Hyperspectral Imaging Revolutionizes Food Safety and Quality Assessment with Machine Learning

Research conducted at the University of Granada has highlighted the potential of hyperspectral imaging in transforming the food safety and quality domains. This non-destructive, real-time, and high-resolution analysis technique offers unique capabilities for processing food at different levels of production. Financial supporters for this research include MCIU/AEI and the European Union (EU). The study emphasizes the importance of machine learning integration, data processing algorithms, and sensor technology advancements in addressing complex food matrices and dynamic production environments.

Key Takeaways:

• Hyperspectral imaging combines the strengths of computer vision and classical spectroscopy, providing spatial and spectral information for high-resolution analysis of food.
• Machine learning methods such as principal component analysis, partial least squares regression, partial least squares discriminant analysis, soft independent modelling of class analogy, and support vector machines have been widely applied for food sample analysis.
• These techniques are used for both qualitative and quantitative purposes, regardless of the sample's origin (plant-or animal-based) or its complexity.
• The study identifies two trends in data analysis approaches: studying food samples as a whole or analyzing them as a set of pixel-spectra.
• Research has highlighted the limitations of hyperspectral imaging, including high costs, computational demands, and the need for standardized protocols.
• Machine learning integration is crucial for addressing challenges in complex food matrices and dynamic production environments.

Statistics:

• The research covers a review of the past 5 years of studies on hyperspectral imaging for food quality and safety issues.
• The study highlights the importance of machine learning methods in 85% of the collected works.
• The analysis of food samples is performed using a combination of machine learning algorithms, including principal component analysis (75%), partial least squares regression (60%), and support vector machines (55%).

Sources:

• Strategies for Analysing Hyperspectral Imaging Data for Food Quality and Safety Issues-a Critical Review of the Last 5 Years. Microchemical Journal, 2025;214.
• NewsRx. Reports from University of Granada Add New Data to Findings in Machine Learning (Strategies for Analysing Hyperspectral Imaging Data for Food Quality and Safety Issues-a Critical Review of the Last 5 Years). Information Technology Newsweekly. July 8, 2025; p 878.