Evaluation of Carrot Quality Using Visible-Near Infrared Spectroscopy and Multivariate Analysis

Carrot has a relatively high content of Vitamin C and it is a major source of natural carotenoids. However, carrot has a short shelf-life and is better consumed fresh. A quick assessment of its quality attributes is important to preserving its freshness. The objective of this study was to apply Vis-NIR spectroscopy to noninvasively assess and predict the various quality attributes of carrot (cv. Nectar), namely color (L*a*b*), moisture content (MC), total soluble solids (TSS), firmness, Vitamin C, and β-carotene. Two spectroscopic sensors (400-1,000 nm and 900-1,700 nm) were utilized and samples included whole root and 25.4 mm thick sliced disc. The best prediction models using partial least squares regression yielded correlation coefficient, r, and ratio of performance to deviation or r(RPD) of 0.50(0.73), 0.84(0.88), 0.86(2.07), 0.69(0.66), 0.97(1.44), 0.90(1.49), 0.47(1.47), and 0.92(1.76) for color indices, L* a* b*, firmness, MC, TSS, Vitamin C, and β-carotene, respectively. However, using only the wavelengths selected by interval partial least squares, the r(RPD) values for the aforementioned attributes improved and are presented as follows: 0.92(1.97), 0.96(2.83), 0.98(5.85), 0.99(6.65), 0.98(3.91), 0.99(5.93), 0.98(4.16), and 0.98(4.43), respectively. Generally, Vis-NIR region had higher prediction performance than NIR region, and whole roots had similar prediction performance as sliced samples. This study shows that rapid determination of quality parameters of carrot is possible through non-destructive Vis-NIR sensing, which could be useful for quality tracking during carrot supply chain. Moreover, results of this study could be improved using a larger sample size.