Nondestructive measurement of fresh tomato lycopene content and other physicochemical characteristics using visible NIR spectroscopy

Measurement of fresh tomato fruit overall quality, and particularly lycopene content, is challenging in the context of high-volume production. An experiment was conducted to simultaneously measure various quality parameters of tomato in a nondestructive manner using vis-NIR reflectance spectroscopy and chemometrics. The sampling set included different cultivars that are obtainable from both retailers' shelves and two greenhouse producers. Results indicate that lycopene content was accurately predicted [r2 = 0.98; root mean square error of cross-validation (RMSECV) = 3.15 mg/kg], along with color variables such as Hunter (r2 = 0.98), L, and b (2 = 0.92). Tomato color index (TCI) was better predicted (r2 = 0.96) than the a/b ratio (r2 = 0.89). Firmness prediction, with an r2 of 0.75, is comparable to what is reported in the literature for other fruits and may have a practical interest. Prediction of internal quality such as pH, soluble solids, titratable acidity, and electrical conductivity was less accurate, partly due to a low variability of these parameters among samples. Predictions were robust with regard to cultivars, except for pink variety tomato. The 400-1000 nm range gave results almost as accurate as the 400-1500 nm range. © 2008 American Chemical Society.