Application of handheld NIR spectrometer for simultaneous identification and quantification of quality parameters in intact mango fruits

Abstract

There are several varieties of mango fruits, and the most important quality indicators for determining mango maturity are pH and total soluble solids (TSS). The study examined the possibility of using a handheld NIR spectrometer (NIRS) with a wavelength range of 740 nm to 1070 nm and multivariate algorithms in combination with a smartphone to determine the varieties and maturity of mangoes. A total of 198 intact mango fruits were scanned with the NIR spectrometer, while a digital refractometer and pH meter were used to measure TSS and pH from the extracted mango juice. After using several preprocessing methods, multivariate classification models were created using support vector machine (SVM), linear discriminant analysis (LDA), random forest (RF), neural network (NN), and a new classifier (LDA-SVM) to identify the varieties. Partial-least-square regressions methods, such as; interval partial-least-square (IPLS), synergy partial-least-square (Si-PLS), and back interval-partial-least-square (Bi-PLS), were used to build quantitative models for determining TSS and pH of the fruits. Among the identification techniques, the RAW, MC, SNV, FD, and SD plus LDA-SVM could be used to identify mango fruit varieties 100% accurately in the training set and 97.44% in the prediction set. For quantification, the best model for TSS and pH measurements in mango is Si-PLS with an R2 value of 0.63, an RMSEP value of 1.83, an R2 value of 0.81, and an RMSEP value of 0.49, respectively. The study demonstrated that rapid and non-destructive assessment of TSS and pH could be achieved using handheld NIR coupled with suitable chemometric tools.