Commenced in January 2007
Frequency: Monthly
Edition: International
Paper Count: 72555
Chemical Image Processing for Food Safety, Classification and Authentication

Authors: Isik Riza Türkmen, Simon Friedrich, Ervinatasia Djaw


The application of chemical imaging for food safety and authentication studies has recently become a popular research topic in the food industry, as recent technological advances have led to increased computational power to run data- and computational-intensive machine learning algorithms on laboratory computers. As a result, data intensive hyperspectral image acquisition and processing has gained acceptance and prestige in the food industry, enabling a wide range of digital transformation applications that accompany various research efforts from farm to fork. We focus on two case studies: In the first, we address the application of digital image processing for sorting coffee brands based on aroma characteristics and origin. In the second application, we address the safety and authentication of minced meat by processing hyperspectral images captured in the near-infrared region of the electromagnetic spectrum. Methods are developed for the processing of hyperspectral image data that enables visual data exploration using unsupervised clustering and for the prediction of various food safety and authenticity properties based on supervised learning.

Keywords: hyperspectral imaging, food safety and authentication, unsupervised learning, supervised learning

Procedia PDF Downloads 22