Application of Portable NIR Spectroscopy for Instant Prediction of the Product Quality of Apple Slices During Hot Air–Assisted Radio Frequency Drying

Hot air–assisted radio frequency drying (HARFD) is a recently developed food dehydration method to enhance thermal efficiency while retaining the quality of the fresh produce. This investigation is aimed at exploring the potential for the application of portable near-infrared (NIR) spectroscopy for rapid real-time assessment of the quality of apple slices during HARFD. Both moisture content (MC) and vitamin C content (VCC) were selected as the critical indicators to assess the quality of the dried apple slices. Principal component regression (PCR), partial least squares regression (PLSR), and back propagation-artificial neural network (BP-ANN) models were developed and compared to establish the relationships between the NIR spectrum and the selected quality indicators of dried products. Model fitting results indicate that the BP-ANN model achieved the highest prediction accuracy for MC (lowest RMSE_P = 0.331 and highest \({R}_{P}^{2}\) $R_P^2$ =0.976) and VCC (lowest RMSE_P = 0.605 and highest \({R}_{P}^{2}\) $R_P^2$ =0.933) of apple slices during HARFD. This work highlights the use of portable NIR spectroscopy as an efficient and non-destructive tool for smart prediction of the quality parameters of apple slices during the HARFD process via optimized statistical modeling.