Non-destructive method for wood identification using conventional X-ray computed tomography data

Researchers have developed a non-destructive method for identifying wood species using conventional X-ray computed tomography (CT) data, achieving high accuracy even at low resolution. The study applied machine learning techniques—specifically gray-level co-occurrence matrix (GLCM) and local binary patterns (LBP)—to CT images of six hardwood species used in the Tripitaka Koreana woodblocks. While anatomical microstructures were not visually discernible, the LBP system outperformed GLCM slightly, with accuracy further improved by calibrating gray levels to density. This approach offers a novel way to analyze wooden historical objects without physical damage, provided candidate species and sufficient training data are available.

Kobayashi, K., Hwang, S.-W., Okochi, T., Lee, W.-H., & Sugiyama, J. (2019). Non-destructive method for wood identification using conventional X-ray computed tomography data. Journal of Cultural Heritage, 38, 88–93. https://doi.org/10.1016/j.culher.2019.02.001