83.37%、84.52%和73.11%，分别高于MYD35产品的83.25%、83.31%和72.66%.为了进一步验证多层感知机结合辐射传输模型云检测方法的有效性，将辐射传输模型模拟得到的训练样本分别用于k-最近邻、朴素贝叶斯以及随机森林算法，并与本文多层感知机算法进行对比.结果表明，将多层感知机和辐射传输模型相结合具有更高的正确率.

关键词: 云检测, 多层感知机, 辐射传输模拟, MERSI II, MODIS

Abstract:

Cloud detection is a key step in the preprocessing of satellite remote sensing data. This paper proposes a cloud detection method by combining a multilayer perceptron with a radiative transfer model. The method is to identify cloud from moderate resolution satellite image using visible and near-infrared band reflectance information. In this method, firstly, the santa barbara DISORT atmospheric radiative transfer model(SBDART) is used to simulate and obtain datasets of reflectance values for a variety of complex terrestrial surfaces, which provides training samples for the multilayer perceptron. Secondly, the trained network model is used to distinguish cloud pixels from total pixels of the advanced medium Resolution Spectral Imager(MERSI II) image in the FengYun3D satellite MERSI II image, and then verified using vertical feature mask(VFM) product of the cloud-aerosol LIDAR infrared pathfinder satellite observations satellite(CALIPSO) and compared horizontally with the cloud mask product(MYD35) of the moderate resolution imaging spectroradiometer(MODIS). The results show that the accuracy of cloud detection for the multilayer perceptron is 76.25%, and especially this method works best in summer and low latitudes, achieves an accuracy of 91.74% for surface identification near the equator. In this paper, the method is more effective in detecting clouds under complex surface type conditions such as urban, farmland and bare soil, with accuracies of 83.37%, 84.52% and 73.11% respectively, which are higher than the 83.25%, 83.31% and 72.66% of the MYD35 product respectively. To further validate the effectiveness of the multilayer perceptron combined with the radiative transfer model, the training samples obtained from the radiative transfer model simulations are used in the k-nearest neighbors, Naive Bayesian, and Random Forest algorithms, respectively, and compared with the multilayer perceptron algorithm in this paper. The results show that the combination of the multilayer perceptron and the radiative transfer model has a higher accuracy.

Key words: cloud detection, multilayer perceptron, radiative transfer simulation, MERSI II, MODIS

中图分类号:

TP751

邓梦娇, 徐新, 马盈盈, 等. 多层感知机结合辐射传输模型的复杂陆地表面云检测[J]. 电子学报, 2022, 50(4): 932-942.

Meng-jiao DENG, Xin XU, Ying-ying MA, et al. Multi-layer Perceptron Combined with Radiative Transfer Model for Complex Land Surface Cloud Detection[J]. Acta Electronica Sinica, 2022, 50(4): 932-942.

图/表 9

参考文献 44

1	ZHUZ, WOODCOCKC E. Object-based cloud and cloud shadow detection in Landsat imagery[J]. Remote Sensing of Environment, 2012, 118: 83-94.
2	JINS, ZHANGM, MAY, et al. Adapting the dark target algorithm to advanced MERSI sensor on the FengYun-3-D satellite: Retrieval and Validation of Aerosol Optical Depth Over Land[J]. IEEE Transactions on Geoscience and Remote Sensing, 2021, 59(10): 8781-8797.
3	胡根生, 查慧敏, 梁栋, 等. 结合分类与迁移学习的薄云覆盖遥感图像地物信息恢复[J]. 电子学报, 2017, 45(12): 2855-2862.
	HUG S, ZHAH M, LIANGD, et al. Ground object information recovery for thin cloud contaminated remote sensing images by combining classification with transfer learning[J]. Acta Electronica Sinica, 2017, 45(12): 2855-2862.(in Chinese)
4	JEDLOVECG J, HAINESS L, LAFONTAINEF J. Spatial and temporal varying thresholds for cloud detection in GOES imagery[J]. IEEE Transactions on Geoscience and Remote Sensing, 2008, 46(6): 1705-1717.
5	HAGOLLEO, HUC M, PASCUALD V, et al. A multi-temporal method for cloud detection, applied to FORMOSAT-2, VENμS, LANDSAT and SENTINEL-2 images[J]. Remote Sensing of Environment, 2010, 114(8): 1747-1755.
6	SAUNDERSR W, KRIEBELK T. An improved method for detecting clear sky and cloudy radiances from AVHRR data[J]. International Journal of Remote Sensing, 1988, 9(1): 123-150.
7	KRIEBELK-T, GESELLG, STNERM KA, et al. The cloud analysis tool APOLLO: Improvements and validations[J]. International Journal of Remote Sensing, 2003, 24(12): 2389-2408.
8	WEIJ, HUANGW, LIZ Q, et al. Cloud detection for Landsat imagery by combining the random forest and superpixels extracted via energy-driven sampling segmentation approaches[J]. Remote Sensing of Environment, 2020, 248: 1-14.
9	LIY, CHENW, ZHANGY, et al. Accurate cloud detection in high-resolution remote sensing imagery by weakly supervised deep learning[J]. Remote Sensing of Environment, 2020, 250: 1-18.
10	YANGZ, ZHANGP, GUS, et al. Capability of Fengyun-3D satellite in earth system observation[J]. Journal of Meteorological Research, 2019, 33(6): 1113-1130.
11	ROSSOWW B, GARDERL C. Cloud detection using satellite measurements of infrared and visible radiances for ISCCP[J]. Journal of Climate, 1993, 6(12): 2341-2369.
12	IRISHR R. Landsat 7 automatic cloud cover assessment[C]//Algorithms for Multispectral, Hyperspectral, and Ultraspectral Imagery VI. Orlando: International Society for Optics and Photonics, 2000: 348-355.
13	IRISHR R, BARKERJ L, GOWARDS N, et al. Characterization of the Landsat-7 ETM+ automated cloud-cover assessment (ACCA) algorithm[J]. Photogrammetric Engineering & Remote Sensing, 2006, 72(10): 1179-1188.
14	ACKERMANS A, STRABALAK I, MENZELW P, et al. Discriminating clear sky from clouds with MODIS[J]. Journal of Geophysical Research: Atmospheres, 1998, 103(D24): 32141-32157.
15	SUNL, WEIJ, WANGJ, et al. A universal dynamic threshold cloud detection algorithm(UDTCDA) supported by a prior surface reflectance database[J]. Journal of Geophysical Research: Atmospheres, 2016, 121(12): 7172-7196.
16	VERMOTEE F, TANRÉD, DEUZEJ L, et al. Second simulation of the satellite signal in the solar spectrum, 6S: an overview[J]. IEEE Transactions on Geoscience and Remote Sensing, 1997, 35(3): 675-686.
17	VERMOTEE, TANRÉD, DEUZÉJ L, et al. Second simulation of a satellite signal in the solar spectrum-vector (6SV)[J]. 6S User Guide Version, 2006, 3(2): 1-55.
18	WANGH, HEY, GUANH. Application of support vector machines in cloud detection using EOS/MODIS[C]//Remote Sensing Applications for Aviation Weather Hazard Detection and Decision Support. San Diego: International Society for Optics and Photonics, 2008: 1-8.
19	MENGF, YANGX, ZHOUC, et al. A sparse dictionary learning-based adaptive patch inpainting method for thick clouds removal from high-spatial resolution remote sensing imagery[J]. Sensors, 2017, 17(9): 1-16.
20	徐少平, 林珍玉, 张贵珍, 等. 采用深度学习与图像融合混合实现策略的低照度图像增强算法[J]. 电子学报, 2021, 49(1): 72-76.
	XUS P, LINZ Y, ZHANGG Z, et al. A low-light image enhancement algorithm using the hybrid strategy of deep learning and image fusion[J]. Acta Electronica Sinica, 2021, 49(1): 72-76. (in Chinese)
21	周涛, 霍兵强, 陆惠玲, 等. 残差神经网络及其在医学图像处理中的应用研究[J]. 电子学报, 2020, 48(7): 1436-1447.
	ZHOUT, HUOB Q, LUH L, et al. Research on residual neural network and its application on medical image processing[J]. Acta Electronica Sinica, 2020, 48(7): 1436-1447. (in Chinese)
22	刘启超, 肖亮, 刘芳, 等. SSCDenseNet: 一种空-谱卷积稠密网络的高光谱图像分类算法[J]. 电子学报, 2020, 48(4): 751-762.
	LIUQ C, XIAOL, LIUF, et al. SSCDenseNet: A spectral-spatial convolutional dense network for hyperspectral image classification[J]. Acta Electronica Sinica, 2020, 48(4): 751-762. (in Chinese)
23	XIEF, SHIM, SHIZ, et al. Multilevel cloud detection in remote sensing images based on deep learning[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2017, 10(8): 3631-3640.
24	SEGAL-ROZENHAIMERM, LIA, DAS K, et al. Cloud detection algorithm for multi-modal satellite imagery using convolutional neural-networks(CNN)[J]. Remote Sensing of Environment, 2020, 237: 1-17.
25	LONGJ, SHELHAMERE, DARRELLT. Fully convolutional networks for semantic segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Boston: IEEE, 2015: 3431-3440.
26	CHAID, NEWSAMS, ZHANGH K, et al. Cloud and cloud shadow detection in Landsat imagery based on deep convolutional neural networks[J]. Remote Sensing of Environment, 2019, 225: 307-316.
27	MATEO-GARCÍAG, LAPARRAV, LÓPEZ-PUIGDOLLERSD, et al. Transferring deep learning models for cloud detection between Landsat-8 and Proba-V[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2020, 160: 1-17.
28	LIZ, SHENH, CHENGQ, et al. Deep learning based cloud detection for medium and high resolution remote sensing images of different sensors[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2019, 150: 197-212.
29	WIELANDM, LIY, MARTINISS. Multi-sensor cloud and cloud shadow segmentation with a convolutional neural network[J]. Remote Sensing of Environment, 2019, 230: 1-12.
30	YANGJ, GUOJ, YUEH, et al. CDnet: CNN-based cloud detection for remote sensing imagery[J]. IEEE Transactions on Geoscience and Remote Sensing, 2019, 57(8): 6195-6211.
31	LUOTAMOM, METSÄMÄKIS, KLAMIA. Multiscale cloud detection in remote sensing images using a dual convolutional neural network[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 59(6): 4972-4983.
32	RICCHIAZZIP, YANGS, GAUTIERC, et al. SBDART: a research and teaching software tool for plane-parallel radiative transfer in the earth's atmosphere[J]. Bulletin of the American Meteorological Society, 1998, 79(10): 2101-2114.
33	STAMNESK, HAMREB, STAMNESJ J, et al. Modeling of radiation transport in coupled atmosphere-snow-ice-ocean systems[J]. Journal of Quantitative Spectroscopy and Radiative Transfer, 2011, 112(4): 714-726.
34	ATKINSONP M, TATNALLA R L. Introduction neural networks in remote sensing[J]. International Journal of Remote Sensing, 1997, 18(4): 699-709.
35	HET, DONGZ, MENGK, et al. Accelerating multi-layer perceptron based short term demand forecasting using graphics processing units[C]//Proceedings of the 2009 Transmission & Distribution Conference & Exposition: Asia and Pacific. Seoul: IEEE, 2009: 1-4.
36	任进军, 王宁. 人工神经网络中损失函数的研究[J]. 甘肃高师学报, 2018, 23(2): 61-63.
	RENJ J, WANGN. Research on cost function in artificial neural network[J]. Journal of Gansu Normal Colleges, 2018, 23(2): 61-63. (in Chinese)
37	XUN, NIUX, HUX, et al. Prelaunch calibration and radiometric performance of the advanced MERSI II on FengYun-3D[J]. IEEE Transactions on Geoscience and Remote Sensing, 2018, 56(8): 4866-4875.
38	SALOMONSONV V, BARNESW, MASUOKAE J. Introduction to MODIS and An Overview of Associated Activities[M]. Berlin: Springer, 2006: 12-32.
39	GAOB C, GOETZA F H, WISCOMBEW J. Cirrus cloud detection from airborne imaging spectrometer data using the 1.38 μm water vapor band[J]. Geophysical Research Letters, 1993, 20(4): 301-304.
40	GUOG, WANGH, BELLD, et al. KNN model-based approach in classification[C]//On The Move to Meaningful Internet Systems Confederated International Conferences. Berlin: Springer, 2003: 986-996.
41	SUCARL E. Probabilistic Graphical Models[M]. Berlin: Springer, 2021: 43-69.
42	PARMARA, KATARIYAR, PATELV. A review on random forest: An ensemble classifier[C]//International Conference on Intelligent Data Communication Technologies and Internet of Things. Coimbatore: Springer, 2018: 1-6.
43	ZHUY, MAY, LIUB, et al. Retrieving the vertical distribution of PM2.5 mass concentration from lidar via a random forest Model[J]. IEEE Transactions on Geoscience and Remote Sensing, 2021, 60: 1-9.
44	陈喆, 贾春福, 宗楠, 等. 随机森林在程序分支混淆中的应用[J]. 电子学报, 2018, 46(10): 2458-2466.
	CHENZ, JINC F, ZONGN, et al. Branch obfuscation using random forest[J]. Acta Electronica Sinica, 2018, 46(10): 2458-2466. (in Chinese)

参数名	取值范围	单位
太阳天顶角	0~90	(°)
卫星天顶角	0~90	(°)
相对方位角	0~180	(°)
气溶胶光学厚度	0~3
云光学厚度	1~50
云滴有效半径	5~40	μm
下垫面高程	0~5	km

参数名	取值范围	单位
太阳天顶角	0~90	(°)
卫星天顶角	0~90	(°)
相对方位角	0~180	(°)
气溶胶光学厚度	0~3
云光学厚度	1~50
云滴有效半径	5~40	μm
下垫面高程	0~5	km

数据	数据用处	下载链接
MCD43C1和MCD12C1产品	辐射传输模型模拟的输入参数	https://ladsweb.modaps.eosdis.nasa.gov/
MERSI II真实观测数据	多层感知机模型的测试数据样本	http://www.nsmc.org.cn/nsmc/cn/home/index.html
VFM产品	多层感知机模型的测试数据标签	https://eosweb.larc.nasa.gov/

数据	数据用处	下载链接
MCD43C1和MCD12C1产品	辐射传输模型模拟的输入参数	https://ladsweb.modaps.eosdis.nasa.gov/
MERSI II真实观测数据	多层感知机模型的测试数据样本	http://www.nsmc.org.cn/nsmc/cn/home/index.html
VFM产品	多层感知机模型的测试数据标签	https://eosweb.larc.nasa.gov/

地表类型	多层感知机		MYD35产品
地表类型	正确率/%	样本数	正确率/%	样本数
常绿针叶林	82.22	48 575	81.86	26 725
常绿阔叶林	87.89	189 812	85.32	119 952
落叶针叶林	75.14	23 484	76.56	11 749
落叶阔叶林	78.14	146 206	79.07	84 908
混合森林	81.15	300 831	80.28	168 210
浓密灌木丛	77.32	1 036	91.76	1 141
稀疏灌木丛	79.23	39 085	78.41	43 289
热带稀树草原	78.79	465 934	79.78	273 813
热带草原	80.35	318 991	80.66	189 121
草地	71.44	1 058 269	75.97	676 581
永久性湿地	83.38	10 911	83.19	8 376
农田	77.31	632 714	78.69	420 477
城镇用地	83.37	21 165	83.25	15 016
天然植被和农田	84.52	46 529	83.31	32 477
冰雪地	75.09	5 901	80.54	5 606
裸土或稀疏植被	73.11	553 863	72.66	369 922

多层感知机结合辐射传输模型的复杂陆地表面云检测

Multi-layer Perceptron Combined with Radiative Transfer Model for Complex Land Surface Cloud Detection

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 9

参考文献 44

相关文章 15

编辑推荐

Metrics

[1]	王金伟, 胡冰涛, 张家伟, 马宾, 罗向阳. 基于解压缩模块的JPEG同步重压缩检测[J]. 电子学报, 2023, (): 1-10.
[2]	李滔, 董秀成, 林宏伟. 基于深监督跨尺度注意力网络的深度图像超分辨率重建[J]. 电子学报, 2023, 51(1): 128-138.
[3]	吴靖, 叶晓晶, 黄峰, 陈丽琼, 王志锋, 刘文犀. 基于深度学习的单帧图像超分辨率重建综述[J]. 电子学报, 2022, 50(9): 2265-2294.
[4]	任莎莎, 刘琼. 小目标特征增强图像分割算法[J]. 电子学报, 2022, 50(8): 1894-1904.
[5]	张云, 化青龙, 姜义成, 徐丹. 基于混合型复数域卷积神经网络的三维转动舰船目标识别[J]. 电子学报, 2022, 50(5): 1042-1049.
[6]	李玉, 王姝运, 赵泉华. 基于斑点统计特性保持的SAR影像迭代滤波[J]. 电子学报, 2021, 49(9): 1809-1817.
[7]	薛丹, 苑玮琦. 基于空间拓扑和形态相似关系的长裂缝拼接算法[J]. 电子学报, 2021, 49(4): 672-678.
[8]	张绍泉, 黄志浩, 邓承志, 李璠, 徐晨光, 吴朝明, 汪胜前. 光谱加权协同稀疏和全变差正则化高光谱图像解混[J]. 电子学报, 2020, 48(12): 2453-2461.
[9]	盖杉, 王俊生. 基于深度学习的非局部注意力增强网络图像去雨算法研究[J]. 电子学报, 2020, 48(10): 1899-1908.
[10]	黄鸿, 徐科杰, 石光耀. 联合多尺度多特征的高分遥感图像场景分类[J]. 电子学报, 2020, 48(9): 1824-1833.
[11]	秦福强, 王丽芳. 基于变分的多尺度遥感图像融合算法[J]. 电子学报, 2020, 48(6): 1084-1090.
[12]	黄鸿, 王丽华, 石光耀. 面向高光谱遥感影像分类的监督多流形鉴别嵌入方法[J]. 电子学报, 2020, 48(6): 1099-1107.
[13]	刘启超, 肖亮, 刘芳, 徐金环. SSCDenseNet:一种空-谱卷积稠密网络的高光谱图像分类算法[J]. 电子学报, 2020, 48(4): 751-762.
[14]	石雪, 李玉, 赵泉华. 自适应类别的层次高斯混合模型遥感影像分割[J]. 电子学报, 2020, 48(1): 131-136.
[15]	李庆武, 马云鹏, 周亚琴, 邢俊. 基于无监督栈式降噪自编码网络的显著性检测算法[J]. 电子学报, 2019, 47(4): 871-879.

方法	正确率/%	F1/%
RF	69.51	79.27
NB	73.39	79.46
KNN	74.27	79.28
MLP	76.25	82.32

方法	正确率/%	F1/%
RF	69.51	79.27
NB	73.39	79.46
KNN	74.27	79.28
MLP	76.25	82.32