Abstract:
Handheld near-infrared spectroscopy (NIRS) is emerging as a key technology for food analysis in Africa. This study explores its effectiveness, combined with chemometric techniques, for rapid and non-destructive evaluation of mango fruits and products. It focuses on developing predictive models for variety differentiation, classification of organic and inorganic samples, and assessing quality attributes such as total soluble solids (TSS) and pH. Additionally, it identifies ethephon residues and categorizes organic and inorganic pineapple juice. The study also examines the physicochemical and microbial changes in expired and unexpired commercial fruit juices using conventional laboratory methods. For mango variety identification, NIRS combined with multivariate algorithms achieved 97.44% accuracy. The synergy partial least squares model yielded r² values of 0.63 and 0.81 for TSS and pH predictions, with RMSEP values of 1.83 and 0.49, respectively. In detecting ethephon residues, the neural network model with multiplicative scatter correction reached 100% classification accuracy, while the partial least squares model demonstrated strong predictive performance (r² = 0.996, RMSEP = 0.068). The random forest algorithm classified organic and inorganic mango products with varying accuracy levels. When preprocessed using the second derivative, it achieved 88.76% accuracy for fresh fruit, 77.98% for chips, and 87.53% for juice without preprocessing. The combination of dual NIR spectrometers effectively distinguished organic and inorganic pineapple juice with 100% accuracy. Furthermore, a comparative assessment of expired and unexpired commercial fruit juices showed notable declines in titratable acidity (apple juice decreased from 0.60% to 0.12%) and vitamin C (a 57.6% reduction in pineapple juice), alongside an increase in microbial load. These findings highlight the potential of handheld NIRS as a reliable tool for quality control, product authentication, and food safety assurance. Its application could improve postharvest monitoring, mitigate food fraud, and enhance regulatory compliance within the fruit industry.
