AUTHOR=Chen Zhenan , Xue Xiaoming , Wu Haoqi , Gao Handong , Wang Guangyu , Ni Geyi , Cao Tianyi 

TITLE=Visible/near-infrared hyperspectral imaging combined with machine learning for identification of ten Dalbergia species

JOURNAL=Frontiers in Plant Science

VOLUME=Volume 15 - 2024

YEAR=2024

URL=https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2024.1413215

DOI=10.3389/fpls.2024.1413215

ISSN=1664-462X

ABSTRACT=This study tackles the urgent need for non-destructive identification of commercially valuable Dalbergia species, threatened by illegal logging. We integrate Visible/Near-Infrared (Vis/NIR) Hyperspectral Imaging (HSI) with advanced machine learning techniques to significantly enhance the precision and efficiency of wood species identification. Our methodology employs various modeling approaches, including Principal Component Analysis (PCA), Partial Least Squares Discriminant Analysis (PLS-DA), Support Vector Machine (SVM), and Convolutional Neural Networks (CNN), to analyze spectral data across Vis (383-982 nm), NIR (982-2386 nm), and full spectral ranges (383 nm to 2386 nm). We also evaluate the impact of diverse preprocessing techniques-Standard Normal Variate (SNV), Savitzky-Golay (SG) smoothing, normalization, and Multiplicative Scatter Correction (MSC)-on model performance. Notably, with optimal preprocessing, both SVM and CNN models achieve 100% accuracy across NIR and full spectral ranges. Moreover, the selection of an appropriate wavelength range is critical; utilizing the full spectrum captures a wider array of wood's chemical and physical properties, significantly enhancing model accuracy and predictive power. These results highlight the effectiveness of Vis/NIR HSI in wood species identification and emphasize the importance of precise wavelength selection and preprocessing techniques to maximize both accuracy and cost-efficiency. This research offers substantial contributions to ecological conservation, biodiversity preservation, and the regulation of the timber trade.