The preservation of forage is critical for obtaining consistent feed supplies for ruminant animals. High quality hay can be obtained monitoring the moisture level at harvest time after conditioning. Near-infrared reflectance spectroscopy (NIRS) is a good alternative for providing timely forage moisture content. The aim of this study was to evaluate the potentiality of the in-field Vis/NIR hyperspectral imaging combined with chemometric to predict moisture content of alfalfa after conditioning.
Several combinations of conditioning level, time of day (morning and afternoon), and time after the conditioning (0, 15 and 120 min) were considered to carry out hyperspectral acquisitions. Principal component analysis (PCA) was applied to group the samples according to time of day and minutes after the conditioning, while partial least square regression (PLSr) model was developed to predict the moisture content reporting R 2 =0.840 and RMSECV=2.05%. The reduction of the number of variables does not affect the goodness of the models.
Results confirm the potentiality of the in-field and non-destructively determination of alfalfa moisture content by using HSI combined with linear chemometric techniques.