Abstract:
The aim of this study was to examine the possibility of non-destructive estimation of lycopene and β-carotene
content in intact tomato fruit. Representative tomato fruits were harvested at different maturity stages and
measurement of visible/near infrared (VIS/NIR) spectra and color variables were followed by reference analysis
of lycopene and β-carotene. Models predicting lycopene and β-carotene from color variables and VIS/NIR
spectra of intact tomato were developed. Regression coefficient between color variables (a*, a*/b*, and (a*/
b*)2
) and reference values of lycopene and β-carotene content were (0.90, 0.98 and 0.52) and (0.75, 0.81 and
0.58), respectively for the prediction set. Meanwhile, after observing predictive p-values in multiple regression,
best equations were developed to predict the contents of lycopene and β-carotene from color values with regression coefficient of 0.97 and 0.85 for lycopene and β-carotene, respectively. On the other hand, the selected
partial least square (PLS) model of VIS/NIR spectra had good predictive power for lycopene and β-carotene
showing high correlation coefficient of 0.85 and 0.77, respectively, between measured and predicted samples.
This study revealed that, estimation of the lycopene and β-carotene content in tomatoes could be achieved by a
portable chroma meter and VIS/NIR spectroscopy, with a possible application at field and agricultural processing centers, respectively
