Vis-NIR spectroscopy for the on-site prediction of wood properties

Abstract

We investigated the feasibility of visible-near-infrared (Vis-NIR) spectroscopy for estimation of wood qualities of fast-growing Eucalyptus grandis. Partial least squares regression (PLSR) models are applied to predict the diameter at the breast height (DBH), lateral growth rate (LGR) and propagation velocity of stress waves (PVSW). It was possible to estimate LGR and PVSW with appropriate accuracy. This suggested that perhaps information in terms of maturation is included in Vis-NIR spectra. The key factors in the validation of PVSW and LGR were the water and cellulose condition in wood.