Comparison of different CCD detectors and chemometrics for predicting total anthocyanin content and antioxidant activity of mulberry fruit using visible and near infrared hyperspectral imaging technique.

This study investigated the potential of using hyperspectral imaging technique in tandem with chemometrics for rapid and invasive predicting total anthocyanin content and antioxidant activity of mulberry fruit. Two calibration methods of partial least square regression and least-squares support vector machines and three wavelength selection algorithms of successive projections algorithm, uninformation variable elimination, and competitive adaptive reweighted sampling were applied. The best prediction models for the analysis of total anthocyanin content and antioxidant activity had R of 0.959 and 0.995 respectively. The performances of two CCD detectors named silicon (Si) and indium gallium arsenide (InGaAs) were compared. The results show that hyperspectral imaging has a great potential for the assessment of total anthocyanin content and antioxidant activity of mulberry fruit.