Rapid and non-destructive prediction of adulterated chilli powder by employing pocket-sized NIR spectrometer and chemometrics

Chilli pepper is an economically important vegetable. This study aims to develop a non-destructive and rapid technique to detect the adulteration of chilli pepper powder with avocado pear seed and cola nut powder using handheld near-infrared spectroscopy (NIRS) in the range of 740–1070 nm. A clear cluster trend was achieved for chilli pepper and the adulterated samples by using principal component analysis (PCA). Classification models such as random forest (RF), support vector machine (SVM) and partial least square discriminant analysis (PLS-DA) together with various pre-processing methods were used to predict adulteration of chilli pepper samples. The optimum classification rate of 91.25 % was found for chilli pepper adulterated with pear seeds using MC-SVM while MC-PLSDA had 86.25 % for chilli pepper adulterated with cola nuts. The partial least square (PLS) regression model had a coefficient of determinant (R 2 ) and root mean square error of prediction (RMSEP) values of 0.9853 and 5.06 % for pear seed and 0.9724 and 6.91 % for cola nut, respectively. The current study shows that NIR spectroscopy can classify and determine chilli pepper authenticity and adulteration levels of pear seed and cola nuts.