Prediction of Ripening Stage and Key Physicochemical Quality Parameters in Red-Fleshed Kiwifruits using VNIR Spectroscopy

초록

Red-fleshed kiwifruit has attracted increasing consumer interest, but cultivar-dependent ripening and conventional destructive measurements limit routine eating-quality assessment. This study developed visible-near infrared (VNIR) spectroscopy-based prediction models for key maturity indices in two commercially important red kiwifruit cultivars, ‘Hongyang’ and ‘Qihong’. Fruit were monitored during ripening at ambient conditions to generate a wide range of maturity stages, and an independent set was stored under cold conditions for external evaluation. VNIR spectra (400－1100 nm) were acquired from four orientations per fruit and combined with reference measurements of firmness, soluble solids content (SSC), titratable acidity (TA), SSC/TA ratio, and internal flesh color parameters (L*, a*, b*). Several preprocessing strategies (mean normalization, SNV, MSC, and Savitzky-Golay derivatives) were compared, and partial least squares regression (PLSR) models were developed and evaluated using calibration, cross- validation, and prediction statistics. Ripening parameters showed clear statistical relationships, with firmness negatively correlated with SSC and TA inversely associated with the SSC/TA ratio, indicating that multiple correlated traits can be combined to better characterize ripening status. PLSR models provided reliable prediction for firmness and SSC in both cultivars, whereas TA and SSC/TA ratio exhibited reduced robustness. Among color variables, L* showed the most stable prediction performance, while a* and b* were more cultivar-dependent. Overall, VNIR spectroscopy coupled with PLSR enables rapid, non-destructive estimation of maturity-related quality attributes in red-fleshed kiwifruit, with cultivar-specific calibration improving model stability and supporting practical application for postharvest monitoring and grading.

키워드