不同飞行高度冠层高光谱小波变换估算薄壳山核桃叶片氮素含量

doi:10.3964/j.issn.1000-0593(2025)08-2200-10

摘要
参考文献
相关文章 (15)

全文: PDF (30674 KB)
输出: BibTeX | EndNote (RIS)

摘要：氮素是植物体内氨基酸、蛋白质和叶绿素等物质的组成元素，在植物光合作用中起着重要的作用。利用无人机高光谱技术无损、高效地估算植物氮素含量，对及时掌控树体长势、实施精准管理具有重要意义。飞行高度直接影响植物信息获取精度与效率。通过设置40、60和80 m三个飞行高度获取薄壳山核桃（长林和建德系列）开花期不同分辨率的无人机遥感影像，得到对应高度下的薄壳山核桃冠层光谱，运用连续小波变换（CWT）对原始光谱进行预处理；进一步联合两波段光谱指数（NDSI）分析薄壳山核桃叶片氮含量（LNC）与光谱的响应关系；最后，利用竞争自适应重加权采样-迭代保留信息变量（CARS_IRIV）算法筛选特征变量，构建不同高度下的薄壳山核桃LNC光谱响应的BP神经网络（BPNN）、随机森林（RF）估算模型，以期揭示无人机飞行高度对薄壳山核桃冠层光谱特征及LNC的影响机制。结果表明：CWT处理后的冠层光谱与薄壳山核桃LNC相关性得到提升；CWT结合NDSI在提高与LNC的相关性方面表现更佳；随着飞行高度的增加（从40、60到80 m），无论是单波段光谱还是双波段光谱与LNC的相关性都有所上升；最优LNC估算模型是40 m飞行高度下的CWT-scale 3-NDSI-BPNN模型，模型预测集R²_P=0.73，RMSEP=1.13 g·kg^-1，RPD=1.97。研究结果可为提高薄壳山核桃氮含量遥感估测的准确性提供技术支持，进一步为利用无人机搭载传感设备获取作物信息设定合适的飞行高度提供参考。

关键词：薄壳山核桃；氮素；无人机高光谱；小波变换；飞行高度；机器学习

Abstract：Nitrogen is a constituent element of amino acids, proteins, and chlorophyll in plants, which plays an important role in plant photosynthesis. UAV hyperspectral technology can estimate plant nitrogen content non-destructively and efficiently, which is significant for the timely control of tree growth and precise management. Flight height directly affects the accuracy and efficiency of plant information acquisition. In this study, UAV remote sensing images of different resolutions were acquired during the flowering stage of Carya illinoinensis (Changlin and Jiande series) by setting three flight heights (i.e., 40, 60, and 80 m). Thus, the canopy spectra of Carya illinoinensis at the corresponding heights were obtained. Raw hyperspectral data were preprocessed using the continuous wavelet transform (CWT). Furthermore, the response relationship between the LNC of Carya illinoinensis and the canopy spectrum was analyzed by combining two-band spectral indices (i.e., normalized difference spectral index, NDSI). Finally, the competitive adaptive reweighted sampling-iteratively retaining informative variables (CARS-IRIV) algorithm was used to screen the feature variables. Back propagation neural network (BPNN) and random forest (RF) algorithms were used to construct spectral response estimation models for Carya illinoinensis LNC at different heights, to reveal the impact mechanism of UAV flight heights on the canopy spectral characteristics of Carya illinoinensis and LNC. Results showed improved correlation between the canopy spectrum after CWT pretreatment and Carya illinoinensis LNC. CWT combined with NDSI performed better in improving the correlation with LNC. As the flight height increased (from 40, 60 to 80 m), the correlation with the LNC increases for both single-band and two-band spectra.The optimal LNC estimation model was CWT-scale 3-NDSI-BPNN at 40 m flight height, R²_P=0.73, RMSEP=1.13 g·kg^-1, and RPD=1.97. The research results can provide technical support for improving the accuracy of remote sensing estimation of Carya illinoinensis LNC, and further provide a reference for researchers to use a UAV equipped with sensing devices to obtain crop information and set appropriate flight heights.

Key words：Carya illinoensis; Nitrogen; UAV hyperspectral technology; Wavelet transform; Flight heights; Machine learning

收稿日期: 2024-09-10 修订日期: 2025-01-25

中图分类号:

O657.3

基金资助: 国家自然科学基金项目(32171783）资助

通讯作者: 唐雪海 E-mail: tangxuehai@ahau.edu.cn

作者简介: 孔令瑗，女，2000年生，安徽农业大学林学与园林学院硕士研究生 e-mail: lyk@stu.ahau.edu.cn

引用本文:

孔令瑗，黄庆丰，倪辰，徐佳佳，唐雪海. 不同飞行高度冠层高光谱小波变换估算薄壳山核桃叶片氮素含量[J]. 光谱学与光谱分析, 2025, 45(08): 2200-2209.
KONG Ling-yuan, HUANG Qing-feng, NI Chen, XU Jia-jia, TANG Xue-hai. Estimation of Nitrogen Content of Carya illinoinensis Leaves Based on Canopy Hyperspectral and Wavelet Transform at Different Flight Heights. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2025, 45(08): 2200-2209.

链接本文:

https://www.gpxygpfx.com/CN/10.3964/j.issn.1000-0593(2025)08-2200-10 或 https://www.gpxygpfx.com/CN/Y2025/V45/I08/2200

[1] Zhang X D, Chang J, Ren H D, et al. Frontiers in Plant Science, 2022, 13: 1003728.
[2] PENG Fang-ren, LI Yong-rong, HAO Ming-zhuo, et al(彭方仁, 李永荣, 郝明灼, 等). Journal of Forestry Engineering(林业工程学报), 2012, 26(4): 1.
[3] JIANG Ying-chun, DU Shi-ping, ZOU Ying-wu, et al(江迎春, 杜拾平, 邹英武, 等). Non-wood Forest Research(经济林研究), 2024, 42(1): 77.
[4] TIAN Ting, ZHANG Qing, ZHANG Hai-dong(田婷, 张青, 张海东). Crops(作物杂志), 2020, (5): 1.
[5] HE Yong, DU Xiao-yue, ZHENG Li-yuan, et al(何勇, 杜晓月, 郑力源, 等). Transactions of the Chinese Society of Agricultural Engineering(农业工程学报), 2022, 38(24): 63.
[6] WEI Peng-fei, XU Xin-gang, LI Zhong-yuan, et al(魏鹏飞, 徐新刚, 李中元, 等). Transactions of the Chinese Society of Agricultural Engineering(农业工程学报), 2019, 35(8): 126.
[7] LI Chang-chun, CHEN Peng, LU Guo-zheng, et al(李长春, 陈鹏, 陆国政, 等). Chinese Journal of Applied Ecology(应用生态学报), 2018, 29(4): 1225.
[8] CHEN Peng-fei, LIANG Fei(陈鹏飞, 梁飞). Scientia Agricultura Sinica(中国农业科学), 2019, 52(13): 2220.
[9] JING Yu-hang, GUO Yan, ZHANG Hui-fang, et al(井宇航, 郭燕, 张会芳, 等). Journal of Henan Agricultural Sciences(河南农业科学), 2022, 51(2): 147.
[10] Awais M, Li W, Cheema M J M, et al. International Journal of Environmental Science and Technology, 2022, 19: 2703.
[11] Xu J J, Fu G S, Yan L P, et al. Journal of Soil Science and Plant Nutrition, 2024, 24(1): 1407.
[12] XIA Guo-hua, HUANG Jian-qin, XIE Hong-en, et al(夏国华, 黄坚钦, 解红恩, 等). Journal of Fruit Science(果树学报), 2014, 31(5): 854.
[13] LI Chang-chun, SHI Jin-jin, MA Chun-yan, et al(李长春, 施锦锦, 马春艳, 等). Transactions of the Chinese Society for Agricultural Machinery(农业机械学报), 2021, 52(8): 172.
[14] YANG Fu-qin, FENG Hai-kuan, LI Zhen-hai, et al(杨福芹, 冯海宽, 李振海, 等). Remote Sensing Technology and Application(遥感技术与应用), 2021, 36(2): 353.
[15] PENG Hai-gen, JIN Ying, ZHAN You-guo, et al(彭海根, 金楹, 詹莜国, 等). Journal of Instrumental Analysis(分析测试学报), 2020, 39(10): 1305.
[16] Yun Y H, Wang W T, Tan M L, et al. Analytica Chimica Acta, 2014, 807: 36.
[17] Sun Z, Wang J, Nie L, et al. Chemometrics and Intelligent Laboratory Systems, 2018, 181: 64.
[18] Qi H X, Wu Z Y, Zhang L, et al. Computers and Electronics in Agriculture, 2021, 187: 106292.
[19] CUI Ri-xian, LIU Ya-dong, FU Jin-dong(崔日鲜, 刘亚东, 付金东). Spectroscopy and Spectral Analysis(光谱学与光谱分析), 2016, 36(6): 1837.
[20] Wang F L, Yang M, Ma L F, et al. Remote Sensing, 2022, 14(5): 1251.