Monitoring of Nitrogen Indices in Wheat Leaves Based on the Integration of Spectral and Canopy Structure Information

Canopy spectral reflectance can indicate both crop nutrient and canopy structural information. Differences in structure affect reflectance. However, a non-imaging spectrometer cannot distinguish such differences while monitoring nutrients, because the results are likely to be influenced by structure. In addition, nitrogen application rate is one of main factors influencing crops. Strong correlations exist between indices leaf nitrogen, thus, these used compensate for content wheat leaves. this study, (CSI) as coverage, height, textural features were obtained based on RGB height images RGB-D camera. Moreover, was an ASD hyperspectral spectrometer, which two vegetation indices—ratio index (RVI) angular insensitivity (AIVI)—were constructed. With CSIs input parameters, model established predict (LNC) accumulation (LNA) partial least squares (PLS) random forest (RF) regression algorithms. The showed that RF with RVI CSI inputs had highest prediction accuracy LNA, coefficient determination (R2) reached 0.79, root mean square error (RMSE) 1.54 g/m2. coverage relatively important model. PLS AIVI parameters LNC, R2 0.78 RMSE 0.35%, among indices. parts important. suggested algorithms effectively integrate information, information supplement improving LNA LNC.