EUVIMEDEUVIMED entdecken
Crossref

Research on wheat impurity identification method based on terahertz imaging technology

Guangming Li, Hongyi Ge, Yuying Jiang, Yuan Zhang, Mengdie Jiang, Xixi Wen, Qingcheng Sun

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy · 2025

DOI: 10.1016/j.saa.2024.125205 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 2025. Verfasst von Guangming Li und 6 weiteren Autor:innen. Beteiligte Einrichtung: Key Laboratory of Grain Information Processing and Control (Henan University of Technology), Ministry of Education, Zhengzhou 450001, China; Henan Pr…. Hauptthemen laut MEDLINE: Triticum, Algorithms, Terahertz Imaging. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 25 Literaturangaben.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The traditional detection of impurities in wheat has difficulties such as low precision, time-consuming, and cumbersome, therefore, it is important to study the method of rapid and accurate detection of impurities in wheat for correctly assessing the quality grade of wheat. Terahertz (THz) technology has many superior properties such as transient, broadband, low-energy, and penetrating, which can realize rapid and nondestructive detection of wheat quality. In this study, a classification and recognition algorithm AHA-RetinaNet-X for wheat impurity terahertz images based on RetinaNet and Artificial hummingbird algorithm (AHA) is proposed.A THz three-dimensional tomography imaging system is used to image wheat and its impurities, and two THz image datasets, respectively the wheat and impurity dataset for verifying the classification effect of wheat and impurities and the impurity dataset for verifying the classification effect of impurities. The experimental results show that the AHA-RetinaNet-X model outperforms other detection and classification models in terms of accuracy, F1-score, precision, recall, and specificity, and is able to achieve 96.1%, 94.9%, 95.2%, 95.8%, 95.5%, 95.3%, and 93.3% for the wheat and impurity dataset and the impurity dataset, respectively, 95.6%, 96.3%, and 95.2%, and the mAP value of AHA-RetinaNet-X is also higher than the other models and can reach 92.1%. The combination of THz imaging technology and AHA-RetinaNet-X can realize the classification and identification of wheat and impurities, which provides a new method for the non-contact rapid nondestructive detection and identification of wheat and impurities, and also provides a reference for the research of the identification and detection methods of other substances.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

TriticumAlgorithmsTerahertz ImagingFood Contamination

Stichwörter der Autor:innen

Artificial Hummingbird AlgorithmRetinaNet networkTarget detectionTerahertz imagingWheat impurities

Autor:innen und Institutionen

  • Key Laboratory of Grain Information Processing and Control (Henan University of Technology), Ministry of Education, Zhengzhou 450001, China; Henan Provincial Key Laboratory of Grain Photoelectric Detection and Control,…

Zitieren

APA

Li, G., Ge, H., Jiang, Y., Zhang, Y., Jiang, M., Wen, X., & Sun, Q. (2025). Research on wheat impurity identification method based on terahertz imaging technology. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy. https://doi.org/10.1016/j.saa.2024.125205

Vancouver

Li G, Ge H, Jiang Y, Zhang Y, Jiang M, Wen X, et al. Research on wheat impurity identification method based on terahertz imaging technology. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy. 2025. doi:10.1016/j.saa.2024.125205

BibTeX
@article{Li2025,
  author = {Li, Guangming and Ge, Hongyi and Jiang, Yuying and Zhang, Yuan and Jiang, Mengdie and Wen, Xixi and Sun, Qingcheng},
  title = {Research on wheat impurity identification method based on terahertz imaging technology},
  journal = {Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy},
  year = {2025},
  doi = {10.1016/j.saa.2024.125205}
}
Kostenlos starten →