基于GEE多源遥感数据的干旱区植被地物类型提取

doi:10.13866/j.azr.2024.01.15

干旱区研究 ›› 2024, Vol. 41 ›› Issue (1): 157-168.doi: 10.13866/j.azr.2024.01.15

基于GEE多源遥感数据的干旱区植被地物类型提取

姚金玺^1,²(),肖成志³,张志³(),王浪⁴,张焜²

1.中交第二公路勘察设计研究院有限公司，湖北武汉 430056
2.青海省青藏高原北部地质过程与矿产资源重点实验室，青海西宁 810300
3.中国地质大学（武汉）地球物理与空间信息学院，湖北武汉 430074
4.中国地质调查局乌鲁木齐自然资源综合调查中心，新疆乌鲁木齐 830057

收稿日期:2023-02-14 修回日期:2023-08-17 出版日期:2024-01-15 发布日期:2024-01-24
通讯作者: 张志. E-mail: 3slab@cug.edu.cn
作者简介:姚金玺（1997-），男，硕士研究生，主要从事遥感信息提取研究. E-mail: 20656yjx@cug.edu.cn
基金资助:
青海省青藏高原北部地质过程与矿产资源重点实验室开放课题(2019-kz-01);青海省科技厅创新平台建设专项项目(2019-ZJ-T04);湖北省重点研发计划项目(2021BAA185);武汉市重点研发计划项目(2022012202015071)

Vegetation feature type extraction in arid regions based on GEE multi-source remote sensing data

YAO Jinxi^1,²(),XIAO Chengzhi³,ZHANG Zhi³(),WANG Lang⁴,ZHANG Kun²

1. CCCC Second Highway Consultants Co., Ltd., Wuhan 430056, Hubei, China
2. Key Laboratory of The Northern Qinghai-Tibet Plateau Geological Processes and Mineral Resources, Xining 810300, Qinghai, China
3. School of Geophysics and Geomatics, China University of Geosciences, Wuhan 430074, Hubei, China
4. Urumqi Comprehensive Survey Center on Natural Resources, China Geological Survey, Urumqi 830057, Xinjiang, China

Received:2023-02-14 Revised:2023-08-17 Online:2024-01-15 Published:2024-01-24

摘要/Abstract

摘要：

诺木洪地区是青海省重要的枸杞种植基地，针对主要植被类型进行准确和快速提取对于种植业的可持续发展具有关键意义。然而，诺木洪地区所属的干旱区具有稀疏的植被覆盖和土壤背景影响显著的特点，仅使用少数遥感源或部分特征无法满足干旱区植被提取的要求。因此，整合多种遥感数据源，挖掘植被分类显著特征，并尝试不同的分类优化方法，在提高干旱区植被分类精度和可靠性方面具有重要意义。本研究基于谷歌地球引擎（Google Earth Engine，GEE）平台，使用Sentinel-1合成孔径雷达（Synthetic Aperture Radar，SAR）数据、Sentinel-2光学数据，探讨了红边光谱、纹理以及雷达特征对干旱区植被类型提取的重要性，验证了利用基尼指数（Gini Index，Gini）寻找最优特征组合的可行性，结合支持向量机算法对2021年青海诺木洪地区地物类型进行提取，并对最终的分类结果优化处理。研究表明：（1）Sentinel-2红边指数、纹理信息和Sentinel-1雷达波段有利于植被信息提取，分类总体精度和Kappa系数分别达到了95.51%和0.9406；（2）根据Gini指数得到特征重要性，将分类特征由29个压缩至17个，且表明雷达极化特征、光谱特征和纹理特征对于分类的重要性依此递减；（3）使用简单非迭代聚类算法以及邻域滤波投票决策融合方法，不仅最优总体精度和Kappa系数达到了96.06%和0.9479，且针对孤点类型的噪声也有较好的抑制效果。本研究利用GEE遥感云平台和多源遥感数据以及机器学习算法，能够准确、快速、高效地提取大尺度范围干旱区地物信息，具有较大的应用潜力。

关键词: 地表覆盖, 特征选择, 支持向量机, 分类优化, Google Earth Engine

Abstract:

Nuomuhong region is an important Wolfberry cultivation base in Qinghai Province, China. Accurate and rapid extraction of the primary vegetation types is of critical significance for the sustainable development of agriculture in this region. However, the arid nature of the Nuomuhong area, characterized by sparse vegetation cover and significant soil background effects, presents challenges for vegetation extraction using only a limited number of remote sensing sources or partial features. Therefore, integrating multiple remote sensing data sources, exploring significant features for vegetation classification, and experimenting with different classification and optimization methods are paramount for enhancing the accuracy and reliability of vegetation classification in arid regions. Based on the Google Earth Engine (GEE) platform, this study used Sentinel-1 Synthetic Aperture Radar and Sentinel-2 optical data to explore the importance of red edge, texture, and radar features in extracting vegetation types in arid regions. Additionally, it verifies the feasibility of using the GINI index (GINI) to determine the optimal feature combination. The main geospatial types in Nomu Hong, Qinghai, China, in 2021 were extracted by combining them with the support vector machine algorithm. The classification results were processed using decision fusion methods. The results showed that: (1) Sentinel-2 red edge index, texture data, and Sentinel-1 radar band were beneficial for the extraction of vegetation-related information, with an overall classification accuracy and Kappa coefficient of 95.51% and 0.9406, respectively. (2) Based on the importance obtained by the GINI index, the features involved in the classification were reduced from 29 to 17, and the significance was radar polarization features > spectral features > texture features. (3) Using a simple noniterative clustering algorithm and neighborhood filtering voting decision fusion method not only achieved the optimal overall accuracy and Kappa coefficient but also had an excellent suppression effect on isolated noise. Using the GEE remote sensing cloud platform, multisource remote sensing data, and machine learning algorithms, this study can accurately, quickly, and efficiently extract large-scale arid region geospatial information, which can have great application potential.

Key words: surface coverage, feature selection, support vector machine, classification optimization, Google Earth Engine

姚金玺, 肖成志, 张志, 王浪, 张焜. 基于GEE多源遥感数据的干旱区植被地物类型提取[J]. 干旱区研究, 2024, 41(1): 157-168.

YAO Jinxi, XIAO Chengzhi, ZHANG Zhi, WANG Lang, ZHANG Kun. Vegetation feature type extraction in arid regions based on GEE multi-source remote sensing data[J]. Arid Zone Research, 2024, 41(1): 157-168.

图/表 11

图1

图2

图3

表1

图4

图5

图6

图7

表2

表3

图8

参考文献 31

[1]	Zhang P, Hu S, Li W, et al. Improving parcel-level mapping of smallholder crops from vhsr imagery: An ensemble machine-learning-based framework[J]. Remote Sensing, 2021, 13(11): 2146. doi: 10.3390/rs13112146
[2]	Denize J, Hubert-Moy L, Betbeder J, et al. Evaluation of using sentinel-1 and-2 time-series to identify winter land use in agricultural landscapes[J]. Remote Sensing, 2018, 11(1): 37. doi: 10.3390/rs11010037
[3]	Xiao X, Boles S, Frolking S, et al. Mapping paddy rice agriculture in South and Southeast Asia using multi-temporal MODIS images[J]. Remote Sensing of Environment, 2006, 100(1): 95-113. doi: 10.1016/j.rse.2005.10.004
[4]	王文静, 张霞, 赵银娣, 等. 综合多特征的Landsat 8时序遥感图像棉花分类方法[J]. 遥感学报, 2017, 21(1): 115-124.
	[Wang Wengjing, Zhang Xia, Zhao Yindi, et al. Cotton extraction method of integrated multi-features based on multi-temporal Landsat 8 images[J]. Journal of Remote Sensing, 2017, 21(1): 115-124.]
[5]	Zhang H, Du H, Zhang C, et al. An automated early-season method to map winter wheat using time-series Sentinel-2 data: A case study of Shandong, China[J]. Computers and Electronics in Agriculture, 2021, 182: 105962. doi: 10.1016/j.compag.2020.105962
[6]	王德军, 姜琦刚, 李远华, 等. 基于Sentinel-2A/B时序数据与随机森林算法的农耕区土地利用分类[J]. 国土资源遥感, 2020, 32(4): 236-243.
	[Wang Dejun, Jiang Qigang, Li Yuanhua, et al. Land use classification of farming areas based on time series Sentinel-2A/B data and random forest algorithm[J]. Remote Sensing for Natural Resources, 2020, 32(4): 236-243.]
[7]	闵钰魁, 柯樱海, 韩月, 等. 融合Sentinel-2和GF-1时序影像的入侵植物互花米草清除动态监测[J]. 遥感学报, 2023, 27(6): 1467-1479.
	[Min Yukui, Ke Yinhai, Han Yue, et al. Dynamic monitoring of invasive Spartina alterniflora clearance via fusion of Sentinel-2 and GF-1 time series images[J]. National Remote Sensing Bulletin, 2023, 27(6): 1467-1479.]
[8]	Fatchurrachman, Rudiyanto, Soh N C, et al. High-resolution mapping of paddy rice extent and growth stages across peninsular malaysia using a fusion of sentinel-1 and 2 time series data in Google Earth Engine[J]. Remote Sensing, 2022, 14(8): 1875. doi: 10.3390/rs14081875
[9]	Chakhar A, Hernández-López D, Ballesteros R, et al. Improving the accuracy of multiple algorithms for crop classification by integrating sentinel-1 observations with sentinel-2 data[J]. Remote Sensing, 2021, 13(2): 243. doi: 10.3390/rs13020243
[10]	明義森, 刘启航, 柏荷, 等. 利用光学和SAR遥感数据的若尔盖湿地植被分类与变化监测[J]. 遥感学报, 2023, 27(6): 1414-1425.
	[Ming Yiseng, Liu Qihang, Bai He, et al. Classification and change detection of vegetation in the Ruoergai Wetland using optical and SAR remote sensing data[J]. National Remote Sensing Bulletin, 2023, 27(6): 1414-1425.]
[11]	张琍, 罗文庭, 张皓寰, 等. 时序Sentinel-1和Sentinel-2 数据支持下的鄱阳湖湿地草本植物群落制图分类[J]. 遥感学报, 2023, 27(6): 1362-1375.
	[Zhang Li, Luo Wenting, Zhang Haohuan, et al. Classification scheme for mapping wetland herbaceous plant communities using time series Sentinel-1 and Sentinel-2 data[J]. National Remote Sensing Bulletin, 2023, 27(6): 1362-1375.]
[12]	曹文梅, 刘廷玺, 王喜喜, 等. 科尔沁沙丘草甸相间地区土地利用与覆被识别[J]. 干旱区研究, 2021, 38(2): 526-535.
	[Cao Wenmei, Liu Tingxi, Wang Xixi, et al. Land use and land cover classifications of Horqin Sandy Land dune-meadow areas[J]. Arid Zone Research, 2021, 38(2): 526-535.]
[13]	姚金玺, 王浪, 李建忠, 等. 青海诺木洪地区多源遥感及多特征组合地物分类[J]. 农业工程学报, 2022, 38(3): 247-256.
	[Yao Jinxi, Wang Lang, Li Jianzhong, et al. Multi-source remote sensing and multi-feature combination ground object classification in Nuomuhong areas, Qinghai Province of China[J]. Transactions of the Chinese Society of Agricultural Engineering, 2022, 38(3): 247-256.]
[14]	Awad M. Google Earth Engine (GEE) cloud computing based crop classification using radar, optical images and Support Vector Machine Algorithm (SVM)[C]// IEEE 3rd International Multidisciplinary Conference on Engineering Technology, 2021: 71-76.
[15]	Virnodkar S, Pachghare V K, Patil V C, et al. Performance evaluation of RF and SVM for sugarcane classification using Sentinel-2 NDVI time-series[C]// Progress in Advanced Computing and Intelligent Engineering: Proceedings of ICACIE, 2021: 163-174.
[16]	Ponganan N, Horanont T, Artlert K, et al. Land Cover Classification using Google Earth Engine’s Object-oriented and Machine Learning Classifier[C]// 2021 2nd International Conference on Big Data Analytics and Practices (IBDAP), IEEE, 2021: 33-37.
[17]	Yao J, Wu J, Xiao C, et al. The classification method study of crops remote sensing with deep learning, machine learning, and Google Earth Engine[J]. Remote Sensing, 2022, 14(12): 2758. doi: 10.3390/rs14122758
[18]	梁顺林, 白瑞, 陈晓娜, 等. 2019年中国陆表定量遥感发展综述[J]. 遥感学报, 2020, 24 (6): 618-671.
	[Liang Shunlin, Bai Rui, Chen Xiaona, et al. Review of China’s land surface quantitative remote sensing development in 2019[J]. Journal of Remote Sensing, 2020, 24(6): 618-671.]
[19]	邱凤婷, 过志峰, 张宗科, 等. 基于改进SVM分类法的SAR图像水体面积提取研究[J]. 地球信息科学学报, 2022, 24(5): 940-948. doi: 10.12082/dqxxkx.2022.210095
	[Qiu Fengting, Guo Zhifeng, Zhang Zongke, et al. Water body area extraction from SAR image based on improved SVM classification method[J]. Journal of Geo-information Science, 2022, 24(5): 940-948.] doi: 10.12082/dqxxkx.2022.210095
[20]	庞海威, 余殿, 任成宝, 等. 宁夏东部半干旱区典型植物群落遥感分类特征[J]. 干旱区研究, 2022, 39(6): 1930-1941.
	[Pang Haiwei, Yu Dian, Ren Chengbao, et al. Remote sensing classification characteristics of typical plant communities in the semi-arid areas of eastern Ningxia[J]. Arid Zone Research, 2022, 39(6): 1930-1941.]
[21]	杨维涛, 孙建国, 马恒利, 等. 地貌形态多尺度综合分类方法[J]. 干旱区研究, 2022, 39(2): 638-645.
	[Yang Weitao, Sun Jianguo, Ma Hengli, et al. A multi-scale integrated classification method for landforms[J]. Arid Zone Research, 2022, 39(2): 638-645.]
[22]	Tassi A, Vizzari M. Object-oriented lulc classification in google earth engine combining snic, glcm, and machine learning algorithms[J]. Remote Sensing, 2020, 12(22): 3776. doi: 10.3390/rs12223776
[23]	刘通, 任鸿瑞. GEE平台下利用物候特征进行面向对象的水稻种植分布提取[J]. 农业工程学报, 2022, 38(12): 189-196.
	[Liu Tong, Ren Hongrui. Object-oriented extraction of paddy rice planting areas using phenological features from the GEE platform[J]. Transactions of the Chinese Society of Agricultural Engineering, 2022, 38(12): 189-196.]
[24]	肖森天, 依力亚斯江·努尔麦麦提, 努尔比耶·穆合塔尔, 等. 基于光学和雷达多源遥感的于田绿洲土壤盐渍化时空分析[J]. 干旱区研究, 2023, 40(1): 59-68.
	[Xiao Sentian, Ilyas Nurmemet, Nuerbiye Muhetaer, et al. Spatial and temporal analysis of soil salinity in Yutian Oasis by combined optical and radar multi-source remote sensing[J]. Arid Zone Research, 2023, 40(1): 59-68.]
[25]	Yang G, Zhao Y, Xing H, et al. Understanding the changes in spatial fairness of urban greenery using time-series remote sensing images: A case study of Guangdong-Hong Kong-Macao Greater Bay[J]. Science of The Total Environment, 2020, 715: 136763. doi: 10.1016/j.scitotenv.2020.136763
[26]	Achanta R, Susstrunk S. Superpixels and polygons using simple non-iterative clustering[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 4651-4660.
[27]	穆学青, 郭向阳, 明庆忠, 等. 黄河流域旅游生态安全的动态演变特征及驱动因素[J]. 地理学报, 2022, 77(3): 714-735. doi: 10.11821/dlxb202203015
	[Mu Xueqing, Guo Xiangyang, Ming Qingzhong, et al. Dynamic evolution characteristics and driving factors of tourism ecological security in the Yellow River Basin[J]. Acta Geographica Sinica, 2022, 77(3): 714-735.] doi: 10.11821/dlxb202203015
[28]	王振武, 孙佳骏, 于忠义, 等. 基于支持向量机的遥感图像分类研究综述[J]. 计算机科学, 2016, 43(9): 11-17. doi: 10.11896/j.issn.1002-137X.2016.09.002
	[Wang Zhenwu, Sun Jiajun, Yu Zhongyi, et al. Review of remote sensing image classification based on support vector machine[J]. Computer Science, 2016, 43(9): 11-17.] doi: 10.11896/j.issn.1002-137X.2016.09.002
[29]	戴声佩, 易小平, 罗红霞, 等. 基于GEE和Landsat时间序列数据的海南岛土地利用分类研究[J]. 热带作物学报, 2021, 42(11): 3351-3357.
	[Dai Shengpei, Yi Xiaoping, Luo Hongxia, et al. Mapping land use in Hainan Island based on Google Earth Engine and Landsat time series data[J]. Chinese Journal of Tropical Crops, 2021, 42(11): 3351-3357.]
[30]	宁晓刚, 常文涛, 王浩, 等. 联合GEE与多源遥感数据的黑龙江流域沼泽湿地信息提取[J]. 遥感学报, 2022, 26(2): 386-396. doi: 10.11834/jrs.20200033
	[Ning Xiaogang, Chang Wentao, Wang Hao, et al. Extraction of marsh wetland in Heilongjiang Basin based on GEE and multi-source remote sensing data[J]. National Remote Sensing Bulletin, 2022, 26(2): 386-396.] doi: 10.11834/jrs.20200033
[31]	杨泽航, 王文, 鲍健雄. 融合多源遥感数据的黑河中游地区生长季早期作物识别[J]. 地球信息科学学报, 2022, 24(5): 996-1008. doi: 10.12082/dqxxkx.2022.210527
	[Yang Zehang, Wang Wen, Bao Jianxiong. Identifying crop types in early growing season in the middle reaches of Heihe River by fusing multi-source remote sensing data[J]. Journal of Geo-information Science, 2022, 24(5): 996-1008.] doi: 10.12082/dqxxkx.2022.210527

特征变量	简写	特征说明
光谱特征	B	B2~B8、B8A、B9、B11、B12
	NDVI	（B8A-B4）/（B8A+B4）
	EVI	2.5×（B8A-B4）/（B8A+6×B4-7.5×B2+1）
	GNDVI	（B8A-B3）/（B8A+B3）
	NDVIre1	（B8A-B5）/（B8A+B5）
	NDVIre2	（B8A-B6）/（B8A+B6）
	NDVIre3	（B8A-B7）/（B8A+B7）
	NDre1	（B6-B5）/（B6+B5）
	NDre2	（B7-B5）/（B7+B5）
	CIre	B7/B5-1
纹理特征	Gray_asm	角二阶矩
	Gray_contrast	对比度
	Gray_corr	相关性
	Gray_ent	熵
	Gray_var	方差
	Gray_idm	逆差矩
	Gray_savg	总和平均值
雷达特征	VV	VV极化后向散射系数
雷达特征	VH	VH极化后向散射系数

类别		草地	防护林	建筑物	裸地	梭梭林	枸杞种植园地	总体精度/%	Kappa系数
方案1	UA/%	90.41	93.12	94.71	87.23	83.84	90.75	91.42	0.8865
方案1	PA/%	74.16	85.49	97.39	85.60	93.96	89.53	91.42	0.8865
方案2	UA/%	75.79	97.05	93.72	85.60	86.96	96.12	93.21	0.9105
方案2	PA/%	80.90	94.18	95.86	87.20	95.76	90.47	93.21	0.9105
方案3	UA/%	93.67	97.86	96.34	88.60	90.48	97.04	95.51	0.9406
方案3	PA/%	83.15	94.87	97.44	91.20	96.08	95.14	95.51	0.9406

类别		草地	防护林	建筑物	裸地	梭梭林	枸杞种植园地	总体精度/%	Kappa系数
优化1	UA/%	96.00	96.58	96.27	87.47	90.42	97.17	95.23	0.9370
优化1	PA/%	80.90	94.67	97.01	91.20	95.43	95.20	95.23	0.9370
优化2	UA/%	95.06	97.28	96.46	89.03	91.21	98.14	95.89	0.9457
优化2	PA/%	84.62	95.74	97.72	90.69	95.72	95.82	95.89	0.9457
优化3	UA/%	93.33	99.05	98.27	91.25	90.65	95.12	96.06	0.9479
优化3	PA/%	78.65	92.40	99.07	91.73	94.94	97.43	96.06	0.9479

基于GEE多源遥感数据的干旱区植被地物类型提取

Vegetation feature type extraction in arid regions based on GEE multi-source remote sensing data

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 11

参考文献 31

相关文章 4

Metrics

本文评价

推荐阅读 0

[1]	赵雨琪, 魏天兴. 1990—2020年黄土高原典型县域植被覆盖变化及影响因素[J]. 干旱区研究, 2024, 41(1): 147-156.
[2]	李小雨, 贾科利, 魏慧敏, 陈睿华, 王怡婧. 基于随机森林算法的土壤含盐量预测[J]. 干旱区研究, 2023, 40(8): 1258-1267.
[3]	贾松涛,黄生志,王浩,李紫妍,黄强,梁浩. 基于BP-SVM模型的植被变化模拟研究[J]. 干旱区研究, 2021, 38(4): 1085-1093.
[4]	陈春秀,陈蜀江,徐世薇,陈孟禹,贾翔,黄铁成,李春蕾. 多特征辅助下的GF-6 WFV影像准噶尔山楂识别研究[J]. 干旱区研究, 2021, 38(2): 553-561.