Estimation of Soluble Solids for Stone Fruit Varieties Based on Near-Infrared Spectra Using Machine Learning...

Area of Science:

• Agricultural Science
• Spectroscopy
• Machine Learning

Background:

• Fruit quality control is crucial for packaging and trade.
• Soluble Solids (SS) content is a key indicator of fruit ripeness and consumer acceptability.
• Non-destructive Visible and Near-Infrared (VIS-NIR) spectroscopy is a popular technique for fruit quality assessment.

Purpose of the Study:

• To develop accurate, non-destructive methods for evaluating soluble solids (SS) in stone fruits.
• To improve fruit maturity inspection using VIS-NIR spectra and machine learning.
• To classify fruit species and estimate SS content for varieties of peaches, nectarines, and plums.

Main Methods:

• A Convolutional Neural Network (CNN) was used for fruit species classification.
• A Feedforward Neural Network (FNN) was employed to estimate SS content from VIS-NIR spectra.
• The models were trained and validated on VIS-NIR spectral data from stone fruits.

Main Results:

• The CNN model achieved a high classification accuracy of 98.9% for fruit species.
• The FNN models demonstrated a strong correlation (Rc > 0.7109) for estimating SS content.
• The combined approach showed significant potential for on-line fruit classification and SS estimation.

Conclusions:

• The proposed VIS-NIR spectroscopy combined with CNN and FNN offers a fast and accurate method for non-destructive fruit quality assessment.
• This technique can be applied for on-line sorting and quality control in the fruit industry.
• The study highlights the effectiveness of machine learning in analyzing spectral data for precise fruit maturity evaluation.