Yang Jian, Du Lin, Gong Wei, Shi Shuo, Sun Jia
Artificial Intelligence School, Wuchang University of Technology, Wuhan, Hubei 430223, People's Republic of China.
School of Geography and Information Engineering, China University of Geosciences, Wuhan, Hubei 430074, People's Republic of China.
R Soc Open Sci. 2020 Feb 26;7(2):191941. doi: 10.1098/rsos.191941. eCollection 2020 Feb.
Leaf nitrogen concentration (LNC) is a major indicator in the estimation of the crop growth status which has been diffusely applied in remote sensing. Thus, it is important to accurately obtain LNC by using passive or active technology. Laser-induced fluorescence can be applied to monitor LNC in crops through analysing the changing of fluorescence spectral information. Thus, the performance of fluorescence spectrum (FS) and first-derivative fluorescence spectrum (FDFS) for paddy rice (Yangliangyou 6 and Manly Indica) LNC estimation was discussed, and then the proposed FS + FDFS was used to monitor LNC by multivariate analysis. The results showed that the difference between FS ( = 0.781, s.d. = 0.078) and FDFS ( = 0.779, s.d. = 0.097) for LNC estimation by using the artificial neural network is not obvious. The proposed FS + FDFS can improved the accuracy of LNC estimation to some extent ( = 0.813, s.d. = 0.051). Then, principal component analysis was used in FS and FDFS, and extracted the main fluorescence characteristics. The results indicated that the proposed FS + FDFS exhibited higher robustness and stability for LNC estimation ( = 0.851, s.d. = 0.032) than that only using FS ( = 0.815, s.d. = 0.059) or FDFS ( = 0.801, s.d. = 0.065).
叶片氮浓度（LNC）是估算作物生长状况的一个主要指标，已广泛应用于遥感领域。因此，利用被动或主动技术准确获取LNC很重要。激光诱导荧光可通过分析荧光光谱信息的变化来监测作物中的LNC。因此，讨论了荧光光谱（FS）和一阶导数荧光光谱（FDFS）对水稻（扬两优6号和曼利籼稻）LNC估算的性能，然后利用提出的FS + FDFS通过多变量分析来监测LNC。结果表明，使用人工神经网络进行LNC估算时，FS（r = 0.781，标准差 = 0.078）和FDFS（r = 0.779，标准差 = 0.097）之间的差异不明显。提出的FS + FDFS在一定程度上可以提高LNC估算的准确性（r = 0.813，标准差 = 0.051）。然后，对FS和FDFS进行主成分分析，并提取主要荧光特征。结果表明，与仅使用FS（r = 0.815，标准差 = 0.059）或FDFS（r = 0.801，标准差 = 0.065）相比，提出的FS + FDFS在LNC估算方面表现出更高的稳健性和稳定性（r = 0.851，标准差 = 0.032）。
