Usability of NIR Hyperspectral Imaging for Evaluating Added Sugar Solution in Pineapple Juice

Abstract

Economically motivated adulteration of pineapple juice with sugar can be successfully detected by chemical analysis, but this is expensive and time consuming. Consequently, this research focused on testing whether the application of near-infrared (NIR) hyperspectral imaging could be successfully used to detect added sugar in pineapple juice. A dataset comprising 149 samples of pure concentrated pineapple juice and 149 samples of concentrated pineapple juice with added sugar solution in various concentrations, ranging from 0.5% to 99.5% w/w, was carried out. The data set of samples was divided into a calibration set and a prediction set for both quantitative and qualitative analyses. Partial least squares regression (PLSR) was employed for establishing the model for detecting the ratio of added sugar in the concentrated pineapple juice, achieving a correlation coefficient (R) of 0.985 with a root mean square error of prediction (RMSEP) of 4.971% w/w. Predictive images using colors was used to present added sugar in concentrated pineapple juice. The principal component analysis (PCA) showed the spectral information was able to be used to classify the group of pure concentrated pineapple juice and adulterated concentrated pineapple juice with the cumulative variance percentage of 98% for the first principal component (PC1) and the second principal component (PC2).  Partial least squares-discriminant analysis (PLS-DA) was also employed for classifying the group of pure concentrated pineapple juice and adulterated concentrated pineapple juice, achieving an impressive total accuracy of 98.98% for the prediction set. It was concluded that this study demonstrated that NIR hyperspectral imaging could be used to detect added sugar adulteration of concentrated pineapple juice.