engleski

Prediction of soybean leaf nitrogen content using proximal field spectroscopy

Calibration models based on multiple linear regression (MLR), partial least squares regression (PLSR) and artificial neural networks (ANN) were developed in order to predict leaf nitrogen (N) content in soybean utilizing hyperspectral reflectance data. Research was conducted on experimental field with 9 treatments of mineral N fertilization (0-300 kg N ha-1). Total of 180 soybean leaf samples in the beginning of flowering (R1) were scanned using field spectroradiometer (350-1050 nm). The highest correlations with soybean N content were found in red edge to NIR region. Validation models revealed strong to very strong correlation between predicted and measured values (MLR: R2=0.39 ; PLSR: R2=0.68 ; ANN: R2=0.70).

leaf reflectance ; soybean ; leaf nitrogen content ; principal component analysis ; multivariate regression

nije evidentirano