基于无人机高光谱融合连续投影算法估算棉花地上部生物量

doi:10.11963/1002-7807.yxzlf.20210428

摘要/Abstract

摘要：

【目的】地上部生物量是表征植物生命活动的重要参数。探索不同的光谱预处理方法和建模方法,实现对棉花地上部生物量快速、无损、准确的估算,对棉花长势监测和大田精准管理具有重要意义。【方法】以新陆早53号、新陆早45号为研究对象,设置不同施氮处理,于出苗后不同阶段获取棉花地上部生物量和无人机高光谱数据,通过连续投影算法（Successive projections algorithm,SPA）筛选不同预处理[一阶导数、二阶导数、Savitzky-Golay（SG平滑）、多元散射校正]后的特征波长,基于筛选出的不同波长组合使用偏最小二乘法回归（Partial least square regression,PLSR）和随机森林回归（Random forest regression,RFR）分别构建棉花地上部生物量估算模型,比较不同预处理后建立模型的精度,确定最优估算模型。【结果】（1）利用SPA算法对不同预处理后的光谱信息筛选出特征波长9~26个,可实现光谱信息降维。（2）基于SG平滑-SPA处理及PLSR方法建立的模型最佳,R²达到了0.63,均方根误差（Root mean square error,RMSE）为0.42,验证集的R²为0.67,RMSE为0.44。（3）一阶导数-SPA处理后,采用RFR构建的模型最佳,R²达到0.87,RMSE为0.45,验证集R²为0.81,RMSE为0.37。【结论】采用一阶导数预处理结合SPA筛选特征波长,经RFR构建的估算模型结果和验证效果均最佳,可用于棉花地上部生物量定量估算。

关键词: 棉花; 地上部生物量; 光谱预处理; 特征波长; 偏最小二乘法回归; 随机森林回归

Abstract:

[Objective] Above-ground biomass is an important parameter of plant life activity. Exploring different spectral pretreatment methods and modeling methods to achieve rapid, nondestructive and accurate estimation of cotton above-ground biomass is of great significance for cotton growth monitoring and field precision management. [Method] Xinluzao 53 and Xinluzao 45 were selected as the research objects, and different nitrogen application treatments were set up to obtain cotton above-ground biomass and UAV hyperspectral data at different stages after emergence. The successive projections algorithm (SPA) was used to select the characteristic wavelengths after different pretreatments [first derivative, second derivative, Savitzky-Golay (SG smoothing), multiple scatter correction]. Based on the selected wavelengths, partial least squares regression (PLSR) and random forest regression (RFR) were used to construct cotton above-ground biomass (dry matter) estimation models, respectively, and the optimal estimation model was determined by comparison of the model estimation accuracy. [Result] (1) The SPA algorithm can effectively remove the redundant bands from the spectra, and the number of above-ground biomass-sensitive characteristic wavelengths after different pretreatments ranges from 9 to 26, which could be used to reduce the dimension of spectral information. (2) Based on SG smoothing-SPA processing, the model established by PLSR is the best, with a coefficient of determination (R²) of 0.63, root mean square error (RMSE) of 0.42, and validation set R², RMSE of 0.67, 0.44. (3) After the first derivative-SPA processing, the model constructed by RFR is the best, with R² of 0.87, RMSE of 0.45, validation set R² and RMSE of 0.81 and 0.37. [Conclusion] Using the first derivative pretreatment combined with SPA to select biomass-sensitive wavelengths, the RFR model has the best results and validation performance, which can be used for quantitative estimation of cotton above-ground biomass.

Key words: cotton; above-ground biomass; spectral pretreatment; sensitive bands; partial least squares; random forest regression

易翔,张立福,吕新,张泽,田敏,印彩霞,马怡茹,范向龙. 基于无人机高光谱融合连续投影算法估算棉花地上部生物量[J]. 棉花学报, 2021, 33(3): 224-234.

Yi Xiang,Zhang Lifu,Lü Xin,Zhang Ze,Tian Min,Yin Caixia,Ma Yiru,Fan Xianglong. Estimation of cotton above-ground biomass based on unmanned aerial vehicle hyperspectral and successive projections algorithm[J]. Cotton Science, 2021, 33(3): 224-234.

0
/ / 推荐

导出引用管理器 EndNote|Reference Manager|ProCite|BibTeX|RefWorks

链接本文: http://journal.cricaas.com.cn/Jweb_mhxb/CN/10.11963/1002-7807.yxzlf.20210428

http://journal.cricaas.com.cn/Jweb_mhxb/CN/Y2021/V33/I3/224

图/表 11

图1

图2

表1

图3

图4

图5

表2

表3

图6

图7

表4

参考文献 30

[1]	李继福, 何俊峰, 陈佛文,等. 中国棉花生产格局与施肥研究现状——基于CNKI数据计量分析[J/OL]. 中国棉花, 2019, 46(4): 17-24, 28 [2020-12-25]. https://doi.org/10.11963/1000-632X.ljfljf.20190402. doi: https://doi.org/10.11963/1000-632X.ljfljf.20190402
	Li Jifu, He Junfeng, Chen Fowen, et al. Status of cotton planting and fertilization research in China——Based on CNKI data analysis[J/OL]. China Cotton, 2019, 46(4): 17-24, 28[2020-12-25]. https://doi.org/10.11963/1000-632X.ljfljf.20190402. doi: https://doi.org/10.11963/1000-632X.ljfljf.20190402
[2]	卫小华, 张立杰. 农业供给侧改革背景下新疆棉花产量变动及其驱动因素分析[J/OL]. 江苏农业科学, 2019, 47(14): 338-342 [2020-12-25]. https://doi.org/10.15889/j.issn.1002-1302.2019.14.076. doi: https://doi.org/10.15889/j.issn.1002-1302.2019.14.076
	Wei Xiaohua, Zhang Lijie. Analysis of cotton yield change in Xinjiang area and its driving factors under agricultural supply-side reform[J/OL]. Jiangsu Agricultural Sciences, 2019, 47(14): 338-342 [2020-12-25]. https://doi.org/10.15889/j.issn.1002-1302.2019.14.076. doi: https://doi.org/10.15889/j.issn.1002-1302.2019.14.076
[3]	刘畅, 杨贵军, 李振海,等. 融合无人机光谱信息与纹理信息的冬小麦生物量估测[J/OL]. 中国农业科学, 2018, 51(16): 3060-3073 [2020-12-25]. https://doi.org/10.3864/j.issn.0578-1752.2018.16.003A. doi: https://doi.org/10.3864/j.issn.0578-1752.2018.16.003A
	Liu Chang, Yang Guijun, Li Zhenhai, et al. Biomass estimation in winter wheat by UAV spectral information and texture information fusion[J/OL]. Scientia Agricultura Sinica, 2018, 51(16): 3060-73[2020-12-25]. https://doi.org/10.3864/j.issn.0578-1752.2018.16.003. doi: https://doi.org/10.3864/j.issn.0578-1752.2018.16.003A
[4]	范云豹, 宫兆宁, 赵文吉,等. 基于高光谱遥感的植被生物量反演方法研究[J/OL]. 河北师范大学学报(自然科学版), 2016, 40(3): 267-271[2020-12-25]. https://doi.org/10.13763/j.cnki.jhebnu.nse.2016.03.015. doi: https://doi.org/10.13763/j.cnki.jhebnu.nse.2016.03.015
	Fan Yunbao, Gong Zhaoning, Zhao Wenji, et al. study on vegetation biomass inversion method based on hyperspectral remote sensing[J/OL]. Journal of Hebei Normal University (Natural Science), 2016, 40(3): 267-271 [2020-12-25]. https://doi.org/10.13763/j.cnki.jhebnu.nse.2016.03.015. doi: https://doi.org/10.13763/j.cnki.jhebnu.nse.2016.03.015
[5]	王利民, 刘佳, 杨玲波,等. 基于无人机影像的农情遥感监测应用[J/OL]. 农业工程学报, 2013, 29(18): 136-145 [2020-12-25]. https://doi.org/10.3969/j.issn.1002-6819.2013.18.017. doi: https://doi.org/10.3969/j.issn.1002-6819.2013.18.017
	Wang Liming, Liu Jia, Yang Lingbo, et al. Applications of unmanned aerial vehicle images on agricultural remote sensing monitoring[J/OL]. Transactions of the Chinese Society of Agricultural Engineering, 2013, 29(18): 136-145 [2020-12-25]. https://doi.org/10.3969/j.issn.1002-6819.2013.18.017. doi: https://doi.org/10.3969/j.issn.1002-6819.2013.18.017
[6]	Yue J B, Feng H K, Yang G J, et al. A comparison of regression techniques for estimation of above-ground winter wheat biomass using near-surface spectroscopy[J/OL]. Remote Sensing, 2018, 10(2): 66. https://doi.org/10.3390/rs10010066. doi: https://doi.org/10.3390/rs10010066
[7]	李冰, 刘镕源, 刘素红,等. 基于低空无人机遥感的冬小麦覆盖度变化监测[J/OL]. 农业工程学报, 2012, 28(13): 160-165 [2020-12-25]. https://doi.org/1002-6819 (2012)-13-0160-06. doi: https://doi.org/1002-6819 (2012)-13-0160-06
	Li Bing, Liu Rongyuan, Liu Suhong, et al. Monitoring vegetation coverage variation of winter wheat by low-altitude UAV remote sensing system[J/OL]. Transactions of the Chinese Society of Agricultural Engineering, 2012, 28(13): 160-165[2020-12-25]. https://doi.org/1002-6819(2012)-13-0160-06. doi: https://doi.org/1002-6819 (2012)-13-0160-06
[8]	颜安, 郭涛, 陈全家,等. 基于无人机影像的棉花株高预测[J/OL]. 新疆农业科学, 2020, 57(8): 1493-1502[2020-12-25]. https://doi.org/10.6048/j.issn.1001-4330.2020.08.014. doi: https://doi.org/10.6048/j.issn.1001-4330.2020.08.014
	An Yan, Guo Tao, Chen Quanjia, et al. Prediction of cotton plant height based on UAV image[J]. Xinjiang Agricultural Sciences, 2020, 57(8): 1493-1502[2020-12-25]. https://doi.org/10.6048/j.issn.1001-4330.2020.08.014. doi: https://doi.org/10.6048/j.issn.1001-4330.2020.08.014
[9]	陈鹏飞, 梁飞. 基于低空无人机影像光谱和纹理特征的棉花氮素营养诊断研究[J/OL]. 中国农业科学, 2019, 52(13): 2220-2229 [2020-12-25]. https://doi.org/10.3864/j.issn.0578-1752.2019.13.003. doi: https://doi.org/10.3864/j.issn.0578-1752.2019.13.003
	Cheng Pengfei, Liang Fei. Cotton nitrogen nutrition diagnosis based on spectrum and texture feature of images from low altitude unmanned aerial vehicle[J/OL]. Scientia Agricultura Sinica, 2019, 52(13): 2220-2229[2020-12-25]. https://doi.org/10.3864/j.issn.0578-1752.2019.13.003. doi: https://doi.org/10.3864/j.issn.0578-1752.2019.13.003
[10]	崔美娜. 基于无人机遥感的棉花螨害动态监测研究[D]. 石河子: 石河子大学, 2019.
	Cui Meina. Study on dynamic monitoring of cotton spider mites based on remote sensing of UAV[D]. Shihezi: Shihezi University, 2019.
[11]	Tao H, Feng H, Xu L, et al. Estimation of crop growth parameters using UAV-based hyperspectral remote sensing data[J/OL]. Sensors, 2020, 20(5): 1296[2020-12-25]. https://doi.org/10.3390/s20051296. doi: https://doi.org/10.3390/s20051296
[12]	Hansen P M, Schjoerring J K. Reflectance measurement of canopy biomass and nitrogen status in wheat crops using normalized difference vegetation indices and partial least squares regression[J/OL]. Remote Sensing of Environment, 2003, 86(4): 542-553 [2020-12-25]. https://doi.org/10.1016/S0034-4257(03)00131-7. doi: https://doi.org/10.1016/S0034-4257(03)00131-7
[13]	邓江, 谷海斌, 王泽,等. 基于无人机遥感的棉花主要生育时期地上生物量估算及验证[J/OL]. 干旱地区农业研究, 2019, 37(5): 55-61, 69[2020-12-25]. https://doi.org/10.7606/j.issn.1000-7601.2019.05.09. doi: https://doi.org/10.7606/j.issn.1000-7601.2019.05.09
	Deng Jiang, Gu Haibin, Wang Ze, et al. Estimation and validation of above-ground biomass of cotton during main growth period using unmanned aerial vehicle (UAV)[J/OL]. Agricultural Research in the Arid Areas, 2019, 37(5): 55-61, 69 [2020-12-25]. https://doi.org/10.7606/j.issn.1000-7601.2019.05.09. doi: https://doi.org/10.7606/j.issn.1000-7601.2019.05.09
[14]	刘明博, 唐延林, 李晓利,等. 水稻叶片氮含量光谱监测中使用连续投影算法的可行性[J/OL]. 红外与激光工程, 2014, 43(4): 1265-1271.
	Liu Mingbo, Tang Yanlin, Li Xiaoli, et al. Feasibility of using successive projections algorithm in spectral monitoring of rice leaves nitrogen contents[J/OL]. Infrared and Laser Engineering, 2014, 43(4): 1265-1271.
[15]	Lei T, Lin X H, Sun D W. Rapid classification of commercial Cheddar cheeses from different brands using PLSDA, LDA and SPA-LDA models built by hyperspectral data[J/OL]. Journal of Food Measurement and Characterization, 2019, 13(4): 3119-3129 [2020-12-25]. https://doi.org/10.1007/s11694-019-00234-0. doi: https://doi.org/10.1007/s11694-019-00234-0
[16]	Geladi P, Kowalski B R. Partial least-squares regression: a tutorial[J]. Analytica Chimica Acta, 1986, 185: 1-17 doi: 10.1016/0003-2670(86)80028-9
[17]	Gleason C J, IM J. Forest biomass estimation from airborne LiDAR data using machine learning approaches[J/OL]. Remote Sensing of Environment, 2012, 125: 80-91[2020-12-25]. https://doi.org/10.1016/j.rse.2012.07.006. doi: https://doi.org/10.1016/j.rse.2012.07.006
[18]	Van Huffel S. Proceedings of the second international workshop on Recent advances in total least squares techniques and errors-in-variables modeling[C]//International Workshop on Recent Advances in Total Least Squares Techniques & Errors-in-variables Modeling. Society for Industrial and Applied Mathematics, 1997.
[19]	陶惠林, 徐良骥, 冯海宽,等. 基于无人机数码影像的冬小麦株高和生物量估算[J/OL]. 农业工程学报, 2019, 35(19): 107-116 [2020-12-25]. https://doi.org/10.11975/j.issn.1002-6819.2019.19.013. doi: https://doi.org/10.11975/j.issn.1002-6819.2019.19.013
	Tao Huilin, Xu Liangji, Feng Haikuan, et al. Estimation of plant height and biomass of winter wheat based on UAV digital image[J/OL]. Transactions of the Chinese Society of Agricultural Engineering, 2019, 35(19): 107-116[2020-12-25]. https://doi.org/10.11975/j.issn.1002-6819.2019.19.013. doi: https://doi.org/10.11975/j.issn.1002-6819.2019.19.013
[20]	Yeniaysupsu O, Goktas A. A comparison of partial least squares regression with other prediction methods[J]. Hacettepe University Bulletin of Natural Sciences and Engineering Series B: Mathematics and Statistics, 2002, 31: 99-111.
[21]	吴芳, 李映雪, 张缘园,等. 基于机器学习算法的冬小麦不同生育时期生物量高光谱估算[J/OL]. 麦类作物学报, 2019, 39(2): 95-102 [2020-12-25]. https://doi.org/10.7606/j.issn.1009-1041.2019.02.13. doi: https://doi.org/10.7606/j.issn.1009-1041.2019.02.13
	Wu Fang, Li Yingxue, Zhang Yuanyuan, et al. Hyperspectral estimation of biomass of winter wheat at different growth stages based on machine learing algorithms[J/OL]. Journal of Triticeae Crops, 2019, 39(2): 95-102[2020-12-25]. https://doi.org/10.7606/j.issn.1009-1041.2019.02.13. doi: https://doi.org/10.7606/j.issn.1009-1041.2019.02.13
[22]	岳继博, 杨贵军, 冯海宽. 基于随机森林算法的冬小麦生物量遥感估算模型对比[J/OL]. 农业工程学报, 2016, 32(18): 175-182 [2020-12-25]. https://doi.org/10.11975/j.issn.1002-6819.2016.18.024. doi: https://doi.org/10.11975/j.issn.1002-6819.2016.18.024
	Yue Jibo, Yang Guijun, Feng Haikuan. Comparative of remote sensing estimation models of winter wheat biomass based on random forest algorithm[J/OL]. Transactions of the Chinese Society of Agricultural Engineering, 2016, 32(18): 175-182 [2020-12-25]. https://doi.org/10.11975/j.issn.1002-6819.2016.18.024. doi: https://doi.org/10.11975/j.issn.1002-6819.2016.18.024
[23]	Belgiu M Dr?, gu? L. Random forest in remote sensing: A review of applications and future directions[J/OL]. Journal of Photogrammetry and Remote Sensing, 2016, 114: 24-31 [2020-12-25]. https://doi.org/10.1016/j.isprsjprs.2016.01.011. doi: https://doi.org/10.1016/j.isprsjprs.2016.01.011
[24]	第五鹏瑶, 卞希慧, 王姿方,等. 光谱预处理方法选择研究[J/OL]. 光谱学与光谱分析, 2019, 39(9): 2800-2806 [2020-12-25]. https://doi.org/10.3964/j.issn.1000-0593(2019)09-2800-07. doi: https://doi.org/10.3964/j.issn.1000-0593(2019)09-2800-07
	Diwu Pengyao, Bian Xihui, Wang Zifang, et al. Study on the selection of spectral preprocessing methods[J/OL]. Spectroscopy and Spectral Analysis, 2019, 39(9): 2800-2806 [2020-12-25]. https://doi.org/10.3964/j.issn.1000-0593(2019)09-2800-07. doi: https://doi.org/10.3964/j.issn.1000-0593(2019)09-2800-07
[25]	王玉娜, 李粉玲, 王伟东,等. 基于连续投影算法和光谱变换的冬小麦生物量高光谱遥感估算[J/OL]. 麦类作物学报, 2020, 40(11): 1389-1398 [2020-12-25]. https://doi.org/10.7606/j.issn.1009-1041.2020.11.14. doi: https://doi.org/10.7606/j.issn.1009-1041.2020.11.14
	Wang Yuna, Li Fenling, Wang Weidong, et al. Hyper-spectral remote sensing stimation of shoot biomass of winter wheat based on SPA and transformation spectra[J]. Journal of Triticeae Crops, 2020, 40(11): 1389-1398 [2020-12-25]. https://doi.org/10.7606/j.issn.1009-1041.2020.11.14. doi: https://doi.org/10.7606/j.issn.1009-1041.2020.11.14
[26]	马文君, 常庆瑞, 田明璐,等. 棉花全生育期叶片SPAD值的遥感估算模型[J/OL]. 干旱地区农业研究, 2017, 35(5): 42-48 [2020-12-25]. https://doi.org/10.7606/j.issn.1000-7601.2017.05.07. doi: https://doi.org/10.7606/j.issn.1000-7601.2017.05.07
	Ma Wenjun, Chang Qingrui, Tian Minglu, et al. Remote sensing estimation model of cotton leaf SPAD value at the whole growth period[J/OL]. Agricultural Research in the Arid Areas, 2017, 35(5): 42-48 [2020-12-25]. https://doi.org/10.7606/j.issn.1000-7601.2017.05.07. doi: https://doi.org/10.7606/j.issn.1000-7601.2017.05.07
[27]	Daniels A J, Poblete-Echeverría C, Opara U L, et al. Measuring internal maturity parameters contactless on intact table grape bunches using NIR spectroscopy[J/OL]. Frontiers in Plant Science, 2019, 10: 1517 [2020-12-25]. https://doi.org/10.3389/ fpls.2019.01517. doi: https://doi.org/10.3389/ fpls.2019.01517
[28]	Yang W, Yang C, Hao Z, et al. Diagnosis of plant cold damage based on hyperspectral imaging and convolutional neural network[J/OL]. IEEE Access, 2019(99): 1-1[2020-12-25]. https://doi.org/10.1109/ACCESS.2019.2936892. doi: https://doi.org/10.1109/ACCESS.2019.2936892
[29]	Han L, Yang G, Dai H, et al. Modeling maize above-ground biomass based on machine learning approaches using UAV remote-sensing data[J/OL]. Plant Methods, 2019, 15(1): 10 [2020-12-25]. https://doi.org/10.1186/s13007-019-0394-z. doi: https://doi.org/10.1186/s13007-019-0394-z
[30]	Lu Ning, Zhou Jie, Han Zixu, et al. Improved estimation of aboveground biomass in wheat from RGB imagery and pointcloud data acquired with a low-cost unmanned aerial vehicle system[J/OL]. Plant Methods, 2019, 15: 17 [2020-12-25]. https://doi.org/10.1186/s13007-019-0402-3. doi: https://doi.org/10.1186/s13007-019-0402-3

棉花品种 Cotton variety	处理 Treatment	样本数量 Sample number	最大值 Maximum	最小值 Minimum	平均值 Mean	标准差 Standard deviation
新陆早53号	N₀	15	2.73	0.32	1.30	0.74
Xinluzao 53	N1	15	2.87	0.34	1.38	0.78
	N₂	15	3.53	0.34	1.43	0.96
	NC	15	3.31	0.51	1.37	0.83
	N₃	15	2.83	0.45	1.37	0.81
	N₄	15	3.09	0.34	1.26	0.84
新陆早45号	N₀	15	0.43	0.28	0.36	0.05
Xinluzao 45	N₁	15	0.76	0.33	0.58	0.13
	N₂	15	1.12	0.56	0.82	0.13
	NC	15	1.41	0.73	0.97	0.16
	N₃	15	1.71	0.68	1.08	0.27
	N₄	15	2.19	0.87	1.35	0.41

预处理方法 Pretreatment	特征波长数量 Number of characteristic wavelengths	光谱样本提取的特征波长 Characteristic wavelength/nm
原始光谱Original spectrum	10	873、938、962、958、758、942、960、940、727、556
一阶导数First derivative	9	940、931、942、947、971、511、976、936、949
二阶导数 Second derivative	24	942、962、929、758、967、938、654、918、958、769、980、971、936、993、978、878、940、920、922、1 000、820、911、900、924
SG平滑 Savitzky-Golay smoothing	25	916、951、900、676、907、993、820、876、982、973、953、962、765、991、958、938、987、1 000、889、931、971、942、447、922、700
多元散射校正Multiplicative scatter correction	26	873、678、949、534、984、931、907、998、576、971、902、924、762、760、936、944、1 000、727、982、891、967、958、976、916、978、991

预处理 Pretreatment	偏最小二乘法回归（PLSR） Partial least squares regression		随机森林回归（RFR） Random forest regression
预处理 Pretreatment	R²	RMSE	R²	RMSE
原始光谱Original spectrum	0.53	0.49	0.84	0.47
一阶导数First derivative	0.48	0.47	0.87	0.45
二阶导数Second derivative	0.52	0.46	0.84	0.45
SG平滑Savitzky-Golay smoothing	0.63	0.42	0.65	0.46
多元散射校正Multiplicative scatter correction	0.32	0.54	0.79	0.50

预处理 Pretreatment	偏最小二乘法回归（PLSR） Partial least squares regression		随机森林回归（RFR） Random forest regression
预处理 Pretreatment	R²	RMSE	R²	RMSE
原始光谱Original spectrum	0.63	0.44	0.73	0.45
一阶导数First derivative	0.55	0.50	0.81	0.37
二阶导数Second derivative	0.57	0.49	0.75	0.40
SG平滑Savitzky-Golay smoothing	0.67	0.44	0.71	0.62
多元散射校正Multiplicative scatter correction	0.21	0.66	0.57	0.56