[Detection of onion soluble solids content based on the near-infrared reflectance spectra]

Area of Science:

• Agricultural Science
• Analytical Chemistry
• Spectroscopy

Background:

• Onion quality is significantly influenced by soluble solids content (SSC).
• Accurate and non-destructive methods for SSC determination are crucial for the agricultural industry.
• Traditional refractometry requires sample preparation, limiting high-throughput analysis.

Purpose of the Study:

• To evaluate the feasibility of near-infrared (NIR) reflectance spectroscopy for quantifying onion SSC.
• To compare different spectral pre-processing techniques for optimizing prediction models.
• To develop a robust statistical model for SSC determination in onions.

Main Methods:

• NIR reflectance spectra (924-1720 nm) were collected from 268 onion samples across three cultivars and harvest periods.
• Soluble solids content (SSC) was measured using a temperature-compensated refractometer as the reference method.
• Partial Least Squares Regression (PLSR) was employed to build predictive models, with pre-processing methods including Savitzky-Golay (S-G) smoothing and scatter correction.

Main Results:

• Savitzky-Golay smoothing (window 32, span 10) showed efficiency.
• Spectra scatter correction yielded improved prediction results (R² = 0.88, RMSEP = 2.31 °Brix) compared to derivation methods.
• The optimal model, utilizing crossing validation and scatter correction, achieved high prediction accuracy (R² = 0.90, RMSEP = 1.84 °Brix, RPD = 3).

Conclusions:

• Near-infrared reflectance spectroscopy is a viable, non-destructive technique for determining onion soluble solids content (SSC).
• Spectra scatter correction pre-processing significantly enhances the accuracy of SSC prediction models.
• The developed PLSR model demonstrates high reliability for quality assessment in onions.