<![CDATA[A simple, portable visible reflectance spectrophotometer was developed for classifying the quality of Curcuma xanthorrhiza Roxb. The device was assembled using consumer electronic components: an LED as the light source, a DVD optical layer as the diffraction grating, and a webcam as the detector, with web-based software (Spectral Workbench) as the signal processor. The spectrophotometer's performance was evaluated using standard Sudan III samples and powdered C. xanthorrhiza samples from two quality classes (A and B). Spectral data were analysed using principal component analysis (PCA) and partial least squares (PLS) regression. PCA successfully grouped Sudan III samples by concentration and classified C. xanthorrhiza by curcuminoid content. A PLS regression model was developed for quantifying Sudan III, yielding R² values of 0.9909 for both calibration and cross-validation, with RMSEs of 0.3821% and 0.4152%, respectively. The results closely matched those from a commercial spectrophotometer. A PLS-discriminant analysis (PLS-DA) classification model for C. xanthorrhiza was also developed, achieving sensitivity and specificity values of 1. Additionally, semi-quantitative parameters such as decision limit (26.15% B), detection capability (41.06% B), and unreliability range (26.15–41.06% B) were calculated. The classification model showed strong sensitivity, specificity, and detection capabilities within acceptable limits. These findings suggest that this low-cost reflectance spectrophotometer, combined with chemometric methods, holds promise as a practical tool for the quality control of raw herbal materials.]]>
