基于Sentinel-2A影像的县域冬小麦种植面积遥感监测
|
摘要
小麦种植面积遥感监测是小麦估产的基本要素,准确而及时地提取不同灌溉类型冬小麦种植面积及其空间分布信息可为冬小麦长势监测以及产量评估提供科学依据。以山西省闻喜县冬小麦为研究对象,以Sentinel-2A影像为基础数据源,选择主成分(PCA)、红边归一化植被指数(RENDVI)、纹理特征等3个特征变量,结合实地调查样本点,采用随机森林算法,提取冬小麦种植面积,并结合数字高程模型(DEM)提取雨养区和灌溉区冬小麦种植面积。结果表明,Sentinel-2A遥感数据适合作为县域尺度冬小麦监测的数据源;主成分分析、纹理特征和RENDVI的引入可以提高单时相遥感影像对县域冬小麦分类的识别能力;随机森林算法和数字高程模型结合可以实现雨养区和灌溉区冬小麦种植面积的提取。
Monitoring of wheat planting area by using remote sensing technology is an essential element for wheat yield estimation.Accurately and timely extraction of winter wheat planting area and spatial distribution information of different irrigation types can provide a scientific basis for winter wheat growth monitoring and yield assessment. In this study, the winter wheat in Wenxi county of Shanxi province and the Sentinel-2 A satellite were made as the research object and the image data, respectively. The combination of three characteristic variables, including principal components(PC), red edge normalized difference vegetation index(RENDVI), and texture features, was selected and combined with the field survey samples. Moreover, digital elevation model(DEM)was combined with random forest algorithm to extract the area of winter wheat in rainfed and irrigation land. The results showed that Sentinel-2 A remote sensing data was suitable as a data source for winter wheat monitoring at a county scale. Principal component analysis, texture features and the introduction of RENDVI could improve the ability of single-temporal remote sensing imagery to classify winter wheat in a county. The combination of random forest algorithm and digital elevation model could realize the extraction of winter wheat planting area in rainfed and irrigated areas.
Monitoring of wheat planting area by using remote sensing technology is an essential element for wheat yield estimation.Accurately and timely extraction of winter wheat planting area and spatial distribution information of different irrigation types can provide a scientific basis for winter wheat growth monitoring and yield assessment. In this study, the winter wheat in Wenxi county of Shanxi province and the Sentinel-2 A satellite were made as the research object and the image data, respectively. The combination of three characteristic variables, including principal components(PC), red edge normalized difference vegetation index(RENDVI), and texture features, was selected and combined with the field survey samples. Moreover, digital elevation model(DEM)was combined with random forest algorithm to extract the area of winter wheat in rainfed and irrigation land. The results showed that Sentinel-2 A remote sensing data was suitable as a data source for winter wheat monitoring at a county scale. Principal component analysis, texture features and the introduction of RENDVI could improve the ability of single-temporal remote sensing imagery to classify winter wheat in a county. The combination of random forest algorithm and digital elevation model could realize the extraction of winter wheat planting area in rainfed and irrigated areas.
引文
[1]刘佳,王利民.农作物面积遥感监测原理与实践[M].北京:科学出版社,2017.
[2]黄双燕,杨辽,陈曦,等.机器学习法的干旱区典型农作物分类[J].光谱学与光谱分析,2018,38(10):3169-3176.
[3]刘哲,刘帝佑,朱德海,等.作物遥感精细识别与自动制图研究进展与展望[J].农业机械学报,2018(12):1-12.
[4]田海峰,周伯燕,陈燕芬,等.县域尺度上基于GF-1PMS影像的冬小麦种植面积遥感监测[J].中国农业大学学报,2017,22(10):139-146.
[5] BELGIU M,CSILLIK O. Sentinel-2 cropland mapping using pixel-based and object-based time-weighted dynamic time warping analysis[J]. Remote Sensing of Environment,2018,204:509-523.
[6] ZHANG M,LIN H. Object-based rice mapping using time-series and phenological data[J]. Advances in Space Research,2019,63(1):190-202.
[7] SONOBE R,YAMAYA Y,TANI H,et al. Crop classification from Sentinel-2-derived vegetation indices using ensemble learning[J].Journal of Applied Remote Sensing,2018,12(2):1-17.
[8]史飞飞,雷春苗,肖建设,等.基于多源遥感数据的复杂地形区农作物分类[J].地理与地理信息科学,2018,34(5):49-55.
[9]吕雅慧,张超,郧文聚,等.高分辨率遥感影像农田林网自动识别[J].农业机械学报,2018,49(1):157-163.
[10] CHUANG Y C M,SHIU Y S. A comparative analysis of machine learning with WorldView-2 pan-sharpened imagery for tea crop mapping[J]. Sensors,2016,16(5):594.
[11]GAO J,NUYTTENS D,LOOTENS P,et al. Recognising weeds in a maize crop using a random forest machine-learning algorithm and near-infrared snapshot mosaic hyperspectral imagery[J]. Biosystems Engineering,2018,170:39-50.
[12]杨武德,宋艳暾,宋晓彦,等.基于3S和实测相结合的冬小麦估产研究[J].农业工程学报,2009,25(2):131-135,315.
[13] BELGIU M,DR魤GUT L. Random forest in remote sensing:A review of applications and future directions[J]. ISPRS Journal of Photogrammetry and Remote Sensing,2016,114:24-31.
[14] HAO P,ZHAN Y,WANG L,et al. Feature selection of time series MODIS data for early crop classification using random forest:A case study in Kansas,USA[J]. Remote Sensing,2015,7(5):5347-5369.
[15] PELLETIER C,VALERO S,INGLADA J,et al. Assessing the robustness of Random Forests to map land cover with high resolution satellite image time series over large areas[J]. Remote Sensing of Environment,2016,187:156-168.
[16]王利军,郭燕,贺佳,等.基于决策树和SVM的Sentinel-2A影像作物提取方法[J].农业机械学报,2018,49(9):146-153.
[17]邓书斌. ENVI遥感图像处理方法[M]. 2版.北京:科学出版社,2014.
[18]刘丽,匡纲要.图像纹理特征提取方法综述[J].中国图象图形学报,2009,14(4):622-635.
[19]侯群群,王飞,严丽.基于灰度共生矩阵的彩色遥感图像纹理特征提取[J].国土资源遥感,2013,25(4):26-32.
[20]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.
[21]王雷,龙永清,杨勤科.重采样方法对DEM数据质量的影响[J].水土保持通报,2016,36(4):72-77.
[22]赵英时.遥感应用分析原理与方法[M]. 2版.北京:科学出版社,2013.
[23] VAN DER LINDEN S,RABE A,HELD M,et al. The EnMAP-Box-A toolbox and application programming interface for EnMAP data processing[J]. Remote Sensing,2015,7(9):11249-11266.
[24]张晓羽,李凤日,甄贞,等.基于随机森林模型的陆地卫星-8遥感影像森林植被分类[J].东北林业大学学报,2016,44(6):53-57,74.
[25]黄健熙,侯矞焯,武洪峰,等.基于时间序列MODIS的农作物类型空间制图方法[J].农业机械学报,2017,48(10):142-147,285.
[26]冯美臣,杨武德,张东彦,等.基于TM和MODIS数据的水旱地冬小麦面积提取和长势监测[J].农业工程学报,2009,25(3):103-109,313.
[2]黄双燕,杨辽,陈曦,等.机器学习法的干旱区典型农作物分类[J].光谱学与光谱分析,2018,38(10):3169-3176.
[3]刘哲,刘帝佑,朱德海,等.作物遥感精细识别与自动制图研究进展与展望[J].农业机械学报,2018(12):1-12.
[4]田海峰,周伯燕,陈燕芬,等.县域尺度上基于GF-1PMS影像的冬小麦种植面积遥感监测[J].中国农业大学学报,2017,22(10):139-146.
[5] BELGIU M,CSILLIK O. Sentinel-2 cropland mapping using pixel-based and object-based time-weighted dynamic time warping analysis[J]. Remote Sensing of Environment,2018,204:509-523.
[6] ZHANG M,LIN H. Object-based rice mapping using time-series and phenological data[J]. Advances in Space Research,2019,63(1):190-202.
[7] SONOBE R,YAMAYA Y,TANI H,et al. Crop classification from Sentinel-2-derived vegetation indices using ensemble learning[J].Journal of Applied Remote Sensing,2018,12(2):1-17.
[8]史飞飞,雷春苗,肖建设,等.基于多源遥感数据的复杂地形区农作物分类[J].地理与地理信息科学,2018,34(5):49-55.
[9]吕雅慧,张超,郧文聚,等.高分辨率遥感影像农田林网自动识别[J].农业机械学报,2018,49(1):157-163.
[10] CHUANG Y C M,SHIU Y S. A comparative analysis of machine learning with WorldView-2 pan-sharpened imagery for tea crop mapping[J]. Sensors,2016,16(5):594.
[11]GAO J,NUYTTENS D,LOOTENS P,et al. Recognising weeds in a maize crop using a random forest machine-learning algorithm and near-infrared snapshot mosaic hyperspectral imagery[J]. Biosystems Engineering,2018,170:39-50.
[12]杨武德,宋艳暾,宋晓彦,等.基于3S和实测相结合的冬小麦估产研究[J].农业工程学报,2009,25(2):131-135,315.
[13] BELGIU M,DR魤GUT L. Random forest in remote sensing:A review of applications and future directions[J]. ISPRS Journal of Photogrammetry and Remote Sensing,2016,114:24-31.
[14] HAO P,ZHAN Y,WANG L,et al. Feature selection of time series MODIS data for early crop classification using random forest:A case study in Kansas,USA[J]. Remote Sensing,2015,7(5):5347-5369.
[15] PELLETIER C,VALERO S,INGLADA J,et al. Assessing the robustness of Random Forests to map land cover with high resolution satellite image time series over large areas[J]. Remote Sensing of Environment,2016,187:156-168.
[16]王利军,郭燕,贺佳,等.基于决策树和SVM的Sentinel-2A影像作物提取方法[J].农业机械学报,2018,49(9):146-153.
[17]邓书斌. ENVI遥感图像处理方法[M]. 2版.北京:科学出版社,2014.
[18]刘丽,匡纲要.图像纹理特征提取方法综述[J].中国图象图形学报,2009,14(4):622-635.
[19]侯群群,王飞,严丽.基于灰度共生矩阵的彩色遥感图像纹理特征提取[J].国土资源遥感,2013,25(4):26-32.
[20]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.
[21]王雷,龙永清,杨勤科.重采样方法对DEM数据质量的影响[J].水土保持通报,2016,36(4):72-77.
[22]赵英时.遥感应用分析原理与方法[M]. 2版.北京:科学出版社,2013.
[23] VAN DER LINDEN S,RABE A,HELD M,et al. The EnMAP-Box-A toolbox and application programming interface for EnMAP data processing[J]. Remote Sensing,2015,7(9):11249-11266.
[24]张晓羽,李凤日,甄贞,等.基于随机森林模型的陆地卫星-8遥感影像森林植被分类[J].东北林业大学学报,2016,44(6):53-57,74.
[25]黄健熙,侯矞焯,武洪峰,等.基于时间序列MODIS的农作物类型空间制图方法[J].农业机械学报,2017,48(10):142-147,285.
[26]冯美臣,杨武德,张东彦,等.基于TM和MODIS数据的水旱地冬小麦面积提取和长势监测[J].农业工程学报,2009,25(3):103-109,313.