Non-destructive determination of β-carotene content in mango by near-infrared spectroscopy compared with colorimetric measurements

• β-Carotene content in mango can be estimated by b* and h° colour values of the flesh.
• High accuracy prediction of β-carotene is possible by near-infrared spectroscopy.
• A multiple linear regression model from selected wavelengths shows high potential to predict β-carotene.
• Effects of spectral pretreatment and wavelength selection are presented.

Non-destructive applications for the detection of food quality, especially internal properties, are highly relevant for process control in the food industry. In this respect, colour measurement and near-infrared spectroscopy (NIRS) were evaluated and compared for their ability to predict β-carotene content in mango cv. ‘Nam Dokmai’. Colorimetric analysis of peel and flesh colour as well as NIR measurements in the short- (700–1100 nm) and long-wave regions (1000–2500 nm) were analysed for prediction ability. It was found that β-carotene content could be estimated by multiple linear regression (MLR) models developed from b* and hue angle (h°) values of the flesh with good results for coefficient of determination (R2) and standard errors of cross validation (SECV) of 0.941 and 10.2 retinol equivalents (RE) 100 g−1 edible part (EP), respectively, while peel colour showed poor results. However, flesh colour measurement is a destructive method. NIRS calibration showed good results with R2 > 0.800 and standard error of prediction (SEP) 11.642–20.2 RE 100 g−1 EP. Long-wave NIR provided better prediction ability than short-wave. From these results, NIRS can be recommended for non-destructive and reliable determination of β-carotene content in mango. The results have implications for quality control in the industrial handling and processing of fruits.