基于深度神经网络的蒙古国色楞格河流域水体信息提取

doi:10.12082/dqxxkx.2022.210031

地球信息科学学报 ›› 2022, Vol. 24 ›› Issue (5): 1009-1017.doi: 10.12082/dqxxkx.2022.210031

基于深度神经网络的蒙古国色楞格河流域水体信息提取

姚锦一¹^,²(), 王卷乐¹^,⁴^,^*(), 严欣荣¹^,³, 魏海硕¹, Altansukh Ochir⁵, Davaadorj Davaasuren⁶

1.中国科学院地理科学与资源研究所中国科学院资源与环境信息系统国家重点实验室,北京 100101
2.山东理工大学建筑工程学院, 淄博 255049
3.中国科学院大学,北京 100049
4.江苏省地理信息资源开发与利用协同创新中心,南京 210023
5.蒙古国立大学工程与应用科学学院,乌兰巴托 210646
6.蒙古国立大学艺术与科学学院,乌兰巴托 210646

收稿日期:2021-01-20 修回日期:2021-03-20 出版日期:2022-05-25 发布日期:2022-07-25
通讯作者: * 王卷乐（1976— ）,男,河南洛阳人,博士,研究员,主要从事资源环境科学数据集成与共享研究。 E-mail: wangjl@igsnrr.ac.cn
作者简介:姚锦一（1996— ）,男,江苏苏州人,硕士生,主要从事遥感科学与技术方面的研究。E-mail: yaojy@lreis.ac.cn
基金资助:
国家自然科学基金面上项目(41971385);中国科学院A类战略性先导科技专项(XDA2003020302);中国工程科技知识中心建设项目(CKCEST-2021-2-18)

Water Information Extraction of Selenga River Basin in Mongolia based on Deep Neural Network

YAO Jinyi¹^,²(), WANG Juanle¹^,⁴^,^*(), YAN Xinrong¹^,³, WEI Haishuo¹, Altansukh Ochir⁵, Davaadorj Davaasuren⁶

1. State Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China
2. School of Civil and Architectural Engineering, Shandong University of Technology, Zibo 255049, Shandong
3. China University of Chinese Academy of Sciences, Beijing 100049, China
4. Jiangsu Center for Collaborative Innovation in Geographical Information Resource Development and Application, Nanjing 210023,China
5. School of Engineering and Applied Sciences,National University of Mongolia, Ulaanbaatar 210646, Mongolia
6. School of Art & Sciences, National University of Mongolia, Ulaanbaatar 210646, Mongolia

Received:2021-01-20 Revised:2021-03-20 Online:2022-05-25 Published:2022-07-25
Supported by:
National Natural Science Foundation of China(41971385);Strategic Priority Research Program (Class A) of the Chinese Academy of Sciences(XDA2003020302);Construction Project of the China Knowledge Center for Engineering Sciences and Technology(CKCEST-2021-2-18)

摘要/Abstract

摘要：

蒙古高原地处干旱半干旱地区,河流水系对该区域的资源环境格局及其生态环境影响重大。发源于蒙古国的色楞格河是蒙古高原最主要的水资源来源,准确掌握该流域的水体信息对东北亚地区生态环境问题及资源保护具有重要意义。本文以蒙古高原色楞格河流域为研究对象,基于谷歌地球引擎（Google Earth Engine,GEE）云平台,使用 Sentinel-2 多光谱卫星遥感影像,利用深度神经网络（Deep Neural Network, DNN）方法对色楞格河流域的水体信息进行提取,并与支持向量机方法进行对比;利用全球30 m SRTM数据生成水系分布矢量图,通过空间分析形成河流提取目标区,结合深度神经网络分类结果,绘制蒙古国色楞格河流域2019年河流分布图。研究结果表明：① 该方法能够准确地完成大流域范围内的水体制图,提取结果能够体现色楞格河流域河流的空间分布,且能够减少河流断流、空洞现象;② 深度神经网络模型中批量大小设置为8时,在处理数据速度与精度中达到最优,而神经网络结构中隐含层数达到4层时,在精度评价指标测试数据集上达到0.9666,保证了模型特征挖掘能力;③ 经样本点的验证,结果总体精度达到97.65%,可以满足实际应用需求。本研究预期可以为蒙古高原的水体提取提供方法支持和相关数据支持。

关键词: 高原水体提取, 深度神经网络, 遥感解译, 色楞格河, 蒙古高原, 中蒙俄经济走廊, Google Earth Engine, 水体指数

Abstract:

Arid and semi-arid climate zones of the Mongolian Plateau support multiple river systems that significantly impact the distribution of resources and ecological environments of the region. The Selenga River is the most important water resource on the Mongolian Plateau and the only river that enters the largest freshwater lake in the world——Lake Baikal in Russia. Thus, an accurate understanding of the hydrologic characteristics of the Selenga River Basin is critical in monitoring the ecology, environment, and water resources in Northeast Asia. Obtaining the plane morphology of curved rivers in the upper reaches of the plateau region is helpful to analyze the mathematical and physical mechanism of the evolution of curved rivers, which is of great significance to the study of river geomorphology. In this study, the Selenga River Basin in the Mongolia was taken as the study area. Based on the Google Earth Engine (GEE) cloud platform, Sentinel-2 multi-spectral satellite remote sensing images, Normalized Difference Water Index (NDWI), Modified Normalized Difference Water Index (MNDWI), and automated Water Extraction Index (AWEI) were used to construct a deep neural network model for the extraction of fine water body data in the Selenga River Basin. Three different types of water bodies, including lakes, main streams, and tributaries, were selected to compare the extraction results of deep neural network with the extraction results of support vector machines and Global Surface Water data. Global 30 m SRTM data were then used to generate a river system vector distribution map, identify river extraction target areas, and draw a 10 m spatial resolution river distribution map for the 2019 Selenga River Basin using output from the deep neural network model. Results indicate that, firstly, this method can accurately produce a water regime map for a large basin. The extraction results can reflect the spatial distribution of rivers in the Selenga River Basin in Mongolia, as well as reduce the drying-up and empty phenomenon of rivers. Secondly, the batch size of eight in the deep neural network model optimize data processing speed and accuracy while the four hidden layers in the neural network structure produce an accuracy evaluation index of 0.9666, which guarantees feature mining capabilities of models. Thirdly, after the verification of sample points, the overall accuracy reaches 97.65%, meeting actual application requirements. Thus, this research provides a method and data support for water body extraction for the Mongolian Plateau, which will play an important role in revealing long-term changes in river hydrological conditions, and impact related research on resources and environmental benefits.

Key words: water extraction of plateau, deep neural network, remote sensing interpretation, selenga river, mongolian plateau, China-Mongolia-Russia Economic Corridor, Google Earth Engine, water index

姚锦一, 王卷乐, 严欣荣, 魏海硕, Altansukh Ochir, Davaadorj Davaasuren. 基于深度神经网络的蒙古国色楞格河流域水体信息提取[J]. 地球信息科学学报, 2022, 24(5): 1009-1017.DOI:10.12082/dqxxkx.2022.210031

YAO Jinyi, WANG Juanle, YAN Xinrong, WEI Haishuo, Altansukh Ochir, Davaadorj Davaasuren. Water Information Extraction of Selenga River Basin in Mongolia based on Deep Neural Network[J]. Journal of Geo-information Science, 2022, 24(5): 1009-1017.DOI:10.12082/dqxxkx.2022.210031

图/表 12

图1

表1

水体指数计算公式

水体指数	计算方法	编号
NDWI	NDWI =( $ρ green - ρ NIR$ )/( $ρ green + ρ NIR$ )	（1）
MNDWI	MNDWI =( $ρ green - ρ MIR$ )/( $ρ green + ρ MIR$ )	（2）
AWEI	AWEInsh =4×( $ρ green - ρ SWIR 1$ ) $-$ (0.25× $ρ NIR$ +2.75× $ρ SWIR 2$ )	（3）
AWEI	AWEIsh= $ρ blue$ +2.5× $ρ green -$ 1.5×( $ρ NIR + ρ SWIR 1$ )-0.25× $ρ SWIR 2$	（4）

表1

图2

图3

表2

表3

图4

表4

图5

表5

图6

图7

参考文献 24

[1]	朱长明, 骆剑承, 沈占锋, 等. DEM辅助下的河道细小线性水体自适应迭代提取[J]. 测绘学报, 2013, 42(2):277-283.
	[ Zhu C M, Luo J C, Shen Z F, et al. River linear water adaptive auto-extraction on remote sensing image aided by DEM[J]. Acta Geodaetica et Cartographica Sinica, 2013, 42(2):277-283. ] DOI: 1001-1595(2013)02-0277-07 doi: 1001-1595(2013)02-0277-07
[2]	孙娜, 高志强, 王晓晶, 等. 基于高分遥感影像的黄土高原地区水体高精度提取[J]. 国土资源遥感, 2017, 29(4):173-178.
	[ Sun N, Gao Z Q, Wang X J, et al. High-precise extraction for water on the Loess Plateau region from high resolution satellite image[J]. Remote Sensing for Land & Resources, 2017, 29(4):173-178. ] DOI: 10.6046/gtzyyg.2017.04.26 doi: 10.6046/gtzyyg.2017.04.26
[3]	肖茜, 杨昆, 洪亮. 基于多源数据的云贵高原湖泊水体提取方法对比研究[J]. 云南师范大学学报(自然科学版), 2018, 38(2):73-78.
	[ Xiao Q, Yang K, Hong L. Comparative study of extraction methods of lake water in Yunnan-Guizhou plateau based on multi-source images[J]. Journal of Yunnan Normal University (Natural Sciences Edition), 2018, 38(2):73-78. ] DOI: 10.7699/j.ynnu.ns-2018-028 doi: 10.7699/j.ynnu.ns-2018-028
[4]	张帅旗, 周秉荣, 史飞飞, 等. 基于高分一号卫星遥感数据的青藏高原高寒湿地信息提取方法研究——以玛多县为例[J]. 高原气象, 2020, 39(6):1309-1317. doi: 10.7522/j.issn.1000-0534.2019.00131
	[ Zhang S Q, Zhou B R, Shi F F, et al. Study on information extraction method of alpine wetland in Qinghai-Xizang plateau based on remote sensing data of GF-1 satellite—taking Maduo County for example[J]. Plateau Meteorology, 2020, 39(6):1309-1317. ] DOI: 10.7522/j.issn.1000-0534.2019.00131 doi: 10.7522/j.issn.1000-0534.2019.00131
[5]	Hinton G E, Salakhutdinov R R. Reducing the dimensionality of data with neural networks[J]. Science (New York, N Y), 2006, 313(5786):504-507. DOI: 10.1126/science.1127647 doi: 10.1126/science.1127647
[6]	周岩, 董金玮. 陆表水体遥感监测研究进展[J]. 地球信息科学学报, 2019, 21(11):1768-1778. doi: 10.12082/dqxxkx.2019.190518
	[ Zhou Y, Dong J W. Review on monitoring open surface water body using remote sensing[J]. Journal of Geo-information Science, 2019, 21(11):1768-1778. ] DOI: 10.12082/dqxxkx.2019.190518 doi: 10.12082/dqxxkx.2019.190518
[7]	陈前, 郑利娟, 李小娟, 等. 基于深度学习的高分遥感影像水体提取模型研究[J]. 地理与地理信息科学, 2019, 35(4):43-49.
	[ Chen Q, Zheng L J, Li X J, et al. Water body extraction from high-resolution satellite remote sensing images based on deep learning[J]. Geography and Geo-Information Science, 2019, 35(4):43-49. ] DOI: 10.3969/j.issn.1672-0504.2019.04.007 doi: 10.3969/j.issn.1672-0504.2019.04.007
[8]	何海清, 杜敬, 陈婷, 等. 结合水体指数与卷积神经网络的遥感水体提取[J]. 遥感信息, 2017, 32(5):82-86.
	[ He H Q, Du J, Chen T, et al. Remote sensing image water body extraction combing NDWI with convolutional neural network[J]. Remote Sensing Information, 2017, 32(5):82-86. ] DOI: 10.3969/j.issn.1000-3177.2017.05.013 doi: 10.3969/j.issn.1000-3177.2017.05.013
[9]	Miao Z M, Fu K, Sun H, et al. Automatic water-body segmentation from high-resolution satellite images via deep networks[J]. IEEE Geoscience and Remote Sensing Letters, 2018, 15(4):602-606. DOI: 10.1109/LGRS.2018.2794545 doi: 10.1109/LGRS.2018.2794545
[10]	Chen Y, Fan R S, Yang X C, et al. Extraction of urban water bodies from high-resolution remote-sensing imagery using deep learning[J]. Water, 2018, 10(5):585. DOI: 10.3390/w10050585 doi: 10.3390/w10050585
[11]	Guo X Y, Chen D, Parker G. Flow directionality of pristine meandering rivers is embedded in the skewing of high-amplitude bends and neck cutoffs[J]. PNAS, 2019, 116(47):23448-23454. doi: 10.1073/pnas.1910874116
[12]	郝斌飞, 韩旭军, 马明国, 等. Google Earth Engine在地球科学与环境科学中的应用研究进展[J]. 遥感技术与应用, 2018, 33(4):600-611.
	[ Hao B F, Han X J, Ma M G, et al. Research progress on the application of google earth engine in geoscience and environmental sciences[J]. Remote Sensing Technology and Application, 2018, 33(4):600-611. ] DOI: 10.11873/j.issn.1004-0323.2018.4.0600 doi: 10.11873/j.issn.1004-0323.2018.4.0600
[13]	Törnqvist R, Jarsjö J, Pietroń J, et al. Evolution of the hydro-climate system in the Lake Baikal basin[J]. Journal of Hydrology, 2014, 519:1953-1962. DOI: 10.1016/j.jhydrol.2014.09.074 doi: 10.1016/j.jhydrol.2014.09.074
[14]	贾文翰, 白建荣, 杨利荣. 河流湖泊湿地边界定位中遥感影像时相选择方法[J]. 测绘与空间地理信息, 2018, 41(1):46-48,54.
	[ Jia W H, Bai J R, Yang L R. Phase selection method of remote sensing images for river and lake wetland boundary location[J]. Geomatics & Spatial Information Technology, 2018, 41(1):46-48,54. ] DOI: 10.3969/j.issn.1672-5867.2018.01.013 doi: 10.3969/j.issn.1672-5867.2018.01.013
[15]	Du Y, Zhang Y H, Ling F, et al. Water bodies' mapping from sentinel-2 imagery with modified normalized difference water index at 10-m spatial resolution produced by sharpening the SWIR band[J]. Remote Sensing, 2016, 8(4):354. DOI: 10.3390/rs8040354 doi: 10.3390/rs8040354
[16]	Yang X C, Zhao S S, Qin X B, et al. Mapping of urban surface water bodies from sentinel-2 MSI imagery at 10 m resolution via NDWI-based image sharpening[J]. Remote Sensing, 2017, 9(6):596. DOI: 10.3390/rs9060596 doi: 10.3390/rs9060596
[17]	McFeeters S K. The use of the Normalized Difference Water Index (NDWI) in the delineation of open water features[J]. International Journal of Remote Sensing, 1996, 17(7):1425-1432. DOI: 10.1080/01431169608948714 doi: 10.1080/01431169608948714
[18]	徐涵秋. 利用改进的归一化差异水体指数(MNDWI)提取水体信息的研究[J]. 遥感学报, 2005, 9(5):589-595.
	[ Xu H Q. A study on information extraction of water body with the modified normalized difference water index (MNDWI)[J]. Journal of Remote Sensing, 2005, 9(5):589-595. ] DOI: 1007-4619(2005)05-0589-07 doi: 1007-4619(2005)05-0589-07
[19]	Feyisa G L, Meilby H, Fensholt R, et al. Automated Water Extraction Index: A new technique for surface water mapping using Landsat imagery[J]. Remote Sensing of Environment, 2014, 140:23-35. DOI: 10.1016/j.rse.2013.08.029 doi: 10.1016/j.rse.2013.08.029
[20]	冈萨雷斯, 伍兹. 数字图像处理(第三版)[M]. 北京: 电子工业出版社, 2017.
	[ Rafael C. G, Richard E. W. Digital image processing (Third Edition)[M]. Beijing: Publishing House of Electronics Industry, 2017. ]
[21]	Pekel J F, Cottam A, Gorelick N, et al. High-resolution mapping of global surface water and its long-term changes[J]. Nature, 2016, 540(7633):418-422. DOI: 10.1038/nature20584 doi: 10.1038/nature20584
[22]	O'Callaghan J F, Mark D M. The extraction of drainage networks from digital elevation data[J]. Computer Vision, Graphics, and Image Processing, 1984, 28(3):323-344. DOI: 10.1016/S0734-189X(84)80011-0 doi: 10.1016/S0734-189X(84)80011-0
[23]	Priestnall G, Aplin P. Cover: Spatial and temporal remote sensing requirements for river monitoring[J]. International Journal of Remote Sensing, 2006, 27(11):2111-2120. DOI: 10.1080/01431160500396139 doi: 10.1080/01431160500396139
[24]	田富有, 吴炳方, 曾红伟, 等. 基于多层神经网络与Sentinel-2数据的大豆种植区识别方法[J]. 地球信息科学学报, 2019, 21(6):918-927. doi: 10.12082/dqxxkx.2019.180424
	[ Tian F Y, Wu B F, Zeng H W, et al. Identifying soybean cropped area with sentinel-2 data and multi-layer neural network[J]. Journal of Geo-Information Science, 2019, 21(6):918-927. ] DOI: 10.12082/dqxxkx.2019.180424 doi: 10.12082/dqxxkx.2019.180424

批量大小	训练集损失值	训练集准确率	测试集损失值	测试集准确率
4	0.1502	0.9508	0.1198	0.9632
8	0.1713	0.9489	0.1205	0.9666
16	0.1981	0.9407	0.1318	0.9615
32	0.2130	0.9381	0.1296	0.9582
64	0.2719	0.9047	0.2225	0.9415

层数	训练集损失值		训练集损失值	测试集准确率
1	0.2418	0.9198	0.2188	0.9197
2	0.1958	0.9381	0.1382	0.9565
3	0.2088	0.9438	0.1371	0.9615
4	0.1713	0.9489	0.1205	0.9666
5	0.2168	0.9426	0.1341	0.9632

水体类型	遥感影像	DNN结果	SVM结果	JRC数据集
湖泊
干流
支流

方法	总体精度/%	生产者精度/%	用户精度/%	Kappa系数
DNN	97.65	91.57	86.36	0.88
SVM	93.48	80.95	96.23	0.84

基于深度神经网络的蒙古国色楞格河流域水体信息提取

Water Information Extraction of Selenga River Basin in Mongolia based on Deep Neural Network

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 12

参考文献 24

相关文章 15

Metrics

本文评价

推荐阅读 0

[1]	张文璇, 王卷乐. 基于SAR影像后向散射特性的中俄黑龙江流域洪水监测[J]. 地球信息科学学报, 2022, 24(4): 802-813.
[2]	林禹, 赵泉华, 李玉. 一种基于深度传递迁移学习的遥感影像分类方法[J]. 地球信息科学学报, 2022, 24(3): 495-507.
[3]	熊皓丽, 周小成, 汪小钦, 崔雅君. 基于GEE云平台的福建省10 m分辨率茶园专题空间分布制图[J]. 地球信息科学学报, 2021, 23(7): 1325-1337.
[4]	李文萍, 王伟, 高星, 伍宇明, 王学成, 刘青. 融合面向对象和分水岭算法的山地湖泊提取方法[J]. 地球信息科学学报, 2021, 23(7): 1272-1285.
[5]	彭妍菲, 李忠勤, 姚晓军, 牟建新, 韩伟孝, 王盼盼. 基于多源遥感数据和GEE平台的博斯腾湖面积变化及影响因素分析[J]. 地球信息科学学报, 2021, 23(6): 1131-1153.
[6]	姜伊兰, 陈保旺, 黄玉芳, 崔佳琪, 郭宇龙. 基于Google Earth Engine和NDVI时序差异指数的作物种植区提取[J]. 地球信息科学学报, 2021, 23(5): 938-947.
[7]	李培林, 刘小平, 黄应淮, 张鸿辉. 基于GEE平台的广州市主城区不透水面时间序列提取[J]. 地球信息科学学报, 2020, 22(3): 638-648.
[8]	陈炜, 黄慧萍, 田亦陈, 杜云艳. 基于Google Earth Engine平台的三江源地区生态环境质量动态监测与分析[J]. 地球信息科学学报, 2019, 21(9): 1382-1391.
[9]	刘凯, 彭力恒, 李想, 谭敏, 王树功. 基于Google Earth Engine的红树林年际变化监测研究[J]. 地球信息科学学报, 2019, 21(5): 731-739.
[10]	王卷乐, 程凯, 祝俊祥, 刘清. 蒙古国30米分辨率土地覆盖产品研制与空间格局分析[J]. 地球信息科学学报, 2018, 20(9): 1263-1273.
[11]	徐晗泽宇, 刘冲, 王军邦, 齐述华. Google Earth Engine平台支持下的赣南柑橘果园遥感提取研究[J]. 地球信息科学学报, 2018, 20(3): 396-404.
[12]	帅晨, 沙晋明, 林金煌, 季建万, 周正龙, 高尚. 不同下垫面遥感指数与地温关系的空间差异性研究[J]. 地球信息科学学报, 2018, 20(11): 1657-1666.
[13]	王九中, 田海峰, 邬明权, 王力, 王长耀. 河南省冬小麦快速遥感制图[J]. 地球信息科学学报, 2017, 19(6): 846-853.
[14]	饶萍, 王建力. 最优分区与最优指数联合的水体信息提取[J]. 地球信息科学学报, 2017, 19(5): 702-712.
[15]	王铭, 宋开山, 邵田田, 吕冬梅, 杜嘉, 臧淑英. 南美洲的地表水体制图研究[J]. 地球信息科学学报, 2015, 17(10): 1260-1268.