Please wait a minute...
img

官方微信

遥感技术与应用  2020, Vol. 35 Issue (5): 1057-1069    DOI: 10.11873/j.issn.1004-0323.2020.5.1057
LAI专栏     
基于模拟退火算法的BP神经网络模型估算高分辨率叶面积指数
薛华柱1(),王昶景1,2,周红敏2(),王锦地2,万华伟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.北京师范大学地理科学学部,遥感科学国家重点实验室,北京市陆表遥感数据产品工程技术 研究中心,北京 100875
3.环境保护部卫星环境应用中心,北京 100094
BP Neural Network based on Simulated Annealing Algorithm for High Resolution LAI Retrieval
Huazhu Xue1(),Changjing Wang1,2,Hongmin Zhou2(),Jingdi Wang2,Huawei Wan3
1.School of Surveying & Land Information Engineering,Henan Polytechnic University,Jiaozuo 454000,China
2.State Key Laboratory of Remote Sensing Science,Beijing Engineering Research Center for Global Land Remote Sensing Products,Faculty of Geographical Science,BNU,Beijing 100875,China
3.Satellite Environmental Center,Ministry of Environmental Protection,Beijing 100785,China
 全文: PDF(5507 KB)   HTML
摘要:

卫星遥感技术的快速发展使得获取全球大范围叶面积指数成为可能,但基于现有的算法和数据估算高分辨率LAI的精度还需要提高。针对农作物、草地和林地等3种典型地表类型,选取地面观测数据较多的4个研究区,包括3个各地类用于建模验证的研究区与一个用于适用性验证的独立研究区,针对4个研究区,分别获取地面测量数据以及对应的30 m空间分辨率地表反射率数据。在3个主要研究区建立并比较了NDVI植被指数经验模型、BP神经网络模型和基于模拟退火算法的BP神经网络模型,利用地面实测数据对模型进行验证。结果表明:在研究所选的3个主要研究区,基于模拟退火算法的BP神经网络模型的估算精度比BP神经网络模型和NDVI经验模型的估算精度高,农田、草地和林地站点估算结果的决定系数分别为0.899、0.858和0.863,BP神经网络模型的估算结果决定系数分别为:0.763、0.710和0.742,NDVI经验模型的精度最差,其估算结果的决定系数分别为0.622、0.536和0.637。为了验证SA-BP神经网络的适用性,选取独立研究区进行验证,结果显示验证精度较高,R2为0.842,RMSE为0.689 5,说明该模型外推能力较好。研究证明了基于模拟退火算法的BP神经网络模型提高了模型泛化能力,有效防止了BP神经网络模型滑入局部最小值,是提高高空间分辨率LAI估算精度的有效手段。

关键词: 叶面积指数反演模拟退火算法神经网络    
Abstract:

The rapid development of satellite remote sensing technology makes it possible to obtain global large-scale Leaf Area Index (LAI). However, it is difficult to estimate high-resolution LAI based on existing algorithms and data. In this paper, four typical research areas consist of three research areas for modeling validation and an independent research area for applicability validation including grassland, farmland and woodland were selected. Field data and correspondent satellite of the areas were then collected. The empirical model of NDVI vegetation index, the BP neural network model and the BP neural network based on simulated annealing algorithm were established and 30 m resolution LAI data were estimated with all models. Estimated results were validated with the Field data in three main research areas. The results indicated that NDVI empirical model has the worst accuracy in the three main research areas selected in this paper. The estimation accuracy of BP neural network model based on simulated annealing algorithm is higher than that of BP neural network model. The determinant coefficients of estimation results of farmland, grassland and woodland sites are 0.899, 0.858 and 0.863 respectively. The determinant coefficients of BP neural network model were 0.763, 0.710 and 0.742 respectively, while the determinant coefficients of NDVI empirical model were 0.622, 0.536 and 0.637 respectively. In order to verify the applicability of SA-BP neural network, an independent research area was selected for further verification. The results show that the validation accuracy is high,R2 is 0.842 0, and RMSE is 0.689 5, which shows that the model has good extrapolation ability. This study proves that the BP neural network model based on simulated annealing algorithm improves the generalization ability of the model, effectively prevents the BP neural network model from sliding into the local minimum, and it is an effective method for LAI estimation in high spatial resolution.

Key words: Leaf Area Index(LAI)    Inversion    Simulated annealing algorithm    BP neural network
收稿日期: 2019-03-21 出版日期: 2020-11-26
ZTFLH:  TP79  
基金资助: 国家重点研发计划课题“全球多时空尺度遥感动态监测与模拟预测”(2016YFB0501502);国家自然科学青年基金项目(41801242);国家重点基础研究发展计划项目“遥感信息动态特征分析与时间尺度扩展”(2013CB733403);国家自然科学基金项目“小麦长势无人机遥感监测关键指标参数反演机理及长势诊断模型研究”(41871333)
通讯作者: 周红敏     E-mail: xhz@hpu.edu.cn;zhouhm@bnu.edu.cn
作者简介: 薛华柱(1977-),男,安徽肥东人,博士,副教授,主要从事地表参数定量反演研究。E?mail:xhz@hpu.edu.cn
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  
薛华柱
王昶景
周红敏
王锦地
万华伟

引用本文:

薛华柱,王昶景,周红敏,王锦地,万华伟. 基于模拟退火算法的BP神经网络模型估算高分辨率叶面积指数[J]. 遥感技术与应用, 2020, 35(5): 1057-1069.

Huazhu Xue,Changjing Wang,Hongmin Zhou,Jingdi Wang,Huawei Wan. BP Neural Network based on Simulated Annealing Algorithm for High Resolution LAI Retrieval. Remote Sensing Technology and Application, 2020, 35(5): 1057-1069.

链接本文:

http://www.rsta.ac.cn/CN/10.11873/j.issn.1004-0323.2020.5.1057        http://www.rsta.ac.cn/CN/Y2020/V35/I5/1057

研究区域建模数据对验证数据对
张北407
Pshenichne14125
Nezer9116
表 1  建模数据量一览表
图 1  乌克兰研究区土地覆盖分布图(2010年)
图 2  BP神经网络模型结构示意图
图3  SA-BP算法流程图
图4  BP神经网络迭代曲线
图5  SA-BP神经网络迭代曲线
研究区域模型类型R2RMSE
PshenichneNDVI经验模型0.58620.9981
BP神经网络0.78360.7331
SA-BP神经网络0.86250.5822
ZhangbeiNDVI经验模型0.53630.5848
BP神经网络0.70530.5171
SA-BP神经网络0.89620.3484
NezerNDVI经验模型0.53170.9141
BP神经网络0.75380.7923
SA-BP神经网络0.87690.6530
表 2  模型精度一览表
图6  Pshenichne研究区验证结果
图7  Zhangbei研究区验证结果
图8  Nezer研究区验证结果
图9  Pshenichne SA-BP 神经网络模型估算结果
图10  Zhangbei SA-BP 神经网络模型估算结果
图 11  Nezer SA-BP 神经网络模型估算结果
图12  Pshenichne 2014年6月6日区域LAI估算结果对比统计图
图13  Southwest研究区验证结果
1 Chen J M, Black T A. Defining Leaf Area Index for Non-Flat Leaves[J]. Plant Cell & Environment, 1992, 15(4):421-429.
2 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.
3 Gu Chengyan, Du Huaqiang, Zhou Guomo, 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.
3 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.
4 Cleugh H A, Leuning R, Mu Q, et al. Regional Evaporation Estimates from Flux Tower and MODIS Satellite Data[J]. Remote Sensing of Environment, 2007, 106(3):285-304.
5 Hu Lei, Zhang Qingxia, Zhang Chonghua, et al. Changes in Leaf Area Index of Early-Season Cropping Rice and Its Relationship with Yield under Climate Warming in Nanchang County [J]. Ecology and Environmental Sciences, 2019, 28(9): 1785-1792.
5 胡磊, 张清霞, 张崇华, 等. 气候变暖下南昌县早稻叶面积指数变化及其与产量的关系[J]. 生态环境学报, 2019, 28(9): 1785-1792.
6 Turner D P, Cohen W B, Kennedy R E, et al. Relationships between Leaf Area Index and Landsat TM Spectral Vegetation Indices across Three Temperate Zone Sites[J]. Remote Sensing of Environment, 1999, 70(1):52-68.
7 Liu Jun, Pang Xin, Li Yanrong, et al. Inversion Study on Leaf Area Index of Summer Maize Using Remote Sensing[J]. Transactions of the Chinese Society for Agricultural Machinery, 2016, 47(9):309-317.
7 刘珺, 庞鑫, 李彦荣, 等. 夏玉米叶面积指数遥感反演研究[J]. 农业机械学报, 2016, 47(9):309-317.
8 Li Xiaotong, Qin Xianlin, Liu Shuchao, et al. Estimation of Forest Leaf Area Index based on GF-1 WFV Data[J]. Remote Sensing for Land & Resources, 2019, 31(3):80-86.
8 李晓彤, 覃先林, 刘树超, 等. 基于GF-1 WFV数据森林叶面积指数估算[J]. 国土资源遥感, 2019, 31(3):80-86.
9 Xu Weixing, Xue Huazhu, Jin Huaan, et al. Retrieval of Leaf Area Index by Fusing Prior Information from Remote Sensing Data[J]. Remote Sensing Technology and Application, 2019, 34(6):1235-1244.
9 徐卫星, 薛华柱, 靳华安, 等. 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.
10 Tian Qingjiu, Min Xiangjun. Advances in Study on Vegetation Indices[J]. Advance in Earthences, 1998, 13(4):327-333.
10 田庆久, 闵祥军. 植被指数研究进展[J]. 地球科学进展, 1998, 13(4):327-333.
11 Liang L, Di L, Zhang L, et al. Estimation of Crop LAI Using Hyperspectral Vegetation Indices and A Hybrid Inversion Method[J]. Remote Sensing of Environment, 2015, 165:123-134.
12 Jiang Hong, Zhang Zhaoming, Wang Xiaoqin, et al. Bamboo Forest LAI Retrieval and Analysis in Mountainous Area based on TAVI[J]. Journal of Geo-Information Science, 2015, 17(4):500-504.
12 江洪, 张兆明, 汪小钦, 等. 基于TAVI的山区毛竹林LAI反演分析[J]. 地球信息科学学报, 2015, 17(4):500-504.
13 Xiang Hongbo, Guo Zhihua, Zhao Zhanqing, et al. Estimating Method of Forest Leaf Area Index on Different Space Scales[J]. Scientia Silvae Sinicae, 2009, 45(6):139-144.
13 向洪波, 郭志华, 赵占轻, 等. 不同空间尺度森林叶面积指数的估算方法[J]. 林业科学, 2009, 45(6):139-144.
14 Wang Dongwei, Meng Xianzhi, Wang Jingdi, et al. The Devel-opment of Leaf Area Index Inversion from Remote Sensing[J]. Journal of Wuyi University, 2009, 23(4):47-52.
14 王东伟, 孟宪智, 王锦地, 等. 叶面积指数遥感反演方法进展[J]. 五邑大学学报(自然科学版), 2009, 23(4):47-52.
15 Ren Zhe, Chen Huailiang, Wang Lianxi, et al. Research on Inversion Model of Wheat LAI Using Cross-Validation[J]. Remote Sensing for Land and Resources, 2015, 27(4):34-40.
15 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.
16 Liang S, Zhang X, Xiao Z, et al. Global LAnd Surface Satellite (GLASS) Products[M]. Berlin: Springer International Publishing, 2014.
17 Knyazikhin Y, Martonchik J V, Myneni R B, et al. Estimating Vegetation Canopy Leaf Area Index and Fraction of Absorbed Photosynthetically Active Radiation from MODIS and MISR Data[J]. Journal of Geophysical Research, 1998, 103(D24): 32239-32256.
18 Baret F, Hagolle O, Geiger B, et al. LAI, FAPAR and FCover CYCLOPES Global Products Derived from VEGET-ATION: Part 1: Principles of the Algorithm[J]. Remote Sensing of Environment, 2007, 110(3):275-286.
19 Martin C, Jessica M, Eric V, et al. A 30+ Year AVHRR LAI and FAPAR Climate Data Record: Algorithm Description and Validation[J]. Remote Sensing, 2016, 8(3):263. doi: 10.3390/rs8030263.
doi: 10.3390/rs8030263
20 Zhang Y, Tian Y, Knyazikhin Y, et al. Prototyping of MISR LAI and FPAR Algorithm with POLDER Data over Africa[J]. IEEE Transactions on Geoence and Remote Sensing, 2000, 38(5): 2402-2418.
21 Baret F, Weiss M, Lacaze R, et al. GEOV1: LAI and FAPAR Essential Climate Variables and FCOVER Global Time Series Capitalizing Over Existing Products. Part1: Principles of Development and Production[J]. Remote Sensing of Environ-ment, 2013, 137(10):299-309.
22 Zhao Jing, Li Jing, Liu Qinhuo, et al. Leaf Area Index Inversion Combining with HJ-1/CCD and Landsat 8/OLI Data in the Middle Reach of the Heihe River Basin[J]. Journal of Remote Sensing, 2015, 19(5):733-749.
22 赵静, 李静, 柳钦火, 等. 联合HJ-1/CCD和Landsat 8/OLI数据反演黑河中游叶面积指数[J]. 遥感学报, 2015, 19(5):733-749.
23 Xiao Z, Liang S, Wang J, et al. Use of General Regression Neural Networks for Generating the GLASS Leaf Area Index Product from Time-Series MODIS Surface Reflectance[J]. IEEE Transactions on Geoscience and Remote Sensing, 2014, 52(1):209-223.
24 Shi Y, Wang J, Wang J, et al. A Prior Knowledge-based Method to Derivate High-resolution Leaf Area Index Maps with Limited Field Measurements[J]. Remote Sensing, 2016, 9(1):13. doi: 10.3390/rs9010013.
doi: 10.3390/rs9010013
25 Li Shuangcheng, Zheng Du. Applications of Artificial Neural Networks to Geoscien-Ces:Review and Prospect[J]. Advances in Earth Science, 2003, 18(1):68-76.
25 李双成, 郑度. 人工神经网络模型在地学研究中的应用进展[J]. 地球科学进展, 2003, 18(1):68-76.
26 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.
27 Xiang Hongbo. Estimating Forest LeafArea Index based on BP-Nueral Networks: A Case Study from Northern Slop of Changbai Moutain[D]. Chongqing:Southwest University, 2009.向洪波.基于BP神经网络森林叶面积指数估算研究—以长白山北坡为例[D].重庆:西南大学,2009.
28 Liu Jun, Meng Qingyan, Ge Xiaosan, et al. Leaf Area Index Inversion of Summer Maize at Multiple Growth Stages based on BP Neural Network[J]. Remote Sensing Technology and Application, 2020, 35(1):174-184.
28 刘俊, 孟庆岩, 葛小三, 等. 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.
29 Ma Yinchi, Yan Guangjian, Ding Wen, et al. Leaf Area Index Retrieval of Winter Wheat Using Artificial Neural Network[J]. Transactions of the CSAE, 2009, 25(12):187-192.
29 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.
30 Ding S, Su C, Yu J. An Optimizing BP Neural Network Algori-thm based on Genetic Algorithm[J]. Artificial Intelligence Review, 2011, 36(2):153-162.
31 Lin Jie, Pan Ying, Yang Min, et al.Quantitative Inversion of Long Sequential Leaf Area Index Using Remote Sensing based on BP Neural Network from 1988 to 2013. Acta Eeologiea Sinica, 2018, 38(10):3534-3542.[林杰, 潘颖, 杨敏,等. 1988-2013年基于BP神经网络的植被叶面积指数遥感定量反演[J]. 生态学报, 2018, 38(10):3534-3542.]
32 Lu Kun, Meng Qingyan, Sun Yunxiao, et al. Estimating Leaf Area Index of Wheat at The Booting Stage Using GF-2 Data: A Case Study of Langfang City, Hebei Province[J]. Remote Sensing for Land & Resources, 2018, 30(1):196-202.
32 陆坤, 孟庆岩, 孙云晓, 等. 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.
33 Kirkpatrick S,Gelatt C D,Vecchi M P. Optimization by Simu-lated Annealing[J].Science,1983,220(4598):671-680.
34 Cui YongLi, Wei Jun. Application of SA Algorithm to Solving Problem of Non-linear Constraint Optimization[J]. Journal of Dalian University, 2003, 24(6):5-8.
34 崔永利, 魏君. SA法在求解非线性约束优化问题中的应用[J]. 大连大学学报, 2003, 24(6):5-8.
35 Lu Yanshan, Zhao Zhengqi. Optimization and Application Research of BP Neural Network[J]. Journal of Beijing University of Chemical Technology(Natural Science Edition), 2001, (1):67-69.吕砚山, 赵正琦. BP神经网络的优化及应用研究[J]. 北京化工大学学报(自然科学版), 2001, 28(1):67-69.
36 Liu Songqing. Research and Application on MATLAB BP Neural Network[J]. Computer Engineering and Design, 2003, 24(11):81-83,88.
36 柳松青. MATLAB神经网络BP网络研究与应用[J]. 计算机工程与设计, 2003, 24(11):81-83, 88.
37 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.
37 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.
38 Zhao X, Huang S, Xiang J, et al. Remote Sensing of Atmospheric Duct Parameters Using Simulated Annealing[J]. Chinese Physics B, 2011, 20(9):468-475.
39 Goffe W L, Ferrier G D, Rogers J. Global Optimization of Statistical Functions with Simulated Annealing[J]. Journal of Econometrics, 1994, 60(1-2):65-99.
40 Qi J, Chehbouni A, Huete A R, et al. A Modified Soil Adjusted Vegetation Index[J]. Remote Sensing of Environment, 1994, 48(2):119-126.
41 Zhu Xuchao, Yuan Guofu, Yi Xiaobo, et al. Leaf Area Index Inversion of Riparian Forest in the Lower basin of Tarim River based on Landsat 8 OLI Images[J]. Arid Land Geography, 2014, 37(6):1248-1256.
41 朱绪超, 袁国富, 易小波, 等. 基于Landsat 8 OLI影像的塔里木河下游河岸林叶面积指数反演[J]. 干旱区地理, 2014, 37(6):1248-1256.
[1] 汪垚,方红亮,张英慧,李思佳. 基于机载LVIS和星载GLAS波形LiDAR数据反演森林LAI[J]. 遥感技术与应用, 2020, 35(5): 1004-1014.
[2] 侯吉宇,周艳莲,刘洋. 不同叶面积指数遥感数据模拟中国总初级生产力的时空差异[J]. 遥感技术与应用, 2020, 35(5): 1015-1027.
[3] 桑宇星,刘刚,江聪,任舒艳,朱再春. 近30 a中国叶面积指数变化趋势的不确定性评估[J]. 遥感技术与应用, 2020, 35(5): 1028-1036.
[4] 刘刚,桑宇星,赵茜,江聪,朱再春. 生态系统模型模拟中国叶面积指数变化趋势及驱动因子的不确定性[J]. 遥感技术与应用, 2020, 35(5): 1037-1046.
[5] 郭利彪,刘桂香,运向军,张勇,孙世贤. 基于数据机理的植被叶面积指数遥感反演研究[J]. 遥感技术与应用, 2020, 35(5): 1047-1056.
[6] 李净,温松楠. 基于3种机器学习法的太阳辐射模拟研究[J]. 遥感技术与应用, 2020, 35(3): 615-622.
[7] 张峰极,吴艳兰,姚雪东,梁泽毓. 基于改进DenseNet网络的多源遥感影像露天开采区智能提取方法[J]. 遥感技术与应用, 2020, 35(3): 673-684.
[8] 程娟,肖青,闻建光,唐勇,游冬琴,卞尊健,吴胜标,郝大磊,钟守熠. 地物波谱数据库应用方法及遥感应用现状[J]. 遥感技术与应用, 2020, 35(2): 267-286.
[9] 王艺晴,韩震,周玮辰,吴义生. SMAP卫星海表面亮温仿真及海表面盐度遥感反演[J]. 遥感技术与应用, 2020, 35(2): 365-371.
[10] 郑贵洲,熊良超,廖艳雯,王红平. 利用MODIS数据反演南海南部海表温度及时空变化分析[J]. 遥感技术与应用, 2020, 35(1): 132-140.
[11] 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.
[12] 郑宗生,胡晨雨,黄冬梅,邹国良,刘兆荣,宋巍. 基于迁移学习及气象卫星云图的台风等级分类研究[J]. 遥感技术与应用, 2020, 35(1): 202-210.
[13] 边玲玲,王卷乐,郭兵,程凯,魏海硕. 基于特征空间的黄河三角洲垦利县土壤盐分遥感提取[J]. 遥感技术与应用, 2020, 35(1): 211-218.
[14] 胡路,赵天杰,施建成,李尚楠,樊东,王平凯,耿德源,肖青,崔倩,陈德清. 基于地基微波辐射观测的土壤水分反演算法评估[J]. 遥感技术与应用, 2020, 35(1): 74-84.
[15] 劳从坤,杨娜,徐少博,汤燕杰,张恒杰. 反演策略对SMOS土壤水分反演算法的影响研究[J]. 遥感技术与应用, 2020, 35(1): 65-73.