[1]夏俊,苏涛,刘丽娜,等.基于多时相Sentinel-1A数据的水稻面积提取[J].江苏农业学报,2022,38(03):666-674.[doi:doi:10.3969/j.issn.1000-4440.2022.03.012]
　XIA Jun,SU Tao,LIU Li-na,et al.Rice area information extraction based on multi-temporal Sentinel-1A data[J].,2022,38(03):666-674.[doi:doi:10.3969/j.issn.1000-4440.2022.03.012]
点击复制

基于多时相Sentinel-1A数据的水稻面积提取()
分享到：

江苏农业学报[ISSN:1006-6977/CN:61-1281/TN]

卷:
38
期数:
2022年03期
页码:
666-674
栏目:
农业信息工程
出版日期:
2022-06-30

文章信息/Info

Title:
Rice area information extraction based on multi-temporal Sentinel-1A data
作者:
夏俊1; 2; 3;  苏涛1; 2; 3;  刘丽娜1; 2; 3;  王建1; 2; 3;  朱菲1; 2; 3;  廖晋一1; 2; 3
（1.安徽理工大学空间信息与测绘工程学院，安徽淮南232001；2.安徽理工大学矿山采动灾害空天地协同监测与预警安徽普通高校重点实验室，安徽淮南232001；3.安徽理工大学矿区环境与灾害协同监测煤炭行业工程研究中心，安徽淮南232001）
Author(s):
XIA Jun1; 2; 3; SU Tao1; 2; 3; LIU Li-na1; 2; 3;  WANG Jian1; 2; 3; ZHU Fei1; 2; 3; LIAO Jin-yi1; 2; 3
(1.School of Spatial Informatics and Geomatics Engineering, Anhui University of Science and Technology, Huainan 232001, China;2.Key Laboratory of Aviation-Aerospace-Ground Cooperative Monitoring and Early Warning of Coal Mining-induced Disasters of Anhui Higher Education Institutes, Anhui University of Science and Technology, Huainan 232001, China;3.Coal Industry Engineering Research Center of Mining Area Environmental and Disaster Cooperative Monitoring,Anhui University of Science and Technology， Huainan 232001, China)
关键词:
水稻面积; 长时间序列; Sentinel-1A; 微分变换; 阈值分类
Keywords:
rice area; long time series; Sentinel-1A; differential transformation; threshold classification
分类号:
TP79
DOI:
doi:10.3969/j.issn.1000-4440.2022.03.012
文献标志码:
A
摘要:
合成孔径雷达（SAR）具有全天候、全天时获取遥感影像的能力，因此在南方多阴雨地区有着较高的应用潜力。本研究以江苏省盐城市建湖县为研究区域，选用当地水稻生长周期内的长时间序列Sentinel-1A影像作为数据源，依据光谱微分变换分析法，采用一种雷达微分变换的方法，通过对长时间序列SAR影像进行一阶和二阶微分变换处理，选取其中水稻与其他地物后向散射系数差异明显的时间段，再利用支持向量机（SVM）模型进行分类从而获取水稻信息。与利用多时相极化SAR影像的阈值分类法进行比较可知，基于二阶微分变换的SVM分类方法优于阈值分类方法，其总体精度为89.88%，Kappa系数和F1值分别为0.841 2和0.879 5，水稻提取面积为525.32 km2，相对误差为11.58%。说明，经过微分变换的时序SAR数据结合SVM模型进行分类可以进一步提高水稻面积提取精度，为作物识别提供了一种新的思路。
Abstract:
Synthetic aperture radar (SAR) has a high application potential in cloudy and rainy areas of South China for its ability of obtaining remote sensing images in all-weather and all-time. In this study, Jianhu County of Yancheng City in Jiangsu province was taken as the research area, the long-time series Sentinel-1A images collected during the local rice growth cycle were selected as the data source. According to the spectral differential transformation analysis approach, a method of radar differential transformation was proposed. Through first-order and second-order differential transformation processings of SAR images of long-time series, time periods with obviously different backscattering coefficients between rice and other ground objects were selected. Then support vector machine (SVM) model was used for classification to obtain rice information. Compared with the threshold classification method based on multi-temporal and polarized SAR images, SVM classification method based on second-order differential transform was better than the threshold classification method. For the SVM classification method based on second-order differential transform, the overall accuracy was 89.88%, the Kappa coefficient and F1 value were 0.841 2 and 0.879 5 respectively, the rice extraction area was 525.32 km2, and the relative error was 11.58%. It can be concluded that the accuracy of rice area extraction can be further improved by classification of time series SAR data treated by differential transform combined with SVM model, which can provide a new idea for crop identification.

参考文献/References:

[1]孙姝娟,李民录,王萍,等. 基于GF-1/WFV EVI时间序列数据的水稻种植面积提取 [J]. 福建农业学报, 2018, 33(6): 575-580.
[2]牛海鹏,王占奇,肖东洋. 基于时空数据融合的县域水稻种植面积提取 [J]. 农业机械学报, 2020, 51(4): 156-163.
[3]张晓忆,李卫国,景元书,等. 多种光谱指标构建决策树的水稻种植面积提取 [J]. 江苏农业学报, 2016, 32(5): 1066-1072.
[4]孙政,周清波,杨鹏,等. 基于星载极化SAR数据的农作物分类识别进展评述 [J]. 中国农业资源与区划, 2019, 40(11): 63-71.
[5]东朝霞. 基于全极化SAR数据的旱地作物识别与生物学参数反演研究 [D].北京：中国农业科学院, 2016.
[6]PARK N W. Accounting for temporal contextual information in land-cover classification with multi-sensor SAR data [J]. International Journal of Remote Sensing, 2010, 31(1/2): 281-298.
[7]CLAUSS K, OTTINGER M, LEINENKUGEL P, et al. Estimating rice production in the Mekong Delta, Vietnam, utilizing time series of Sentinel-1 SAR data [J]. International Journal of Applied Earth Observation and Geoinformation, 2018, 73: 574-585.
[8]ZHANG Y, YANG B, LIU X H, et al. Estimation of rice grain yield from dual-polarization Radarsat-2 SAR data by integrating a rice canopy scattering model and a genetic algorithm [J]. International Journal of Applied Earth Observation And Geoinformation, 2017, 57: 75-85.
[9]杜烨,郭长青,文宁,等. 基于多时相COSMO-SkyMed SAR数据对水稻信息提取方法的研究与应用 [J]. 遥感信息, 2014, 29(3): 30-34.
[10]徐凯健,田庆久,徐念旭,等. 基于时序NDVI与光谱微分变换的森林优势树种识别 [J]. 光谱学与光谱分析, 2019, 39(12): 3794-3800.
[11]FIEUZAL R, BAUP F, MARAISSICRE C. Monitoring wheat and rapeseed by using synchronous optical and radar satellite data—from temporal signatures to crop parameters estimation [J]. Advances in Remote Sensing, 2013, 2(2): 162-180.
[12]张东,塔西甫拉提·特依拜,张飞,等. 分数阶微分算法对盐渍土高光谱数据的影响研究 [J]. 光学学报, 2016, 36(3): 282-289.
[13]徐念旭,田庆久,申怀飞,等. 基于微分变换的高光谱马尾松和杉木识别 [J]. 国土资源遥感, 2018, 30(4): 28-32.
[14]万意,李长春,赵旭辉,等. 基于SVM的光学遥感影像分类与评价 [J]. 测绘地理信息, 2018, 43(6): 74-77.
[15]高燕,周成虎,苏奋振. 基于OLI影像多参数设置的SVM分类研究 [J]. 测绘工程, 2014, 23(6): 1-5，10.
[16]白燕英,高聚林,张宝林. 基于NDVI与EVI的作物长势监测研究 [J]. 农业机械学报, 2019, 50(9): 153-161.
[17]贾明权. 水稻微波散射特性研究及参数反演 [D]. 成都：电子科技大学, 2013.
[18]何泽. 基于多时相RADARSAT-2数据的水稻物候监测 [D]. 成都：电子科技大学, 2019.
[19]刘康,BALZ T,廖明生. 利用后向散射特性从高分辨率SAR影像中提取建筑物高度 [J]. 武汉大学学报(信息科学版), 2012, 37(7): 806-809.
[20]赵凌君. 高分辨率SAR图像建筑物提取方法研究 [D]. 长沙：国防科技大学, 2009.

备注/Memo

备注/Memo:
收稿日期：2021-11-08基金项目：国家重点研发计划项目（2018YFC0407703）；安徽理工大学引进人才科研启动项目（ZY030）；安徽理工大学2021年研究生创新基金项目（2021CX2139）；安徽理工大学青年教师科学研究基金项目（QN201502）作者简介：夏俊（1996-），男，安徽合肥人，硕士研究生，主要研究方向为农业遥感。（E-mail）1849145280@qq.com通讯作者：苏涛，（E-mail）st7162003@163.com
更新日期/Last Update: 2022-07-07