Wavelength selection for predicting physicochemical properties of apple fruit based on near-infrared spectroscopy

Instrumental evaluation tools for fruit quality monitoring are important in the production and postharvest processes as well as in marketing. In the present study, near-infrared spectroscopy (600-1,100 nm) was applied to study the correlation with fruit soluble solid content (SSC ), fruit flesh firmness and water content of apples (cv. "Fuji"). Genetic algorithm and correlation coefficient (r) method were used to select the most sensitive wavelength combinations, and partial least squares regression analysis was applied to calibrate fruit quality parameter. The validation of models based on the most sensitive wavelengths gave good predictions with an r value of 0.94 and a standard error of cross validation (SECV) of 0.85 degrees Brix for SSC; r = 0.89 and SECV = 7.54 N/cm(2) for firmness; and r = 0.96 and SECV = 0.92% for water content. The reduced data set of sensitive wavelengths were found feasible for predicting internal fruit quality.