Please wait a minute...
img

官方微信

遥感技术与应用  2020, Vol. 35 Issue (1): 174-184    DOI: 10.11873/j.issn.1004-0323.2020.1.0174
模型与反演     
基于BP神经网络的夏玉米多生育期叶面积指数反演研究
刘俊1(),孟庆岩2,3(),葛小三1,刘顺喜4,陈旭2,3,孙云晓2,3
1. 河南理工大学测绘与国土信息工程学院,河南 焦作 454000
2. 中国科学院遥感与数字地球研究所,北京 100101
3. 三亚中科遥感研究所,海南 三亚572029
4. 中国土地勘测规划院,北京 100035
Leaf Area Index Inversion of Summer Maize at Multiple Growth Stages based on BP Neural Network
Jun Liu1(),Qingyan Meng2,3(),Xiaosan Ge1,Shunxi Liu4,Xu Chen2,3,Yunxiao Sun2,3
1. School of Surveying and Land Information Engineering, Henan Polytechnic University, Jiaozuo 454000, China
2. Institute of Remote Sensing and Digital Earth, Chinese Academy of Sciences, Beijing 100101, China
3. Sanya Institute of Remote Sensing, Sanya 572029, China
4. China Land Surveying & Planning Institute, Beijing 100035, China
 全文: PDF(6000 KB)   HTML
摘要:

叶面积指数(Leaf Area Index,LAI)是生物地球化学循环中重要的植被结构参数。针对目前基于我国GF-1 WFV卫星影像的夏玉米多生育期LAI反演研究较少的问题,基于不同隐含层构建BP神经网络模型(BP1模型和BP2模型),对比分析BP1模型、BP2模型和6种统计模型(NDVI、RVI、DVI、EVI、SAVI、ARVI)反演之间的精度差异,并根据实测数据绘制BP1模型和BP2模型的夏玉米多生育期LAI动态变化图。结果表明:LAI与6种常用的统计模型均有良好相关性,其中NDVI指数方程式回归模型拟合度最优;BP神经网络模型整体R 2略小于统计模型,而RMSE则小于统计模型,取得了与实测值差异更小的结果,统计模型与BP神经网络模型各有优劣之处;BP2模型在R 2和RMSE均优于BP1模型,能获得更为精确的反演值,BP2整体预测精度更高;基于BP神经网络模拟夏玉米生育期反演,LAI值呈现缓慢升高—快速增长—逐渐减小的S型变化过程,基本符合作物生长规律。该研究结合不同隐含层建立的BP神经网络模型,为GF-1卫星在作物叶面积指数多生育期反演的应用推广提供了方法支撑。

关键词: 夏玉米叶面积指数BP神经网络模型统计模型多生育期    
Abstract:

Leaf Area Index (LAI) is an important vegetation structure parameter in biogeochemical cycling. In view of the lack of LAI inversion in the multiple growth period of summer maize based on GF-1 WFV satellite images in China, this study constructs a BP neural network model (BP1 model and BP2 model) based on different hidden layers, and compares and analyzes the accuracy of the inversion between the BP1 model, BP2 model and 6 statistical models (NDVI、RVI、DVI、EVI、SAVI、ARVI). Based on the measured data, BP1 model and BP2 model are used to map the LAI dynamic changes of summer maize. The results show that LAI has good correlation with 6 common statistical models, and the fitting degree of the NDVI exponential equation regression model is the best. The overallR 2 of BP neural network model is slightly smaller than the statistical model, while RMSE is less than the statistical model, and the errors with the measured value is smaller than the statistical model. So both the statistical model and the BP neural network model have advantages and disadvantages. The BP2 model is superior to the BP1 model inR 2 and RMSE, and can obtain more accurate inversion values, and the overall prediction accuracy of BP2 is higher. Based on the BP neural network simulation of summer maize growth period inversion, the LAI value presents a slow increase to the gradual decrease of S type change process, which is basically in line with the crop growth rules. The study combines with the BP neural network model established by different hidden layers to provide a method for the application of GF-1 satellite in the application of crop leaf area index multiple growth period inversion.

Key words: Summer maize    Leaf Area Index(LAI)    BP neural network model    Statistical model    Multiple growth period
收稿日期: 2018-08-23 出版日期: 2020-04-01
ZTFLH:  TP79  
基金资助: 四川省科技计划项目“基于大数据机器学习与冠层反射率模型结合的水稻叶面积指数提取技术”(2018JZ0054);高分辨率对地观测系统重大专项“GF?6卫星宽幅相机地表覆盖监测及地表覆盖变化快速检测技术”(30?Y20A07?9003?17/18);海南省重点研发计划项目“基于高分辨率数据的农业陆表环境关键参量遥感提取技术”(ZDYF2018231);政府间国际科技创新合作重点专项“基于红外遥感和电离层信息的地震监测预测技术研究”(2016YFE0122200)
通讯作者: 孟庆岩     E-mail: 310122664@qq.com;mengqy@radi.ac.cn
作者简介: 刘 俊(1993-),男,湖南衡阳人,硕士研究生,主要从事定量遥感以及农业遥感研究。E?mail:310122664@qq.com
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  
刘俊
孟庆岩
葛小三
刘顺喜
陈旭
孙云晓

引用本文:

刘俊,孟庆岩,葛小三,刘顺喜,陈旭,孙云晓. 基于BP神经网络的夏玉米多生育期叶面积指数反演研究[J]. 遥感技术与应用, 2020, 35(1): 174-184.

Jun Liu,Qingyan Meng,Xiaosan Ge,Shunxi Liu,Xu Chen,Yunxiao Sun. Leaf Area Index Inversion of Summer Maize at Multiple Growth Stages based on BP Neural Network. Remote Sensing Technology and Application, 2020, 35(1): 174-184.

链接本文:

http://www.rsta.ac.cn/CN/10.11873/j.issn.1004-0323.2020.1.0174        http://www.rsta.ac.cn/CN/Y2020/V35/I1/174

图1  研究区位置及采样点分布
图2  BP神经网络结构框图
图3  S型函数与双曲正切S型函数图像
图4  BP1和BP2 LAI反演分布图
图5  BP神经网络模型反演值与实测值拟合图
图6  相对误差大小正态分布图
图7  BP1和BP2反演值 - LAI实测值拟合图 (注:BP1为一层隐含层的BP神经网络模型,BP2为两层隐含层的BP神经模型;S1为训练样本点,S2为测试样本点,S3为所有样本点)
评价指标 训练样本 测试样本 所有样本
BP1 BP2 BP1 BP2 BP1 BP2
R 2 0.77 0.85 0.52 0.56 0.70 0.76
RMSE 0.23 0.23 0.34 0.31 0.27 0.26
表1  BP1和BP2 - LAI反演精度
植被指数 拟合公式 决定系数(R 2 均方根误差(RMSE)
NDVI y = 5.3675 x + 0.3623 0.78 0.42
y = 1.5534 e 2.1535 x 0.83 0.39
RVI y = 2.1433 x + 2.134 0.71 0.41
y = 0.6642 e 2.132 x 0.72 0.41
DVI y = 3.1933 x + 0.1656 0.68 0.51
y = 3.1654 e 1.1175 x 0.70 0.47
EVI y = 3.2124 x + 0.4045 0.72 0.47
y = 2.4041 e 1.5156 x 0.75 0.45
SAVI y = 0.536 x + 3.431 0.74 0.43
y = 3.534 e 0.3424 x 0.76 0.42
ARVI y = 0.6857 x + 2.837 0.74 0.45
y = 2.849 e 1.1175 x 0.76 0.42
表2  统计模型各植被指数反演精度
图8  BP1和BP2玉米6~9月反演结果
图9  BP1与BP2反演均值拟合图
1 Yang Min , Lin Jie , Gu Zheyan ,et al .Leaf Area Index Retrieval based on Landsat 8 OLI Multi-spectral Image Data and BP Neural Network[J].Science of Soil and Water Conservation,2015,13(4):86-93.杨敏,林杰,顾哲衍, 等 .基于Landsat 8 OLI多光谱影像数据和BP神经网络的叶面积指数反演[J].中国水土保持科学,2015,13(4):86-93.
2 Liang Shunlin , Cheng Jie , Jia Kun ,et al .Recent Progress in Land Surface Quantitative Remote Sensing[J].Journal of Remote Sensing,2016,20(5):875-898.梁顺林,程洁,贾坤, 等 .陆表定量遥感反演方法的发展新动态[J].遥感学报,2016,20(5):875-898.
3 Watson D J .Comparative Physiological Studies on the Growth of Field Crops: I. Variation in Net Assimilation Rate and Leaf Area between Species and Varieties, and within and between Years[J].Annals of Botany,1947,11(41):41-76.
4 Zhu Yunfang , Zhu Li , Li Jiaguo ,et al .The Study of Inversion of Chlorophyll A in Taihu based on GF-1 WFV Image and BP Neural Network[J].Acta Scientiae Circumstantiae,2017,37(1):130-13.朱云芳,朱利,李家国, 等 .基于GF-1WFV影像和BP神经网络的太湖叶绿素a反演[J].环境科学学报,2017,37(1):130-137.
5 Wei Xiangqin , Gu Xingfa , Meng Qingyan ,et al .Leaf Area Index Estimation Using Chinese GF-1 Wide Field View Data in an Agriculture Region[J].Sensors,2017,17(7):1593.
6 Zhang Ying , Meng Qingyan , Wu Jiali ,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.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.
7 Gao Shuai , Niu Zheng , Wu Mingquan .The Neural Network Algorithm for Estimating Plantation Forest Leaf Area Index based on ENVISAT/ASAR[J].Remote Sensing Technology and Application,2013,28(2):205-211.高帅, 牛铮, 邬明权.基于ENVISAT/ASAR的神经网络反演人工林叶面积指数研究[J].遥感技术与应用,2013,28(2):205-211.
8 Su Wei , Hou Ning , Li Qi ,et al .Retrieving Leaf Area Index of Corn Canopy based on Sentinel-2 Remote Sensing Image[J].Transactions of the Chinese Society for Agricultural Machinery,2018(1):151-156.苏伟,侯宁,李琪, 等 .基于Sentinel-2遥感影像的玉米冠层叶面积指数反演[J].农业机械学报,2018(1):151-156.
9 Jiang Chunlei , Zhang Shuqing , Zhang Ce ,et al .Modeling and Predicting of MODIS Leaf Area Index Time Series based on a Hybrid SARIMA and BP Neural Network Method[J],Spectroscopy and Spectral Analysis,2017,37(1):189-193.姜春雷,张树清,张策, 等 .基于SARIMA-BP神经网络组合方法的MODIS叶面积指数时间序列建模与预测[J].光谱学与光谱分析,2017,37(1):189-193.
10 Jacquemoud S , Bacour C , Poilvé H ,et al .Comparison of Four Radiative Transfer Models to Simulate Plant Canopies Reflectance : Direct and Inverse Mode[J].Remote Sensing of Environment,2000,74(3):471-481.
11 Liu Ke , Zhou Qingbo , Wu Wenbin ,et al .Comparison between Multispectral and Hyperspectral Remote Sensing for LAI Estimation[J].Transactions of the Chinese Society of Agricultural Engineering,2016,32(3):155-162.刘轲,周清波,吴文斌, 等 .基于多光谱与高光谱遥感数据的冬小麦叶面积指数反演比较[J].农业工程学报,2016,32(3):155-162.
12 Yun Zengxin , Zheng Guang , Ma Lixia ,et al .Evaluate Quantitatively Effect of Understory on Leaf Area Index (LAI) Estimation Combining Active and Passive Remote Sensing Data[J],Remote Sensing Technology and Application,2019,34(3):583-594.云增鑫,郑光,马丽霞, 等 .联合森林主被动遥感数据定量评价林下植被对叶面积指数估算的影响[J].遥感技术与应用,2019,34(3):583-594.
13 Gao Lin , Li Changchun , Wang Baoshan ,et al .Comparison of Precision in Retrieving Soybean Leaf Area Index based on Multi-source Remote Sensing Data[J].Chinese Journal of Applied Ecology,2016,27(1):191-200.高林,李长春,王宝山, 等 .基于多源遥感数据的大豆叶面积指数估测精度对比[J].应用生态学报,2016,27(1):191-200.
14 Jiang Xiaodong , Shi Jiaojiao , Li Yingxue ,et al .Analysis of Canopy Reflectance Spectra between Winter Wheat Cultivars with Different Gluten Strength[J].Journal of Nuclear Agricultural Sciences,2014,28(8):1503-1509.江晓东,石姣姣,李映雪, 等 .不同筋力冬小麦冠层反射光谱的差异分析[J].核农学报,2014,28(8):1503-1509.
15 Wang Qiang , Pang Yong , Li Zengyuan ,et al .Inversion of Leaf Area Index based on A Simple Physical Model[J].Journal of China University of Mining and Technology,2016,45(3):623-629.王强,庞勇,李增元, 等 .基于一种简单物理模型的叶面积指数反演[J].中国矿业大学学报,2016,45(3):623-629.
16 Wu Jiali , Gu Xingfa , Yu Tao ,et al .The Algorithm Research of LAI Inversion of HJ-1 CCD Camera Data base on SAIL Model[J].Microcomputer Information,2010,26(10):204-206.武佳丽,顾行发,余涛, 等 .基于SAIL模型的HJ-1卫星LAI反演算法研究[J].微计算机信息,2010,26(10):204-206.
17 Liu Yang , Liu Ronggao , Liu Siliang ,et al .Estimation of Crop LAI based on the Neural Network Trained from Physical Model[J].Journal of Geo-information Science,2010,12(3):426-435.刘洋,刘荣高,刘斯亮, 等 .基于物理模型训练神经网络的作物叶面积指数遥感反演研究[J].地球信息科学学报,2010,12(3):426-435.
18 Li Zongnan , Chen Zhongxin , Wang Limin .Inversion of LAI and Chlorophyll Content of Winter Wheat at Different Stages Using Two-layer Canopy Reflectance Model[J].Chinese Journal of Agricultural Resources and Regional Planning,2012,33(1):1-5.李宗南, 陈仲新, 王利民.基于ACRM模型不同时期冬小麦LAI和叶绿素反演研究[J].中国农业资源与区划,2012,33(1):1-5.
19 Chen Yanhua , Zhang Wanchang , Yong Bin .A Feasibility Stuudy of Leaf Area Index Inversion Using Radiative Transfer Model based on TM Data[J].Remote Sensing for Land & Resources,2007,19(2):44-49.陈艳华, 张万昌, 雍斌.基于TM的辐射传输模型反演叶面积指数可行性研究[J].国土资源遥感,2007,19(2):44-49.
20 Wang Liang , Lin Tao , Yan Changrong ,et al .Effects of Plastic Film Residue on Evapotranspiration and Soil Evaporation in Cotton Field of Xinjiang[J].Transactions of the Chinese Society of Agricultural Engineering,2016,32(14):120-128.王亮,林涛,严昌荣, 等 .地膜残留量对新疆棉田蒸散及棵间蒸发的影响[J].农业工程学报,2016,32(14):120-128.
21 Chen Yanhua , Zhang Wanchang , Yong Bin .Retrieving Leaf Area Index Using A Neural Network based on Classification Knowledge[J].Acta Ecologica Sinica,2007,27(7):2785-2793.陈艳华, 张万昌, 雍斌.基于分类知识利用神经网络反演叶面积指数[J].生态学报,2007,27(7):2785-2793.
22 Jia Jieqiong , Liu Wanqing , Meng Qingyan ,et al .Estimation of Maize Leaf Area Index based on GF-1 WFV Image and Machine Learning Random Algorithm[J].Journal of Image and Graphics,2018,23(5):719-729.贾洁琼,刘万青,孟庆岩, 等 .基于GF-1 WFV影像和机器学习算法的玉米叶面积指数估算[J].中国图象图形学报,2018,23(5):719-729.
23 Qi Jiaguo , Yann Kerr , Moran M S ,et al .Leaf Area Index Estimates Using Remotely Sensed Data and BRDF Models in A Semiarid Region[J].Remote Sensing of Environment,2000,73(1):18-30.
24 Ma Yinchi , Yan Guangjian , Ding Wen ,et al .Leaf Area Index Retrieval of Winter Wheat Using Artificial Neural Network[J].Transactions of the Chinese Society of Agricultural Engineering,2009,25(12):187-192.马茵驰,阎广建,丁文, 等 .基于人工神经网络方法的冬小麦叶面积指数反演[J].农业工程学报,2009,25(12):187-192.
25 Bacour C , Baret F , Béal D ,et al .Neural Network Estimation of LAI, fAPAR, fCover, and LAI × Cab, from Top of Canopy MERIS Reflectance Data: Principles and Validation[J].Remote Sensing of Environment,2006,105(4):313-325.
26 Noble P A , Tribou E H .Neuroet : An Easy-to-Use Artificial Neural Network for Ecological and Biological Modeling[J].Ecological Modelling,2007,203(1):87-98.
27 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):331-336.
28 Miao Naizhe .Accuracy Comparison of the Leaf Area Index Inversion Method of Winter Wheat at Different Growth Stages based on Landsat_TM Data[D].Xi'an:Xi'an University of Science and Technology,2012.苗乃哲. 基于Landsat_TM数据的冬小麦不同生育期叶面积指数反演方法精度比较[D].西安: 西安科技大学,2012.
29 He Jia , Liu Bingfeng , Li Jun .Monitoring Model of Leaf Area Index of Winter Wheat based on Hyperspectral Reflectance at Different Growth Stages[J].Transactions of the Chinese Society of Agricultural Engineering,2014,30(24):141-150.贺佳, 刘冰锋, 李军.不同生育时期冬小麦叶面积指数高光谱遥感监测模型[J].农业工程学报,2014,30(24):141-150.
30 Zhao Juan , Huang Wenjiang , Zhang Yaohong ,et al .Inversion of Leaf Area Index during Different Growth Stages in Winter Wheat[J].Spectroscopy and Spectral Analysis,2013,33(9):2546-2552.赵娟,黄文江,张耀鸿, 等 .冬小麦不同生育时期叶面积指数反演方法[J].光谱学与光谱分析,2013,33(9):2546-2552.
31 Gu Zheyan , Zhang Jinchi , Lin Jie ,et al .Effects of Atmospheric Correction on Remote Sensing Estimation of LAI of Broadleaved Forest[J].Journal of South China Agricultural University,2014(3):100-104.顾哲衍,张金池,林杰, 等 .大气校正模型对阔叶林叶面积指数遥感估算的影响[J].华南农业大学学报,2014(3):100-104.
32 Gu Zheyan , Zhang Jinchi , Lin Jie ,et al .Comparative Study of Different Atmospheric Correction Model on Estimation of Leaf Area Index Using Remote Sensing[J],Journal of Arid Land Resources and Environment,2014,28(7) :62-66.顾哲衍,张金池,林杰, 等 .不同大气校正模型对叶面积指数遥感反演精度的影响比较[J].干旱区资源与环境,2014,28(7) :62-66.
33 Rouse J W , Haas R H , Schell J A ,et al .Monitoring Vegetation Systems in the Great Plains with ERTS[C]∥Proceedings of Third Earth Resources Technology Satellite-1 Symposium.Greenbelt,1974 (351):310-317.
34 Rouse J W , Haas R H , Schell J A ,et al .Monitoring Vegetation Systems in the Great Plains with ERTS[C]∥Proceedings of Third Earth Resources Technology Satellite-1 Symposium.Greenbelt,1974 (351):310-317.
35 Huete A R .A Soil-adjusted Vegetation Index (SAVI).[J].Remote Sensing of Environment,1988,25(3):295-309.
36 Jordan C F .Derivation of Leaf Area Index from Quality of Light on the Forest Floor[J].Ecology,1969,50(4):663-666.
37 Kaufman Y J , Tanré D .Atmospherically Resistant Vegetation Index (ARVI) for EOS-MODIS[J].IEEE Transaction on Geoscience Remote Sensing,1992,30(2):261-270.
38 Liu H Q , Huete A .A Feedback based Modification of the NDVI to Minimize Canopy Background and Atmospheric Noise[J].IEEE Transactions on Geoscience and Remote Sensing,1995,33(2):457-465.
39 Li Huidong , Guan Dexin , Yuan Fenghui ,et al .Modeling Canopy Transpiration of Young Poplar Trees ( Populus × Euramericana cv. N3016 ) based on Back Propagation Artificial Neural Network[J].Acta Ecologica Sinica,2015,35(12):4137-4145.李辉东,关德新,袁凤辉, 等 .BP人工神经网络模拟杨树林冠蒸腾[J].生态学报,2015,35(12):4137-4145.
40 Guo Haitao , Zhang Dianlun , Ma Guofang .Some Problems to be Considered in Using BP Algorithm[J].Journal of Jiamusi University(Natural Science Edition),2000,18(4):363-365.郭海涛, 张殿伦, 马国芳.使用BP算法时应考虑的若干问题[J].佳木斯大学学报(自然科学版),2000,18(4):363-365.
[1] 徐卫星,薛华柱,靳华安,李爱农. 融合遥感先验信息的叶面积指数反演[J]. 遥感技术与应用, 2019, 34(6): 1235-1244.
[2] 程雪,贺炳彦,黄耀欢,孙志刚,李鼎,朱婉雪. 基于无人机高光谱数据的玉米叶面积指数估算[J]. 遥感技术与应用, 2019, 34(4): 775-784.
[3] 云增鑫, 郑光, 马利霞, 王晓菲, 卢晓曼, 路璐. 联合主被动遥感数据定量评价林下植被对叶面积指数估算的影响[J]. 遥感技术与应用, 2019, 34(3): 583-594.
[4] 刘洁, 李静, 柳钦火, 何彬彬, 于文涛. 全球典型植被叶片光谱特征及其对LAI反演的影响分析[J]. 遥感技术与应用, 2019, 34(1): 155-165.
[5] 刘振波,邹娴,葛云健,陈健,曹雨濛. 基于高分一号WFV影像的随机森林算法反演水稻LAI[J]. 遥感技术与应用, 2018, 33(3): 458-464.
[6] 谢京凯,王福民,王飞龙,张东尼. 面向水稻LAI监测的植被指数土壤调节参数修正[J]. 遥感技术与应用, 2018, 33(2): 342-350.
[7] 万玲,尤红建,程跃兵,卢晓军. 合成孔径雷达图像分割研究进展[J]. 遥感技术与应用, 2018, 33(1): 10-24.
[8] 褚洪亮,肖青,柏军华,程娟. 基于无人机遥感的叶面积指数反演[J]. 遥感技术与应用, 2017, 32(1): 140-148.
[9] 林岳峰,柳钦火,李静,赵静. 最小二乘法联合光学与雷达遥感数据估算玉米叶面积指数[J]. 遥感技术与应用, 2016, 31(4): 691-701.
[10] 杨勇帅,李爱农,靳华安,尹高飞,赵伟,雷光斌,边金虎. 中国西南山区GEOV1、GLASS和MODISLAI产品的对比分析[J]. 遥感技术与应用, 2016, 31(3): 438-450.
[11] 靳华安,李爱农,边金虎,赵伟,张正健,南希. 西南地区不同山地环境梯度叶面积指数遥感反演[J]. 遥感技术与应用, 2016, 31(1): 42-50.
[12] 李铮,柏延臣,何亚倩. 遥感叶面积指数产品提取自然植被物候期对比[J]. 遥感技术与应用, 2015, 30(6): 1103-1112.
[13] 刘婧怡,汤旭光,常守志,贾明明,董张玉,邵田田,丁智. 森林叶面积指数遥感反演模型构建及区域估算[J]. 遥感技术与应用, 2014, 29(1): 18-25.
[14] 孙晨曦,刘良云,关琳琳. 内蒙古锡林浩特草原GLASS LAI产品的真实性检验[J]. 遥感技术与应用, 2013, 28(6): 949-954.
[15] 赖格英,曾祥贵,刘影,张玲玲,易发钊,潘瑞鑫,盛盈盈. 基于ETM和图像融合的优势植被冠层叶面积指数和消光系数的遥感反演[J]. 遥感技术与应用, 2013, 28(4): 697-706.