EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Quantitative Analysis of Adulteration in Anoectochilus roxburghii Powder Using Hyperspectral Imaging and Multi-Channel Convolutional Neural Network

Ziyuan Liu, Tingsong Zhang, Haoyuan Ding, Zhangting Wang, Hongzhen Wang, Lu Zhou, Yujia Dai, Yiqing Xu

Agronomy · 06.08.2025 · 15 (8) · 1894

DOI: 10.3390/agronomy15081894 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Agronomy, online erschienen am 06.08.2025. Verfasst von Ziyuan Liu und 7 weiteren Autor:innen. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 44 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: The Zhejiang Province Welfare Technology Applied Research Project, The Zhejiang Provincial Natural Science Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Adulteration detection in medicinal plant powders remains a critical challenge in quality control. In this study, we propose a hyperspectral imaging (HSI)-based method combined with deep learning models to quantitatively analyze adulteration levels in Anoectochilus roxburghii powder. After preprocessing the spectral data using raw, first-order, and second-order Savitzky–Golay derivatives, we systematically evaluated the performance of traditional machine learning models (Random Forest, Support Vector Regression, Partial Least Squares Regression) and deep learning architectures. While traditional models achieved reasonable accuracy (R2 up to 0.885), their performance was limited by feature extraction and generalization ability. A single-channel convolutional neural network (CNN) utilizing individual spectral representations improved performance marginally (maximum R2 = 0.882), but still failed to fully capture the multi-scale spectral features. To overcome this, we developed a multi-channel CNN that simultaneously integrates raw, SG-1, and SG-2 spectra, effectively leveraging complementary spectral information. This architecture achieved a significantly higher prediction accuracy (R2 = 0.964, MSE = 0.005), demonstrating superior robustness and generalization. The findings highlight the potential of multi-channel deep learning models in enhancing quantitative adulteration detection and ensuring the authenticity of herbal products.

Förderung

  • The Zhejiang Province Welfare Technology Applied Research Project
  • The Zhejiang Provincial Natural Science Foundation of China

Zitieren

APA

Liu, Z., Zhang, T., Ding, H., Wang, Z., Wang, H., Zhou, L., Dai, Y., & Xu, Y. (2025). Quantitative Analysis of Adulteration in Anoectochilus roxburghii Powder Using Hyperspectral Imaging and Multi-Channel Convolutional Neural Network. Agronomy, 15(8), 1894. https://doi.org/10.3390/agronomy15081894

Vancouver

Liu Z, Zhang T, Ding H, Wang Z, Wang H, Zhou L, et al. Quantitative Analysis of Adulteration in Anoectochilus roxburghii Powder Using Hyperspectral Imaging and Multi-Channel Convolutional Neural Network. Agronomy. 2025;15(8):1894. doi:10.3390/agronomy15081894

BibTeX
@article{Liu2025,
  author = {Liu, Ziyuan and Zhang, Tingsong and Ding, Haoyuan and Wang, Zhangting and Wang, Hongzhen and Zhou, Lu and Dai, Yujia and Xu, Yiqing},
  title = {Quantitative Analysis of Adulteration in Anoectochilus roxburghii Powder Using Hyperspectral Imaging and Multi-Channel Convolutional Neural Network},
  journal = {Agronomy},
  year = {2025},
  volume = {15},
  number = {8},
  pages = {1894},
  doi = {10.3390/agronomy15081894}
}
Kostenlos starten →