基于高斯回归分析的水稻氮素敏感波段筛选及含量估算

doi:10.3964/j.issn.1000-0593(2021)06-1722-08

Abstract
References
Related Citation (15)

Download: PDF (3841 KB)
Export: BibTeX | EndNote (RIS)

Abstract Accurate detection of rice nitrogen content is an important aspect of precision fertilization in rice fields. Nitrogen content variation of rice leaves will cause changes in emissivity of leaves and canopy. Hyperspectral remote sensing is one of the key technologies for non-destructive monitoring of crop nitrogen. This study focuses on the study of nitrogen content monitoring and the sensitive band’s selection through machine learning methods based on 2-year rice nitrogen fertilization experiments carried out in Jianli Hubei during 2018—2019. Hyperspectral reflectance spectral data at the leaf and canopy level and the corresponding leaf nitrogen content data were collected at rice tillering, jointing, booting, flowering and filling stage, respectively. Correlation analysis and Gaussian process regression (GPR) were used to select nitrogen sensitive bands for raw spectra and first-order derivative reflectance (FDR) spectra in rice leaves and canopy level. The nitrogen content estimation models were then constructed through single-band regression analysis, Random Forest (RF) and Support Vector Regression (SVR) method for rice raw spectra data. The Gaussian Process Regression-Random Forest (GPR-RF), Gaussian Process Regression- Support Vector Regression (GPR-SVR), and GPR method were also used to construct the nitrogen estimation model for the nitrogen-sensitive selection bands. The results showed that the GPR method’s sensitive bands were consistent with variation rule of the nitrogen content and spectral changes in rice. The leaf-level model’s over all accuracy was higher than that of the canopy-level model under the same conditions while using FDR spectra was more accurate at canopy level for it could attenuate the effect of background noise in the rice field. R² of calibration datasets and validation sets are 0.79 and 0.84 at the leaf level, while 0.80 and 0.77 at canopy level. Compared with the correlation regression model, the machine learning methods were less affected by rice growth stages (R²>0.80, NRMSE<10%). RF was more suitable than SVR for modeling GPR-selection nitrogen sensitive bands, and the GPR-RF model can use about 1.5% of the bands to reach the accuracy of the RF model using all the bands. The GPR model works well on nitrogen estimation through nitrogen -sensitive bands at leaf and canopy level, not only for the full-growth stage but also for the single-growth stage(R²>0.94, NRMSE<6%). Besides, compared with the other four machine learning methods, the GPR model can improve the accuracy and stability of the estimation of nitrogen content at the canopy level that R² increased by 0.02 and NRMSE decreased by 1.2%. GPR method provides a methodological reference for selecting crop nitrogen hyper-spectrally sensitive bands and inversion of the nitrogen content of leaves and canopy level during rice different growth period.

Key words：Sensitive band; Nitrogen; Gaussian progresses regression (GPR); Random forest (RF); Support vector regression (SVR); Hyperspectral

Received: 2020-06-03 Accepted: 2020-09-28

ZTFLH:

TP79

Corresponding Authors: SONG Xiao-yu, YANG Gui-jun E-mail: songxy@nercita.org.cn; yanggj@nercita.org.cn

Cite this article:

URL:

https://www.gpxygpfx.com/EN/10.3964/j.issn.1000-0593(2021)06-1722-08 OR https://www.gpxygpfx.com/EN/Y2021/V41/I06/1722

[1] YANG Fu-qin, DAI Hua-yang, FENG Hai-kuan, et al（杨福芹, 戴华阳, 冯海宽, 等）. Transactions of the Chinese Society of Agricultural Engineering（农业工程学报）, 2016, 32(23): 161.
[2] TONG Qing-xi, ZHANG Bing, ZHENG Lan-fen（童庆禧, 张兵, 郑兰芬）. Hyperspectral Remote Sensing: Principles, Techniques, and Applications(高光谱遥感：原理技术与应用). Beijing: Higher Education Press（北京：高等教育出版社）, 2006. 38.
[3] Berger K, Verrelst J, Jean-Baptiste Féret, et al. Remote Sensing of Environment, 2020, 242(1): 111758.
[4] LI Lan-tao, WANG Shan-qin, REN Tao, et al（李岚涛, 汪善勤, 任涛, 等）. Transactions of the Chinese Society of Agricultural Engineering（农业工程学报）, 2016, 32(14): 209.
[5] WANG Ren-chao, CHEN Ming-zhen, JIANG Heng-xian（王人潮, 陈铭臻, 蒋亨显）. Journal of Zhejiang University·Agriculture and Life Sciences（浙江农业大学学报）, 1993, 19(Supp-1): 7.
[6] ZHOU Qi-fa, WANG Ren-chao（周启发, 王人潮）. Journal of Zhejiang University·Agriculture and Life Sciences（浙江农业大学学报）, 1993, 19(Supp-1): 40.
[7] FENG Shuai, XU Tong-yu, YU Feng-hua, et al（冯帅, 许童羽, 于丰华, 等）. Spectroscopy and Spectral Analysis（光谱学与光谱分析）, 2019, 39(10): 3281.
[8] Feng W，Yao X，Zhu Y，Tian Y C，Cao W X. European Journal of Agronomy, 2008, 28(3): 394.
[9] YUE Ji-bo, YANG Gui-jun, FENG Hai-kuan（岳继博, 杨贵军, 冯海宽）. Transactions of the Chinese Society of Agricultural Engineering（农业工程学报）, 2016, 32(18): 175.
[10] Smola A J, Schlkopf B. Statistics and Computing, 2004, 14(3): 199.
[11] Jochem V, Malenovsky Zbyněk, Christiaan V D T, et al. Surveys in Geophysics, 2018. 1.
[12] Rasmussen C E, Williams C K I. Gaussian Processes for Machine Learning. The MIT Press. New York，2006. 7.
[13] Camps-Valls G, Verrelst J, Munoz-Mari J, et al. IEEE Geoence and Remote Sensing Magazine, 2016, 4(2): 58.
[14] Breiman L. Machine Learning，2001 45：5.
[15] Vapnik V. Statistical Learning Theory. New York, NY：Wiley，1998. 401.
[16] Verrelst J, Rivera J P, Gitelson A, et al. International Journal of Applied Earth Observation and Geoinformation, 2016, 52: 554.