Non-destructive Prediction of Internal Quality of Heat-treated ‘Irwin’ Mango by Near Infrared Spectroscopy

Abstract

Near infrared (NIR) spectroscopy in the range of 600-1000nm was studied for its ability to predict the internal quality attributes, such as moisture content, hardness, soluble solid content (SSC), acidity and pH, of the ‘Irwin’ mango (Mangifera indica Linn.) following heat treatment. A total of 63 samples were used for calibration models and 124 samples collected from 62 heat-treated mangos were used for prediction models using partial least squares regression (PLS) and principal component regression (PCR) methods. The results showed that models using the PLS method were better than those using the PCR method. There was an influence of the preprocessing spectra on the prediction performance of the models in which the second derivative of Savitzky-Golay method [D₂log(1/R)] resulted in the best calibration models. The correlation coefficients and the SEPs for the prediction of the moisture content, hardness, SSC, acidity and pH of the heat-treated mango were 0.94 (0.384), 0.87 (0.066), 0.94 (0.387), 0.80 (0.053) and 0.83 (0.141), respectively. The average SSC value in the stem-end was significantly higher than that in the blossom-end. The heat-treated mango had significantly higher moisture content, hardness, SSC and acidity levels and had lower pH levels compared to the untreated mango, showing the marked improvement of the fruit quality following the heat treatment due to a reduction in the decline of quality attributes.