[1]孙庆松,张晓楠,陈利东,等.基于Sentinel-2时序谐波特征的县域农作物分类[J].江苏农业学报,2022,38(04):967-975.[doi:doi:10.3969/j.issn.1000-4440.2022.04.013]
　SUN Qing-song,ZHANG Xiao-nan,CHEN Li-dong,et al.Crop classification in counties based on Sentinel-2 temporal harmonic characteristics[J].,2022,38(04):967-975.[doi:doi:10.3969/j.issn.1000-4440.2022.04.013]
点击复制

基于Sentinel-2时序谐波特征的县域农作物分类()
分享到：

江苏农业学报[ISSN:1006-6977/CN:61-1281/TN]

卷:
38
期数:
2022年04期
页码:
967-975
栏目:
农业信息工程
出版日期:
2022-08-31

文章信息/Info

Title:
Crop classification in counties based on Sentinel-2 temporal harmonic characteristics
作者:
孙庆松1; 张晓楠2; 陈利东3; 王坤3; 宋宏利1
（1.河北工程大学地球科学与工程学院，河北邯郸056038；2.河北工程大学矿业与测绘工程学院，河北邯郸056038；3.河北省地矿局第六地质大队，河北石家庄050000）
Author(s):
SUN Qing-song1; ZHANG Xiao-nan2; CHEN Li-dong3; WANG Kun3; SONG Hong-li1
(1.College of Geosciences and Engineering, Hebei University of Engineering, Handan 056038, China;2.College of Mining and Surveying Engineering, Hebei University of Engineering, Handan 056038, China;3.The Sixth Geological Brigade of Hebei Provincial Bureau of Geology and Mineral Resources, Shijiazhang 050000, China)
关键词:
Sentinel-2; 时序曲线; 谐波特征; 红边波段; 农作物精细化分类
Keywords:
Sentinel-2; time series curve; harmonic characteristics; red edge band; fine classification of crops
分类号:
S127
DOI:
doi:10.3969/j.issn.1000-4440.2022.04.013
文献标志码:
A
摘要:
以河北省南宫市主要农作物空间分布状况及面积信息为研究对象，采用多时相Sentinel-2影像数据，构建5种植被指数时序集，并通过傅里叶级数解算各时序曲线中的谐波特征参量，分别采用指数特征和指数特征+谐波特征2种分类依据，在随机森林框架下对10种分类方案进行农作物精细化分类。结果表明：当以指数特征+谐波特征作为分类依据时，5种时序集的总体分类精度比仅利用指数特征分类均有明显提高（最小提高8.14个百分点，最大提高9.21个百分点），表明谐波特征的加入能够有效提高分类精度。当以指数特征+谐波特征作为分类依据时，增强型植被指数（EVI）+归一化植被指数（NDVI）+红边归一化植被指数（NDVI705）组合总体分类精度最高，达到94.95%，比EVI+NDVI组合方案总体分类精度提高了2.57个百分点，说明含有红边波段的NDVI705可以有效提高分类精度。
Abstract:
Using spatial distribution status and area of the main crops in Nangong City of Hebei province as the research objects, multi-temporal Sentinel-2 image data were used to construct five time series sets for vegetation indices, and Fourier series were used to calculate the harmonic characteristic parameters in each time series curve. Two classification bases such as index feature and index feature + harmonic feature were used to classify crops finely by ten kinds of classification schemes under the framework of random forest. The results showed that, when the index feature + harmonic feature was used as the classification basis, the overall classification accuracy of five time series sets was obviously improved compared with using index feature as the classification basis (increased at least 8.14 percentage points and at most 9.21 percentage points). It was indicated that, the harmonic characteristics could improve the classification accuracy effectively. When the index feature + harmonic feature was used as the classification basis, the overall classification accuracy of the combination of enhanced vegetation index (EVI) + normalized difference vegetation index (NDVI) + red edge normalized difference vegetation index (NDVI705) was the highest, which reached 94.95%, and was 2.57 percentage points higher than the overall classification accuracy of the EVI+NDVI combination. The results revealed that, NDVI705 containing red edge band can effectively improve the classification accuracy.

参考文献/References:

[1]郭交，朱琳，靳标. 基于Sentinel-1和Sentinel-2数据融合的农作物分类[J]. 农业机械学报, 2018, 49(4): 192-198.
[2]陈仲新，任建强，唐华俊，等. 农业遥感研究应用进展与展望[J]. 遥感学报, 2016, 20(5): 748-767.
[3]赵春江. 农业遥感研究与应用进展[J]. 农业机械学报, 2014, 45(12): 277-293.
[4]WANG S, AZZAIR G, DAVID B L. Crop type mapping without field-level labels: Random forest transfer and unsupervised clustering techniques[J]. Remote Sensing of Environment, 2019, 222: 303-317.
[5]宋宏利，雷海梅，尚明. 基于Sentinel 2A/B时序数据的黑龙港流域主要农作物分类[J]. 江苏农业学报, 2021, 37(1): 83-92.
[6]郑利娟. 基于高分一/六号卫星影像特征的农作物分类研究[D]. 北京:中国科学院大学(中国科学院遥感与数字地球研究所), 2017.
[7]谷祥辉，张英，桑会勇，等. 基于哨兵2时间序列组合植被指数的作物分类研究[J]. 遥感技术与应用, 2020, 35(3): 702-711.
[8]杨欢，邓帆，张佳华，等. 基于MODIS EVI的江汉平原油菜和冬小麦种植信息提取研究[J]. 国土资源遥感, 2020, 32(3): 208-215.
[9]GOHAR G, OLENA D, FABIAN L, et al．A rule-based approach for crop identification using multi-temporal and multi-sensor phenological metrics[J]. European Journal of Remote Sensing, 2018, 51(1): 511-524.
[10]JAKUBAUSKAS M E, LEGATES D R, KASTENS J H. Crop identification using harmonic analysis of time-series AVHRR NDVI data[J]. Computers and Electronics in Agriculture, 2002, 37: 127-139.
[11]王丹，姜小光. 利用NOAA数据分析中国地区植被覆盖变化周期[J]. 中国图象图形学报, 2006, 11(4): 516-520.
[12]梁守真，邢前国，施平，等. 山东省典型地表覆被NDVI时间序列谐波分析[J]. 生态学杂志, 2011, 30(1): 59-65.
[13]JAKUBAUSKAS M E, LEGATES D R, KASTENS J H. Harmonic analysis of time-series AVHRR NDVI data[J]. Remote Sensing of Environment, 2001, 67(4): 461-467.
[14] 王笑影，周玉科，温日红. 基于Landsat-8影像和随机森林方法的土地分类研究[J]. 测绘与空间地理信息,2020, 43(11): 1-3.
[15]张晶,占玉林，李如仁. 高分一号归一化植被指数时间序列用于冬小麦识别[J]. 遥感信息, 2017, 32(1): 50-56.
[16]郭昱杉，刘庆生，刘高焕，等. 基于MODIS时序NDVI主要农作物种植信息提取研究[J]. 自然资源学报, 2017, 32(10): 1808-1818.
[17]解毅，张永清，荀兰，等. 基于多源遥感数据融合和LSTM算法的作物分类研究[J]. 农业工程学报, 2019, 35(15): 129-137.
[18]蔡文婷，赵书河，王亚梅，等.结合Sentinel-2光谱与纹理信息的冬小麦作物茬覆盖度估算[J].遥感学报,2020,24(9):1108-1119.
[19]杜保佳，张晶，王宗明，等. 应用Sentinel-2A NDVI时间序列和面向对象决策树方法的农作物分类[J].地球信息科学学报, 2019, 21(5):740-751.
[20]李中元，吴炳方，张淼，等. 利用物候差异与面向对象决策树提取油菜种植面积[J].地球信息科学学报,2019, 21(5): 720-730.
[21]吴静，吕玉娜，李纯斌，等. 基于多时相Sentinel-2A的县域农作物分类[J].农业机械学报, 2019, 50(9):194-200.
[22]郭文婷，张晓丽. 基于Sentinel-2时序多特征的植被分类[J]. 浙江农林大学学报, 2019, 36(5): 849-856.
[23]汪小钦，邱鹏勋，李娅丽，等. 基于时序Landsat遥感数据的新疆开孔河流域农作物类型识别[J]. 农业工程学报, 2019, 35(16): 180-188.
[24]韩红珠，白建军，张波，等. 基于MODIS时序的陕西省植被物候时空变化特征分析[J]. 国土资源遥感,2018, 30(4): 125-131.
[25]中央全面深化改革领导小组. 探索实行耕地轮作休耕制度试点方案[R/OL]. (2016-05-20)
[2019-05-20]. https://baidu.com/item/.
[26]EVANGELIDES C, NOBAJAS A. Red-edge normalised difference vegetation index (NDVI705) from sentinel-2 imagery to assess post-fire regeneration[J]. Remote Sensing Applications: Society and Environment, 2020, 17: 100283.
[27]BREIMAN L. Random forests[J]. Machine Learning, 2001, 45(1): 5-32.
[28]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 Sens, 2015, 7: 11249-11266.

相似文献/References:

[1]刘俊伟,陈鹏飞,张东彦,等.基于时序Sentinel-2影像的梨树县作物种植结构[J].江苏农业学报,2020,(06):1428.[doi:doi:10.3969/j.issn.1000-4440.2020.06.011]
　LIU Jun-wei,CHEN Peng-fei,ZHANG Dong-yan,et al.Crop planting structure in Lishu County based on time-series Sentinel-2 images[J].,2020,(04):1428.[doi:doi:10.3969/j.issn.1000-4440.2020.06.011]
[2]李亚妮,曹建君,杨树文,等.基于决策树的大尺度复杂地区夏收作物遥感提取与分析[J].江苏农业学报,2022,38(05):1257.[doi:doi:10.3969/j.issn.1000-4440.2022.05.012]
　LI Ya-ni,CAO Jian-jun,YANG Shu-wen,et al.Extraction and analysis of summer crops in large-scale complex areas based on decision tree[J].,2022,38(04):1257.[doi:doi:10.3969/j.issn.1000-4440.2022.05.012]
[3]于天祥,樊红.基于Sentinel-2多时相遥感影像的冬小麦种植面积监测[J].江苏农业学报,2024,(09):1653.[doi:doi:10.3969/j.issn.1000-4440.2024.09.009]
　YU Tianxiang,FAN Hong.Remote sensing monitoring of winter wheat planting area based on multi-temporal Sentinel-2 imagery[J].,2024,(04):1653.[doi:doi:10.3969/j.issn.1000-4440.2024.09.009]

备注/Memo

备注/Memo:
收稿日期：2021-11-22基金项目：河北省自然科学基金项目(D2019402067)；河北省地矿局创新团队项目（201908）作者简介：孙庆松（1997-），男，河北邢台人，硕士研究生，研究方向为农业遥感。(E-mail)1455995963@qq.com通讯作者：宋宏利，(E-mail)songhongli2003@163.com
更新日期/Last Update: 2022-09-06