Effect of Temperature on the Drying Kinetics of Caturra Coffee: Correlation with Hyperspectral Imaging and Sensory Quality

Coffee processing requires continuous optimization to preserve sensory quality while improving process efficiency. Although hyperspectral imaging has been widely applied for food quality evaluation, its use for predicting coffee cup score during controlled drying remains limited. This study aimed to evaluate the effect of drying temperature on the drying kinetics of Caturra coffee and to develop a predictive model for cup score using hyperspectral imaging combined with Partial Least Squares Regression (PLSR). Coffee samples were dried at four constant temperatures (30, 40, 50, and 60 °C) in forced-convection ovens, and hyperspectral reflectance images (400–1000 nm) were acquired using a Specim FX10 camera. Sensory evaluation was conducted by six certified Q Arabica Graders. Drying times were 52, 34, 30, and 20 h at 30, 40, 50, and 60 °C, respectively, with corresponding cup scores of 83.21, 83.50, 83.60, and 83.26 points. Effective moisture diffusivity ranged from 10−13 to 10−12 m2/s, while mass transfer coefficients were on the order of 10−9 m/s, with activation energies of 28.016 and 19.272 kJ/mol. No significant differences in cup score were observed among drying temperatures (p>0.05). A PLSR-based model was developed to estimate cup score from hyperspectral data, achieving R2 values of 0.770 and 0.855 and RMSE values of 0.515 and 0.518 for calibration and validation, respectively. Key wavelengths at 480, 600, 720, and 940 nm were identified as the most influential spectral regions associated with chemical compounds affecting sensory quality. These findings demonstrate the potential of integrating drying kinetics and hyperspectral imaging as a rapid and non-destructive approach for objective prediction of coffee sensory quality during processing.