Advancing Kiwifruit Maturity Assessment: A Comparative Study of Non-Destructive Spectral Techniques and Predictive...

Area of Science:

• Agricultural Science
• Food Science
• Spectroscopy

Background:

• Accurate assessment of fruit maturity is crucial for optimal harvesting and quality.
• Non-destructive methods are preferred for real-time quality control in the fruit industry.

Purpose of the Study:

• To develop predictive models for estimating the maturity index of gold kiwifruits.
• To evaluate the effectiveness of a portable spectroradiometer and a computer vision system (CVS) for non-destructive fruit analysis.

Main Methods:

• Non-destructive measurements using a portable spectroradiometer (VIS-NIR and SWIR) and a CVS.
• Destructive analysis for reference measurements of soluble solids, sugars, and dry matter.
• Development and comparison of Partial Least Squares (PLS), Support Vector Regression (SVR), and Gaussian Process Regression (GPR) models.

Main Results:

• Sugar content (glucose and fructose) and soluble solids significantly increased with harvest time.
• The CVS could not differentiate harvests due to lack of significant skin color change.
• Hyperspectral data in the NIR and initial SWIR regions effectively predicted sugar content and soluble solids (R² between 0.55-0.60).

Conclusions:

• A portable spectroradiometer measuring up to the SWIR range can rapidly and non-destructively estimate kiwifruit maturity.
• Gaussian Process Regression (GPR) demonstrated superior performance for predicting key quality parameters.
• This technology holds potential for field applications to determine optimal harvest timing.