Discriminating Mung Bean Origins Using Pattern Recognition Methods: A Comparative Study of Raman and NIR Spectroscopy

Area of Science:

• Agricultural Science
• Analytical Chemistry
• Spectroscopy

Background:

• Accurate determination of agricultural product origins is crucial for quality control and combating fraud.
• Spectroscopic techniques offer non-destructive methods for analyzing food products.
• Distinguishing the geographical provenance of crops like mung beans is an ongoing challenge.

Purpose of the Study:

• To compare the effectiveness of Near-Infrared (NIR) and Raman spectroscopy in determining the geographical origins of mung beans.
• To establish a robust traceability model for mung bean provenance.
• To evaluate the discriminatory power of each spectroscopic method.

Main Methods:

• Mung bean samples were collected from three distinct regions in China.
• Spectral data were acquired using both Raman and NIR spectroscopy.
• A traceability model was developed using Principal Component Analysis (PCA) and the K-nearest neighbor (KNN) algorithm.

Main Results:

• NIR spectroscopy explained 99.01% of the cumulative variance with the first three principal components, surpassing Raman spectroscopy by 6.71%.
• The discrimination rate for mung bean origins using NIR data reached 98.67%, which is 22.67% higher than that achieved with Raman spectroscopy.
• The PCA-KNN model demonstrated superior performance when applied to NIR spectral data.

Conclusions:

• Near-Infrared (NIR) spectroscopy is a more effective method than Raman spectroscopy for tracing the provenance of mung beans.
• The developed PCA-KNN model provides a reliable approach for geographical origin discrimination.
• NIR spectroscopy offers a promising tool for ensuring the authenticity and traceability of agricultural products.