Related Experiment Video
Updated: May 20, 2025
07:10
Untargeted Liquid Chromatography-Mass Spectrometry-Based Metabolomics Analysis of Wheat Grain
Published on: March 13, 2020
9.5K
Classification of different gluten wheat varieties based on hyperspectral preprocessing, feature screening, and
Xinghui Qi1, Shaohua Zhang1, Liyang Wang1
1Agronomy College of Henan Agriculture University/State Key Laboratory of High-Efficiency Production of Wheat-Maize Double Cropping, Zhengzhou 450046, Henan, China.
Food Chemistry: X
|March 24, 2025
Summary
Classifying wheat gluten types using hyperspectral data is crucial for food needs. The ReliefF-CWT-SVM model achieved 94.5% accuracy, offering a rapid method for wheat variety classification.
Area of Science:
- Agricultural Science
- Spectroscopy
- Machine Learning
Background:
- Accurate classification of wheat gluten varieties is essential for meeting diverse food industry demands.
- Hyperspectral data offers a promising avenue for rapid and non-destructive analysis of wheat characteristics.
Purpose of the Study:
- To develop and optimize a rapid classification model for wheat gluten types using hyperspectral data.
- To compare the effectiveness of various data preprocessing, feature screening, and machine learning methods.
Main Methods:
- Four preprocessing techniques including continuous wavelet transform (CWT) were evaluated.
- Two feature screening methods, ReliefF and minimum redundancy maximum relevance (mRMR), were applied.
- Four machine learning algorithms, Support Vector Machine (SVM), Convolutional Neural Network (CNN), Random Forest (RF), and K-nearest neighbor (KNN), were tested.
Main Results:
- The ReliefF feature screening method outperformed full wavelength and mRMR methods.
- Continuous Wavelet Transform (CWT) preprocessing yielded higher accuracy than other methods.
- The Support Vector Machine (SVM) classifier demonstrated superior performance among the tested algorithms.
- The optimal model, ReliefF-CWT-SVM, achieved an overall accuracy of 94.5%.
Conclusions:
- The combination of ReliefF feature extraction and CWT preprocessing with SVM classification provides a highly accurate method for classifying wheat gluten types.
- This developed model offers significant theoretical and technical support for the rapid identification of wheat varieties based on gluten characteristics.
