Sviluppo di tecniche di monitoraggio di tipo ottico-digitale nel visibile e nel vicino infrarosso mediante implementazioni di logiche chemiometriche nel settore agroalimentare

Technological innovation is revolutionizing the agri-food sector, making the development of advanced systems for quality control and crop monitoring essential. This PhD thesis focuses on integrating spectroscopic methodologies and artificial intelligence algorithms to enhance agronomic management and non-destructive crop analysis. In particular, the use of visible and near-infrared (Vis-NIR) spectroscopy, combined with machine learning and chemometric techniques, has enabled the extraction of detailed information on key parameters such as water content, nutritional status, and the presence of contaminants in agricultural products. Through a multidisciplinary approach, this study developed and validated predictive models based on spectral data, demonstrating the effectiveness of these tools in providing fast and reliable on-field analysis. The integration of these technologies with advanced optical monitoring strategies represents a significant step toward a more sustainable and data-driven agriculture, capable of optimizing resource use and reducing environmental impact. The results confirm the potential of the proposed techniques to improve the precision and efficiency of agronomic practices, offering an innovative perspective for the future of the sector. This work not only highlights the scientific and practical value of spectral analysis but also paves the way for the development of intelligent decision-support systems in the agricultural context.