基于高光谱小波变换的玉米叶片含水量估测

doi:10.3964/j.issn.1000-0593(2025)10-2875-10

摘要
参考文献
相关文章 (15)

全文: PDF (9044 KB)
输出: BibTeX | EndNote (RIS)

摘要：在植物的生命活动中，水分对于作物产量起着决定性的作用。快速检测并获知植物叶片水分状况，对于了解田间作物的生理需水特性以及相应的水分管理具有重要的意义。高光谱指数是进行作物叶片含水量无损实时估测的重要参数，然而常用的光谱指数在估测叶片含水量中受生育时期影响明显，且稳定性较差，估测精度很难达到生产要求。为了提高玉米叶片水分估测的精度，于2023年—2024年在内蒙古玉米种植的典型区域进行不同水分梯度的田间试验，测定玉米叶片三个关键生育时期的高光谱反射率，通过建立叶片含水量（LWC）与小波函数及光谱指数的关系模型，以确定表现最佳的小波函数和光谱指数，并评估他们在检测玉米叶片含水量方面的稳定性和鲁棒性。利用光谱指数和小波函数对叶片含水量进行相关性分析发现，选用的13个水分指数中MDATT指数的预测结果最好（R²=0.52），但估测精度受生育时期和层位影响较大。相比之下，连续小波变换在提高LWC估测精度的同时克服了生育时期和层位对于预测精度的影响，其中表现最佳的小波函数及其特征为Coif3（S₆W₁₇₂₅）（R²=0.83）。与光谱指数相比小波函数中的Coif3函数在估测玉米叶片含水量方面更为稳定，模型独立验证结果的决定系数R²为0.76，并且验证误差最小，RMSE和RE分别为3.08%和3.51%。研究结果可为玉米生长关键期水分的准确判断和灌溉的精准管理提供指导，从而有助于中国中西部地区玉米水肥一体化种植制度的可持续性发展。

关键词：玉米叶片；含水量；小波函数；光谱指数

Abstract：In the life activities of plants, water plays a decisive role in crop yield. Rapid detection and acquisition of plant leaf water status is of great significance for understanding the physiological water requirements of field crops and corresponding water management. Hyperspectral indices are an important means of non-destructive, real-time estimation of crop leaf water content. However, the commonly used spectral index is significantly affected by the growth period in estimating leaf water content, and the stability is poor. Meeting production requirements is challenging due to the estimation accuracy. To achieve the accuracy of corn leaf water estimation and realize efficient use of corn, this study conducted field experiments with different moisture gradients in typical corn-growing areas in Inner Mongolia from 2023 to 2024, measured the hyperspectral reflectance of corn leaves at three key growth periods, and established a relationship model between leaf water content (LWC) and wavelet function and spectral index to determine the best performing wavelet function and spectral index, and evaluated their stability and robustness in detecting corn leaf water content. The results showed that the correlation analysis of leaf water content using spectral index and wavelet function found that the MDATT index had the best prediction result (R²=0.52) among the 13 selected water indices. Still, the estimation accuracy was greatly affected by the growth period and layer. In contrast, the continuous wavelet transform improved the estimation accuracy of LWC while overcoming the influence of growth period and layer on the prediction accuracy. Among them, the best performing wavelet function and its characteristics were Coif3 (S₆W₁₇₂₅) (R²=0.83). Compared with the spectral index, the Coif3 function in the wavelet function was more stable in estimating the water content of corn leaves. The determination coefficient R² of the independent verification result of the model was 0.76, and the verification error was the smallest, with RMSE and RE of 3.08% and 3.51%, respectively. The research results enable the accurate assessment of water content during the critical growth period of corn and the precise management of irrigation, thereby contributing to the sustainable development of the integrated water-fertilizer corn planting system in central and western China.

Key words：Corn leaf; Moisture content; Wavelet function; Vegetation index

收稿日期: 2024-12-21 修订日期: 2025-07-15

中图分类号:

S127

基金资助: 国家重点研发计划项目（2022YFD1900305-03-01），国家重点研发计划项目（2022YFD1900805-02），国家自然科学基金项目（31030273）资助

通讯作者: 李斐 E-mail: Lifei@imau.edu.cn

作者简介: 肖娅婷，女，1998年生，内蒙古农业大学资源与环境学院硕士研究生 e-mail: XiaoYT@emails.imau.edu.cn

引用本文:

肖娅婷，唐彧哲，白宇飞，王潞，李斐. 基于高光谱小波变换的玉米叶片含水量估测[J]. 光谱学与光谱分析, 2025, 45(10): 2875-2884.
XIAO Ya-ting, TANG Yu-zhe, BAI Yu-fei, WANG Lu, LI Fei. Estimation of Leaf Moisture Content of Maize Based on Spectral Index and Wavelet Transform. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2025, 45(10): 2875-2884.

链接本文:

https://www.gpxygpfx.com/CN/10.3964/j.issn.1000-0593(2025)10-2875-10 或 https://www.gpxygpfx.com/CN/Y2025/V45/I10/2875

[1] Datt B. Australian Journal of Botany, 1999, 47(6): 909.
[2] Furbank R T, Tester M. Trends in Plant Science, 2011, 16(12): 635.
[3] Cheng T, Rivard B, Sanchez-Azofeifa A. Remote Sensing of Environment, 2011, 115(2): 659.
[4] TAN Xian-ming, WANG Zhong-lin, ZHANG Jia-wei, et al(谭先明, 王仲林, 张佳伟, 等). Agricultural Research in the Arid Areas(干旱地区农业研究), 2021, 39(4): 155.
[5] WANG Yan-cang, ZHANG Xiao-yu, JIN Yong-tao, et al(王延仓, 张萧誉, 金永涛, 等). Journal of Triticeae Crops(麦类作物学报), 2020, 40(4): 503.
[6] Gu C, Ji S, Xi X, et al. Frontiers in Plant Science, 2022, 13: 931789.
[7] Yao X, Si H, Cheng T, et al. Frontiers in Plant Science, 2018, 9: 1360.
[8] Li D, Wang X, Zheng H, et al. Plant Methods, 2018, 14: 76.
[9] Hunt Jr·E R, Rock B N, Nobel P S. Remote Sensing of Environment, 1987, 22(3): 429.
[10] Shi B, Yuan Y, Zhuang T, et al. European Journal of Agronomy, 2022, 139: 126548.
[11] LIU Xiao-jing, CHEN Guo-qing, WANG Liang, et al(刘晓静，陈国庆，王良，等). Journal of Triticeae Crops(麦类作物学报)，2018，38(7)：854.
[12] ZHANG Hai-wei, ZHANG Fei, ZHANG Xian-long, et al(张海威, 张飞, 张贤龙，等). Spectroscopy and Spectral Analysis(光谱学与光谱分析), 2018, 38(5): 1540.
[13] LIU Er-hua, ZHOU Guang-sheng, ZHOU Li, et al(刘二华, 周广胜, 周莉, 等). Journal of Applied Meteorological Science(应用气象学报), 2020, 31(1): 52.
[14] Zhang J, Zhang W, Xiong S, et al. Plant Methods, 2021, 17: 34.
[15] Ndlovu H S, Odindi J, Sibanda M, et al. Remote Sensing, 2021, 13(20): 4091.
[16] Thenkabail P S, Smith R B, De Pauw E. Remote Sensing of Environment, 2000, 71(2): 158.
[17] Hatfield J L, Prueger J H. Remote Sensing, 2010, 2(2): 562.
[18] Houborg R, Fisher J B, Skidmore A K. International Journal of Applied Earth Observation and Geoinformation, 2015, 43: 1.