Non-Destructive Soluble Solids Content Determination for 'Rocha' Pear Based on VIS-SWNIR Spectroscopy under 'Real...

Area of Science:

• Agricultural Science
• Spectroscopy
• Horticulture

Background:

• Accurate soluble solids content (SSC) determination is crucial for fruit quality assessment.
• Existing methods for SSC measurement can be time-consuming and destructive.
• Non-destructive methods are needed for efficient fruit sorting in packinghouses.

Purpose of the Study:

• To develop and validate a non-destructive method for determining the SSC of 'Rocha' pears using short-wave NIR reflectance spectra.
• To identify the optimal spectral analysis pipeline, including pre-processing and multivariate regression models.
• To evaluate the impact of fruit temperature and size on SSC prediction accuracy.

Main Methods:

• Acquired 3300 short-wave NIR reflectance spectra (500-1100 nm) from 'Rocha' pears under simulated packinghouse conditions.
• Implemented a computational pipeline with pre-processing, feature selection, and multivariate regression models (Support Vector Machines).
• Utilized internal and external validation strategies to assess prediction accuracy, incorporating fruit temperature and size data.

Main Results:

• Support Vector Machines (SVM) demonstrated the best prediction performance, achieving RMSEP values of ~0.82 °Brix (internal) and ~1.09 °Brix (external validation).
• External validation indicated good prediction accuracy under 'real-world-like' conditions.
• Incorporating fruit temperature and size significantly improved SSC predictability.

Conclusions:

• The developed NIR spectroscopic method offers a reliable and non-destructive approach for single-fruit SSC characterization of 'Rocha' pears.
• The methodology is suitable for implementation in existing packinghouse facilities for automated fruit sorting and quality control.
• This technique supports efficient and accurate quality assessment in the fruit industry.