Epsilon Archive for Student Projects

Abstract

Cheese fraud, which include mislabelling and adulteration, is prevalent, particularly in high-value cheeses such as PDO and PGI cheeses. Novel imaging techniques such as infrared (IR) spectroscopy and Raman spectroscopy has emerged as potentially non-destructive, fast, and accurate ways to authenticate cheeses. These techniques, coupled with chemometric analysis, have been used to gain detailed insights into the chemical and physical properties of cheese. They have also been improved to analyse spatial information, with near-infrared hyperspectral imaging (NIR-HSI) and spatial Raman. Despite their potential, widespread industrial adaption is limited.
This essay explores the application of IR and Raman spectroscopy, coupled with spatial techniques, in cheese authentication. It evaluates their advantages, like being able to analyse samples through their packaging, as well as its disadvantages, such as expensive instruments and complicated data analysis. Necessary steps towards implementing the techniques in industry is integrating advanced analytic tools, such as effective machine learning algorithms, and creating a large public dataset to improve the accuracy and applicability of these imaging techniques for industrial usage.