Identifying Wood Based on Near-Infrared Spectra and Four Gray-Level Co-Occurrence Matrix Texture Features

Identifying wood accurately and rapidly is one of the best ways to prevent product fakes adulterants in forestry products. Wood identification traditionally relies heavily on special experts that spend extensive time laboratory. A new method proposed uses near-infrared (NIR) spectra at a wavelength 780–2300 nm incorporated with gray-level co-occurrence (GLCM) texture feature identify timbers. The NIR spectral features were determined by principal component analysis (PCA), digital image extracted GLCM used create support vector machine (SVM) model results from fusion raw four 25 timbers showed accuracy was 99.43%. sample anisotropy heterogeneity comparative revealed information transverse surface had more characteristics than tangential radial surfaces. Furthermore, short-wavelength pre-processed bands 780–1100 1100–2300 realized high 99.43% 100%, respectively. effective for improving data spatial clustering features.