The aim of this research was to detect adulteration in palm sugar by coconut sugar using FT-NIR spectroscopy with two chemometric methods namely partial least squares regression (PLSR) and principal component analysis (PCA). The absorbance spectra were taken using the FT-NIRFlex-500 Solid. Several spectral pre-processing methods used were the 1st Savitzky-Golay Derivative, Normalization, Standard Normal Variate (SNV), Multiplicative Scatter Correction (MSC), and Baseline. Coconut sugar as adulterant with various concentration ranging from 0 to 100% were added to the palm sugar. A total of 77 spectra of pure and adulterated palm sugar samples were divided into two groups in which 51 samples used for developing calibration model and 26 samples used for developing prediction model. The spectral obtained were pre-processed and analyzed using The Unscrambler X version 10.4. The best transformation of PLSR was MSC with coefficient of determination (Rc2) of 0.93 and the root mean square error (RMSE) of 0.07 for calibration. By using prediction data sets, the model resulted in coefficient of determination of prediction (Rp2) of 0.91 and a root mean square error of prediction (RMSEP) of 0.09. Based on this result, FT-NIR spectroscopy combined with chemometrics is a promising method in food authentication.

No abstract

Determination of pH of intact tomatoes was investigated using a low-cost Vis/NIR spectroscopy in reflectance mode. The best calibration model measured pH in intact tomatoes using wavelength range of 527-799 nm with RC 2 and SEC of 0.90 and 0.04, respectively. The prediction model obtained SEP, Bias, and RPD of 0.11, 0.007, and 1.17, respectively. The low-cost Vis/NIR type instrument is promising to be used for food and agricultural applications.

Soluble solids content (SSC) is one of the most important parameters of banana associated with taste and consumer acceptance. NIR spectroscopy has been applied for nondestructive determination of SSC, but limited studies were conducted for a low-cost and modular VIS/NIR spectroscopy. This study was conducted to develop a calibration model to predict SSC in bananas using a modular type of VIS/NIR spectroscopy in the range of 350-1000 nm by varying distances of fiber optic probe to samples. Two varieties of bananas, namely Musa acuminata × balbisiana and Musa acuminata ‘Lady Finger’ were used. Partial least square regression (PLSR) was used to build a calibration model and to predict SSC of bananas. Normalization, baseline correction, standard normal variate (SNV), and multiple scattered (MSC) correlation were used for spectra preprocessing. The research showed that using 2 cm probe-sample distance and SNV method resulted in the best model with the coefficient correlation of calibration ( R G 2 ) and prediction ( R P 2 ) of 0.95 and 0.87, respectively. This study proved that probe-sample distances affected the efficiency of the model for VIS/NIR spectroscopy. This work concluded that the low-cost modular VIS/NIR spectroscopy is a promising tool for SSC measurement.

Determination of soluble solid content (SSC) and pH of banana was investigated using a modular Vis/NIR spectroscopy in reflectance mode. Vis/NIR spectroscopy has been applied for non-destructive SSC or pH measurement, but limited studies were conducted for a modular VIS/NIR spectroscopy. This study was conducted to develop a calibration model to predict SSC and pH in bananas using a modular type of VIS/NIR spectroscopy at wavelength of 350-1000 nm. Two chemometrics analysis namely partial least square (PLS) and principle component regression (PCR) were used to develop calibration models and to predict SSC and pH of bananas. Normalization, baseline correction, standard normal variate (SNV), and multiplicative scatter correction (MSC) pre-processing were used for spectra transformation. Research showed that PLS regression produced better models compared to PCR in determining SSC and pH contents. PLS regression resulted in RC 2 of 0.95, RMSEC of 1.27, Rp 2 of 0.85, RMSEP of 1.98, and bias of -0.09 for SSC and RC 2 of 0.96, RMSEC of 0.05, Rp 2 of 0.82, RMSEP of 0.11, and bias of 0.11 for pH. PCR resulted in RC 2 of 0.78, RMSEC of 2.63, Rp 2 of 0.76, RMSEP of 2.71, and bias of -0.12 for SSC and RC 2 of 0.71, RMSEC of 0.14, Rp 2 of 0.62, RMSEP of 0.16, and bias of -0.02 for pH. This modular Vis/NIR instrument combined with proper pre-processing method and chemometric analysis is promising to be used for determination of SSC and pH of fruits.

Porang (Amorphophallus muelleri) is a tuber plant used in the industrial production of glucomannan. Porang can be cultivated with bulbil (katak tubers) and spontaneously harvested vegetatively. Under certain harvest and storage conditions, infected bulbils can reduce seed quality, viability, and growth. This study developed a mathematical model to predict the vegetative growth of porang with different bulbil seed qualities. The physical properties and plant growth parameters of 90 seeds of non-infected and infected bulbil were measured and divided into the calibration set comprised of 60 samples. The validation set included the remaining 30 samples. Polynomial models were used to predict plant height which was developed based on the calibration set. In contrast, model evaluation was used Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Means Absolute Percentage Error (MAPE) were used to compare the calibration and validation set with the prediction data of plant height. The mathematical growth model using the polynomial model yielded the coefficient of determination (R 2 ) of 0.9811 and 0.9824, with RMSE 0.4680 and 2.1504, MAE 0.0111 and 0.6694, and MAPE 0.8661 and 5.3096, for the calibration and validation sets of noninfected seed, respectively. Infected bulbil produced R 2 of 0.9931 and 0.9926, with RMSE 0.5641 and 4.7765, MAE 0.0549 and 1.5163, and MAPE 3.4109 and 4.5561, for the calibration and validation sets, respectively. The R 2 and the validation model showed that the mathematical model was feasible to predict the growth of non-infected and infected porang bulbil to assess the vegetative growth of porang seeds of different qualities.

Fermentation is critical in cacao processing which breakdowns sugar compounds in the pulp into organic acids. The produced organic acids stimulate enzymatic reactions in the beans which affect the flavor, taste, and color of the cacao beans. Acidity (pH) and moisture content of cacao beans influence the quality of cacao during fermentation. Those parameters are commonly measured using pH meter and gravimetric methods which are time-consuming and destructive.Therefore, the objective of this study was to develop non-destructive calibration models to predict the pH and moisture content of cacao beans at various fermentation levels using a visible near-infrared spectrometer and chemometric methods. The samples consisted of 315 cacao beans obtained from 3 regions (Lampung, Makasar, and Kulon Progo) at 3 levels of fermentation (non, half, full). The calibration and prediction models were performed by PLS regression which involves X variables (Vis/NIR measurement results) and Y variable (pH and moisture content). Smoothing, normalization baseline correction, standard normal variate (SNV), and multiplicative scattered correction (MSC) were used for spectra pre-processing. The research showed that the MSC method resulted in the best model for pH with the correlation coefficient of calibration ( R C 2 ) and prediction ( R P 2 ) were 0.76 and 0.68, respectively. Original and MSC methods resulted in the best model for moisture content with the same value of R C 2 of 0.55 and R P 2 of 0.41. The results showed the capability of Vis/NIR spectroscopy and the important role of chemometrics in developing models for predicting cacao bean quality parameters during fermentation.

Fermentation is an important process in determining the quality of cocoa beans. Therefore, studies to determine the rate of fermentation of cocoa beans non-destructively and rapidly are needed. This study aimed to develop a robust model using Vis-NIR and PLS-DA to differentiate cocoa bean samples with different fermentation levels. The PLS-DA calibration and prediction classification model showed reliability (Rel) values above 75% and 73%. Accuracy (Acc) of calibration and prediction were higher than 90% and 87%, respectively. A reliable and robust model has been created, which allows determining the rate of fermentation of cocoa beans non-destructively and rapidly.