Please wait a minute...
 
国土资源遥感  2018, Vol. 30 Issue (1): 196-202    DOI: 10.6046/gtzyyg.2018.01.27
  本期目录 | 过刊浏览 | 高级检索 |
基于GF-2卫星数据的孕穗期小麦叶面积指数反演——以河北省廊坊市为例
陆坤1(), 孟庆岩2(), 孙云晓2,3, 孙震辉2,3, 张琳琳2,3
1.山东科技大学测绘科学与工程学院,青岛 266590
2.中国科学院遥感与数字地球研究所,北京 100101
3.中国科学院大学,北京 100049
Estimating leaf area index of wheat at the booting stage using GF-2 data: A case study of Langfang City,Hebei Province
Kun LU1(), Qingyan MENG2(), Yunxiao SUN2,3, Zhenhui SUN2,3, Linlin ZHANG2,3
1.Geotramics College,Shandong University of Science and Technology,Qingdao 266590,China
2.Institute of Remote Sensing and Digital Earth,Chinese Academy of Sciences,Beijing 100101,China
3.University of Chinese Academy of Sciences,Beijing 100049,China
全文: PDF(968 KB)   HTML  
输出: BibTeX | EndNote (RIS)      
摘要 

叶面积指数(leaf area index,LAI)是评价植被长势和预测产量的重要农业生理生态参数。高分2号(GF-2)卫星数据具有高空间分辨率特点,能反映更多细节信息,针对该数据特点的LAI反演方法具有较高的研究价值。以河北省廊坊市万庄镇为研究区,对孕穗期小麦采用了回归模型和神经网络算法反演LAI; 采用4种植被指数与实测LAI值构建回归模型,同时重点探讨了PROSAIL模型结合神经网络方法进行LAI反演。研究结果表明,在回归模型中,归一化植被指数(normalized difference vegetation index,NDVI)的二项式模型估算LAI可以获得最高精度,采用实测数据验证的决定系数(R2)和均方根误差(root mean square error,RMSE)分别为0.719 3和0.393 6; 与回归模型相比,神经网络反演LAI方法更显著提高了精度,R2RMSE分别达到0.900 8和0.273 2。基于GF-2卫星数据,在研究区小麦孕穗期,神经网络反演LAI具有较强可行性和适用性,可为高空间分辨率卫星影像的LAI反演提供参考。

服务
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章
陆坤
孟庆岩
孙云晓
孙震辉
张琳琳
关键词 植被指数神经网络叶面积指数(LAI)反演方法精度验证    
Abstract

Leaf area index (LAI) is an important agricultural parameter to assess crop growing status for production estimation. Due to its very high spatial resolution, GF-2 can be used as a new source of remote sensing data for crop monitoring. It is particularly valuable to develop approaches for LAI estimation using GF-2 data. In this paper, the study of LAI estimation for wheat at the booting stage in Wanzhuang Town, Langfang City of Hebei Province in North China Plain is presented. Canopy LAI of wheat at the booting stage was measured over an experimental field in the town. Regression analysis method was performed with LAI and four different vegetation indexes, and the neural network method combined with the PROSAIL model was also considered. The results show that the best LAI retrieval regression model is the binomial model of normalized difference vegetation index (NDVI). The correlation coefficient (R2) and root mean square error (RMSE) are 0.719 3 and 0.393 6 respectively. Compared with the regression analysis method, the accuracy is greatly improved by using neural network method,and the R2 and RMSE reached 0.900 8 and 0.273 2 respectively. Neural network method has feasibility and applicability based GF-2 data at wheat booting stage, which would provide a reference scheme for LAI inversion from high spatial resolution satellite image.

Key wordsvegetation index    neural network    leaf area index(LAI)    inversion method    precision validation
收稿日期: 2016-08-05      出版日期: 2018-02-08
:  TP79  
基金资助:高分辨率对地观测系统国家重大科技专项项目“基于GF-4卫星数据的特征参数反演技术”(编号: 11-Y20A05-9001-15/16)、广东省省级科技计划项目“中泰农业环境高分辨率遥感监测与示范”(编号: 2014A050503060)和国家自然科学基金项目“基于机载LiDAR数据的城市绿度空间指数模型研究”(编号: 41471310)共同资助
作者简介:

第一作者: 陆 坤(1990-),男,硕士,主要从事定量遥感、农业遥感方面的研究。Email:3298717632@qq.com

引用本文:   
陆坤, 孟庆岩, 孙云晓, 孙震辉, 张琳琳. 基于GF-2卫星数据的孕穗期小麦叶面积指数反演——以河北省廊坊市为例[J]. 国土资源遥感, 2018, 30(1): 196-202.
Kun LU, Qingyan MENG, Yunxiao SUN, Zhenhui SUN, Linlin ZHANG. Estimating leaf area index of wheat at the booting stage using GF-2 data: A case study of Langfang City,Hebei Province. Remote Sensing for Land & Resources, 2018, 30(1): 196-202.
链接本文:  
https://www.gtzyyg.com/CN/10.6046/gtzyyg.2018.01.27      或      https://www.gtzyyg.com/CN/Y2018/V30/I1/196
波段号 光谱
波段
波段范
围/μm
空间分
辨率/m
幅宽/km 侧摆能
力/(°)
重访时
间/d
1 0.45~0.52 4 45 ±35 5
2 绿 0.52~0.59 4 45 ±35 5
3 0.63~0.69 4 45 ±35 5
4 近红外 0.77~0.89 4 45 ±35 5
5 全色 0.45~0.90 1 45 ±35 5
Tab.1  GF-2卫星波段设置
Fig.1  FLAASH大气校正后植被反射率光谱曲线
植被指数类型 计算公式
NDVI[16] NDVI=ρnir-ρredρnir+ρred
比值植被指数[17] RVI=ρnirρred
土壤调节植被指数[18] SAVI=ρnir-ρredρnir+ρred+L(L+1)
EVI219 EVI2=2.5ρnir-ρred(ρnir+2.4ρred)+1
Tab.2  本文使用的植被指数
Fig.2  BP-ANN结构示意图
Fig.3  VI-LAI回归拟合
植被指数
回归模型
拟合方程 R2 RMSE
NDVI-LAI y = 11.312 x - 2.495 9 0.700 1 0.348 7
y = 0.251 1 e5.123x 0.705 8 0.335 3
y = 4.615 5 ln x + 6.287 0.683 7 0.358 2
y = 30.586 x2-14.286 x+2.786 8 0.718 5 0.337 9
RVI-LAI y=1.560 8 x - 1.649 0.641 5 0.381 3
y =0.374 1 e0.700 5x 0.635 2 0.368 8
y =3.719 9 ln x - 1.127 2 0.615 3 0.395 0
y=0.664 3 x2-1.742 6 x+2.380 7 0.664 9 0.368 6
SAVI-LAI y =5.918 6 x - 1.513 0.601 8 0.545 5
y =0.383 5 e2.716 4x 0.623 2 0.382 5
y =3.400 5 ln x + 3.807 7 0.561 1 0.421 9
y =15.543 x2-13.027 x+4.155 4 0.656 9 0.373 0
EVI2-LAI y =4.219 2 x-0.946 3 0.615 0 0.395 2
y =0.501 7 e1.924 6x 0.629 0 0.375 9
y =2.867 3 ln x+3.07 0.569 8 0.417 7
y =6.989 9 x2-5.985 1 x+2.681 3 0.659 2 0.371 8
Tab.3  回归模型构建
Fig.4  BP-ANN模型流程

神经网络预测LAI 实测LAI
预测1 预测2 预测3 预测4 预测5
1 2.573 5 2.868 5 2.255 5 2.548 0 2.487 7 2.282 5
2 3.308 5 3.059 1 3.188 7 3.154 6 3.431 2 3.113 0
3 2.539 3 2.300 0 2.199 3 2.446 2 2.475 5 2.203 0
4 3.135 2 3.039 3 3.018 5 3.172 5 3.622 1 3.374 0
5 2.099 2 2.193 6 2.023 2 2.059 5 2.294 4 1.792 0
6 1.984 0 1.833 9 1.806 7 1.843 0 1.642 2 1.731 0
7 1.950 3 1.598 4 1.903 0 1.764 2 1.767 6 1.626 0
8 2.789 2 2.987 8 2.366 6 2.561 2 2.644 0 2.619 0
9 1.576 6 1.507 2 1.528 5 1.705 6 1.405 1 1.180 0
10 2.320 6 2.482 1 2.206 5 2.118 3 2.010 8 2.334 0
11 2.768 0 2.990 0 2.708 2 2.530 8 2.660 3 3.109 0
Tab.4  神经网络模型测试结果
Fig.5  LAI反演结果对比
Fig.6  LAI反演结果验证
[1] Chen J M,Cihlar J.Retrieving leaf area index of boreal conifer forests using Landsat TM images[J].Remote Sensing of Environment,1996,55(2):153-162.
[2] Pinty B,Lavergne T,Widlowski J L,et al.On the need to observe vegetation canopies in the near-infrared to estimate visible light absorption[J].Remote Sensing of Environment,2009,113(1):10-23.
[3] 李鑫川,徐新刚,鲍艳松,等.基于分段方式选择敏感植被指数的冬小麦叶面积指数遥感反演[J].中国农业科学,2012,45(17):3486-3496.
Li X C,Xu X G,Bao Y S,et al.Retrieving LAI of winter wheat based on sensitive vegetation index by the segmentation method[J].Scientia Agricultura Sinica,2012,45(17):3486-3496.
[4] Yang F,Sun J L,Fang H L,et al.Comparison of different methods for corn LAI estimation over northeastern China[J].International Journal of Applied Earth Observation and Geoinformation,2012,18:462-471.
[5] 任哲,陈怀亮,王连喜,等.利用交叉验证的小麦LAI反演模型研究[J].国土资源遥感,2015,27(4):34-40.doi:10.6046/gtzyyg.2015.04.06.
Ren Z,Chen H L,Wang L X,et al.Research on inversion model of wheat LAI using cross-validation[J].Remote Sensing for Land and Resources,2015,27(4):34-40.doi:10.6046/gtzyyg.2015.04.06.
[6] Meng Q Y,Wang C M,Gu X F,et al.Hot dark spot index method based on multi-angular remote sensing for leaf area index retrieval[J].Environmental Earth Sciences,2016,75:732.
[7] 张瀛,孟庆岩,武佳丽,等.基于环境星CCD数据的环境植被指数及叶面积指数反演研究[J].光谱学与光谱分析,2011,31(10):2789-2793.
Zhang Y,Meng Q Y,Wu J L,et al.Study of environmental vegetation index based on environment satellite CCD data and LAI inversion[J].Spectroscopy and Spectral Analysis,2011,31(10):2789-2793.
[8] 赵娟. 基于多时相多角度遥感数据的植被叶面积指数反演[D].南京:南京信息工程大学,2014.
Zhao J.Study of Leaf Area Index Inversion Based on Multi-Temporal and Multi-angular Remote Sensing Data[D].Nanjing:Nanjing University of Science and Technology,2014.
[9] Kokaly R F,Clark R N.Spectroscopic determination of leaf biochemistry using band-depth analysis of absorption features and stepwise multiple linear regression[J].Remote Sensing of Environment,1999,67(3):267-287.
[10] 宋开山,张伯,王宗明,等.基于人工神经网络的大豆叶面积高光谱反演研究[J].中国农业科学,2006,39(6):1138-1145.
Song K S,Zhang B,Wang Z M,et al.Soybean LAI estimation with in-situ collected hyperspectral data based on BP-neural networks[J].Scientia Agricultura Sinica,2006,39(6):1138-1145.
[11] 陈健,倪绍祥,李云梅.基于神经网络方法的芦苇叶面积指数遥感反演[J].国土资源遥感,2008,20(2):62-67.doi:10.6046/gtzyyg.2008.02.15.
Chen J,Ni S X,Li Y M.LAI retrieval of reed canopy using the neural network method[J].Remote Sensing for Land and Resources,2008,20(2):62-67.doi:10.6046/gtzyyg.2008.02.15.
[12] 包刚,覃志豪,周义,等.基于高光谱数据和RBF神经网络方法的草地叶面积指数反演[J].国土资源遥感,2012,24(2):7-11.doi:10.6046/gtzyyg.2012.02.02.
Bao G,Qin Z H,Zhou Y,et al.The application of hyper-spectral data and RBF neural network method to retrieval of leaf area index of grassland[J].Remote Sensing for Land and Resources,2012,24(2):7-11.doi:10.6046/gtzyyg.2012.02.02.
[13] Adler-Golden S M,Matthew M W,Bernstein L S,et al.Atmospheric correction for shortwave spectral imagery based on MODTRAN4[C]//Proceeding of the SPIE 3753,Imaging Spectrometry V.Denver,CO,United States:SPIE,1999,3753:61-69.
[14] 郝建亭,杨武年,李玉霞,等.基于FLAASH的多光谱影像大气校正应用研究[J]. 遥感信息, 2008(1):78-81.
Hao J T,Yang W N,Li Y X,et al.Atmospheric correction of multi-spectral imagery ASTER[J]. Remote Sensing Information, 2008(1):78-81.
[15] Gonsamo A,Pellikka P.The sensitivity based estimation of leaf area index from spectral vegetation indices[J].ISPRS Journal of Photogrammetry and Remote Sensing,2012,70:15-25.
[16] Deering D W.Rangeland Reflectance Characteristics Measured by Aircraft and Spacecraft Sensors[D].College Station,Texas:Texas A and M University,1978:338.
[17] Chen J M,Pavlic G,Brown L,et al.Derivation and validation of Canada-wide coarse-resolution leaf area index maps using high-resolution satellite imagery and ground measurements[J].Remote Sensing of Environment,2002,80(1):165-184.
[18] Huete A R.A soil-adjusted vegetation index(SAVI)[J].Remote Sensing of Environment,1988,25(3):295-309.
[19] Jiang Z Y,Huete A R,Didan K,et al.Development of a two-band enhanced vegetation index without a blue band[J].Remote Sensing of Environment,2008,112(10):3833-3845.
[20] Baret F,Jacquemoud S,Guyot G,et al.Modeled analysis of the biophysical nature of spectral shifts and comparison with information content of broad bands[J].Remote Sensing of Environment,1992,41(2/3):133-142.
[21] 谷成燕,杜华强,周国模,等.基于PROSAIL辐射传输模型的毛竹林叶面积指数遥感反演[J].应用生态学报,2013,24(8):2248-2256.
Gu C Y,Du H Q,Zhou G M,et al.Retrieval of leaf area index of Moso bamboo forest with Landsat Thematic Mapper image based on PROSAIL canopy radiative transfer model[J].Chinese Journal of Applied Ecology,2013,24(8):2248-2256.
[22] 蔡博峰,绍霞.基于PROSPECT+SAIL模型的遥感叶面积指数反演[J].国土资源遥感,2007,19(2):39-43.doi:10.6046/gtzyyg.2007.02.10.
Cai B F,Shao X.Leaf area index petrieval based on remotely sensed data and PROSPECT+SAIL model[J].Remote Sensing for Land and Resources,2007,19(2):39-43.doi:10.6046/gtzyyg.2007.02.10.
[23] 余金林. 基于PROSAIL模型的青海湖流域草地叶面积指数反演[D].西宁:青海师范大学,2014.
Yu J L.Leaf Area Index Inversion of Grass in Qinghai Lake Basin Based PROSAIL Model[D].Xining:Qinghai Normal University,2014.
[1] 姚金玺, 张志, 张焜. 基于GEE的诺木洪洪积扇植被时空变化特征、成因及趋势分析[J]. 自然资源遥感, 2022, 34(1): 249-256.
[2] 张成业, 邢江河, 李军, 桑潇. 基于U-Net网络和GF-6影像的尾矿库空间范围识别[J]. 自然资源遥感, 2021, 33(4): 252-257.
[3] 于冰, 谭青雪, 刘国祥, 刘福臻, 周志伟, 何智勇. 高分辨率TerraSAR-X时序差分干涉沉降监测及精度验证[J]. 自然资源遥感, 2021, 33(4): 26-33.
[4] 于新莉, 宋妍, 杨淼, 黄磊, 张艳杰. 结合空间约束的卷积神经网络多模型多尺度船企场景识别[J]. 自然资源遥感, 2021, 33(4): 72-81.
[5] 李阳, 袁琳, 赵志远, 张晋磊, 王宪业, 张利权. 基于无人机低空遥感和现场调查的潮滩地形反演研究[J]. 自然资源遥感, 2021, 33(3): 80-88.
[6] 刘万军, 高健康, 曲海成, 姜文涛. 多尺度特征增强的遥感图像舰船目标检测[J]. 自然资源遥感, 2021, 33(3): 97-106.
[7] 吴霞, 王长军, 樊丽琴, 李磊. 基于多光谱遥感的盐渍化评价指数对宁夏银北灌区土壤盐度预测的适用性分析[J]. 国土资源遥感, 2021, 33(2): 124-133.
[8] 宋承运, 胡光成, 王艳丽, 汤超. 基于表观热惯量与温度植被指数的FY-3B土壤水分降尺度研究[J]. 国土资源遥感, 2021, 33(2): 20-26.
[9] 胡新宇, 许章华, 陈文慧, 陈秋霞, 王琳, 刘辉, 刘智才. 基于PROBA/CHRIS影像的归一化阴影植被指数NSVI构建与应用效果[J]. 国土资源遥感, 2021, 33(2): 55-65.
[10] 仇一帆, 柴登峰. 无人工标注数据的Landsat影像云检测深度学习方法[J]. 国土资源遥感, 2021, 33(1): 102-107.
[11] 刘钊, 赵桐, 廖斐凡, 李帅, 李海洋. 基于语义分割网络的高分遥感影像城市建成区提取方法研究与对比分析[J]. 国土资源遥感, 2021, 33(1): 45-53.
[12] 卫虹宇, 赵银娣, 董霁红. 基于改进RetinaNet的冷却塔目标检测[J]. 国土资源遥感, 2020, 32(4): 68-73.
[13] 刘钊, 廖斐凡, 赵桐. 基于PSPNet的遥感影像城市建成区提取及其优化方法[J]. 国土资源遥感, 2020, 32(4): 84-89.
[14] 李宇, 肖春姣, 张洪群, 李湘眷, 陈俊. 深度卷积融合条件随机场的遥感图像语义分割[J]. 国土资源遥感, 2020, 32(3): 15-22.
[15] 高凯旋, 焦海明, 王新闯. 融合影像纹理、光谱与地形特征的森林冠顶高反演模型[J]. 国土资源遥感, 2020, 32(3): 63-70.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
京ICP备05055290号-2
版权所有 © 2015 《自然资源遥感》编辑部
地址:北京学院路31号中国国土资源航空物探遥感中心 邮编:100083
电话:010-62060291/62060292 E-mail:zrzyyg@163.com
本系统由北京玛格泰克科技发展有限公司设计开发