Nondestructive Analysis of Internal Quality in Pears with a Self-Made Near-Infrared Spectrum Detector Combined with...

Area of Science:

• Agricultural Science
• Spectroscopy
• Data Analysis

Background:

• Traditional methods for assessing pear internal quality (SSC, TA, TI) are destructive, time-consuming, and environmentally unfriendly.
• There is a growing demand for rapid, non-destructive techniques to evaluate fruit quality.
• Near-infrared (NIR) spectroscopy offers a promising alternative for non-destructive quality assessment.

Purpose of the Study:

• To develop and validate a novel near-infrared (NIR) spectroscopy method for the non-destructive determination of internal quality attributes in pears.
• To introduce and evaluate the Variable Stability and Cluster Analysis Algorithm (VSCAA) for selecting optimal spectral variables.
• To compare VSCAA with other established variable selection methods (SiPLS, GA, SPA, BOSS) for improved model performance.

Main Methods:

• A self-made NIR spectrum detector was utilized for data acquisition.
• The Variable Stability and Cluster Analysis Algorithm (VSCAA) was developed and applied for infrared spectrum variable selection.
• Partial Least Squares Regression (PLSR) models were established to predict soluble solid content (SSC), titratable acidity (TA), and taste index (TI) in snow pears.
• Performance was evaluated by comparing VSCAA with SiPLS, GA, SPA, and BOSS for variable selection.

Main Results:

• The optimal PLSR model for SSC prediction, using SiPLS-VSCAA selected variables, achieved high accuracy (Rc=0.942, Rp=0.936).
• Optimal models for TA and TI prediction were achieved using GA-BOSS (TA: Rc=0.931, Rp=0.912) and GA-VSCAA (TI: Rc=0.968, Rp=0.968) respectively.
• The developed NIR system demonstrated excellent performance for rapid and non-destructive analysis of internal pear quality.

Conclusions:

• The Variable Stability and Cluster Analysis Algorithm (VSCAA) is effective for selecting informative wavelengths in NIR spectroscopy.
• The self-made NIR-spectrum detector, coupled with multivariate data processing and VSCAA, provides a robust tool for non-destructive pear quality assessment.
• This approach offers a viable alternative to traditional destructive methods, supporting efficient fruit quality management.