Classification and identification of different wood species are demonstrated using a portable near-infrared spectrometer, combined with four spectral pretreatment methods and three pattern recognition methods. Additional chemometric tools were used for comprehensive evaluation of classification model accuracy and complexity.
