Wood moisture content predictability of near Infrared Spectroscopy: Effect of specimen thickness and moisture content

Streszczenie

Brak zatwierdzonego streszczenia. Przeczytaj abstrakt poniżej lub otwórz oryginalną publikację.

Abstrakt

Near Infrared Spectroscopy (NIRS) is an excellent tool being applied on organic materials for quality and process control.Sorting of kiln dried wood based on moisture content has always remained a tedious task during industrial production.The current study is centred on evaluation of suitability of NIRS for real time sorting of kiln dried wood with broad range of thicknesses and moisture content.A partial least square (PLS) regression model was developed including a broad range of moisture content and wood thicknesses.Coefficient of determination, Root mean square error of cross validation (RMSECV) and ratio of performance deviation (RPD) values of the calibration model were 0.966, 2.18 and 5.41 respectively.The results indicated that the NIR model could predict moisture content accurately only for moisture content below 25% for all the thickness ranges.It was found that the deviations in MC from actual values significantly depended upon the thicknesses of the wooden specimens used.

Abstrakt w wersji dostarczonej przez źródło. Pełny tekst może być ograniczony przez wydawcę.