EUVIMEDEUVIMED entdecken
Crossref

Nondestructive Determination for Multiple Quality Indicators of Fermented Fava Bean Paste: Synergistic Fusion Strategies for Combining the NIR and Hyperspectral Information

Min Xu iD, Junwei Xie, Lan Li, Zedong Xu, Yang Kang, Pengfei Jia, Yuzhen Wei, Jie Tang, Hongbin Lin iD, Ping Liu iD

Journal of Food Science · 2026-05 · 91 (5)

DOI: 10.1111/1750-3841.71142 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Food Science, online erschienen am 18.05.2026. Verfasst von Min Xu und 9 weiteren Autor:innen. Beteiligte Einrichtung: College of Food and Bioengineering, Xihua University, Chengdu, China. Hauptthemen laut MEDLINE: Hyperspectral Imaging, Vicia faba, Fermented Foods. Die Arbeit nennt 42 Literaturangaben. Gefördert von: the National Science Foundation for Young Scientists of China, the Sichuan Science and Technology Program, the Talent introduction project of Xihua University.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This study investigates the combined application of near‐infrared spectroscopy (NIR) and hyperspectral imaging (HSI) for monitoring multiple quality indices of fermented fava bean paste including total acidity, ammonia nitrogen, reducing sugars, tartaric acid, and lactic acid. Low‐level fusion (LLF) and mid‐level fusion (MLF) strategies using radial basis function neural networks are introduced for quantitatively analyzing the quality parameters. Genetic algorithm (GA) integrated with principal component analysis (PCA) is applied for obtaining synergetic fused spectral data. The results indicate that models with fusion spectral exhibit superior accuracy than independent information. LLF‐GA‐PCA outperforms in analyzing tartaric acid ( R 2 = 0.9082) and lactic acid ( R 2 = 0.9898), while MLF‐GA‐PCA excels in total acidity ( R 2 = 0.9795), ammonia nitrogen ( R 2 = 0.9855), and reducing sugars ( R 2 = 0.9882). The study highlights the effectiveness of integrating NIR and HSI in rapidly analyzing quality of fermented fava bean paste.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Hyperspectral ImagingVicia fabaFermented FoodsSpectroscopy, Near-InfraredFermentationPrincipal Component AnalysisLactic AcidAlgorithmsTartrates

Stichwörter der Autor:innen

data fusionfava bean pastehyperspectral imagingnear‐infrared spectroscopyquality monitoringradial basis function neural network

Substanzen

Lactic Acidtartaric acidTartrates

Autor:innen und Institutionen

  • College of Food and Bioengineering, Xihua University, Chengdu, China.
  • School of Electrical Engineering, Guangxi University, Nanning, China.
  • School of Information Engineering, Huzhou University, Huzhou, China.

ORCID-Profile: Min Xu, Hongbin Lin, Ping Liu

Förderung

  • the National Science Foundation for Young Scientists of China
  • the Sichuan Science and Technology Program
  • the Talent introduction project of Xihua University

Zitieren

APA

Xu, M., Xie, J., Li, L., Xu, Z., Kang, Y., Jia, P., Wei, Y., Tang, J., Lin, H., & Liu, P. (2026). Nondestructive Determination for Multiple Quality Indicators of Fermented Fava Bean Paste: Synergistic Fusion Strategies for Combining the NIR and Hyperspectral Information. Journal of Food Science, 91(5). https://doi.org/10.1111/1750-3841.71142

Vancouver

Xu M, Xie J, Li L, Xu Z, Kang Y, Jia P, et al. Nondestructive Determination for Multiple Quality Indicators of Fermented Fava Bean Paste: Synergistic Fusion Strategies for Combining the NIR and Hyperspectral Information. Journal of Food Science. 2026;91(5). doi:10.1111/1750-3841.71142

BibTeX
@article{Xu2026,
  author = {Xu, Min and Xie, Junwei and Li, Lan and Xu, Zedong and Kang, Yang and Jia, Pengfei and Wei, Yuzhen and Tang, Jie and Lin, Hongbin and Liu, Ping},
  title = {Nondestructive Determination for Multiple Quality Indicators of Fermented Fava Bean Paste: Synergistic Fusion Strategies for Combining the NIR and Hyperspectral Information},
  journal = {Journal of Food Science},
  year = {2026},
  volume = {91},
  number = {5},
  doi = {10.1111/1750-3841.71142}
}
Kostenlos starten →