In this paper (open access) the authors propose a chemometric workflow for verifying rice varieties.
Authentic rice varieties were sourced from Ente Nazionale Risi, a public economic body supervised by the Italian Ministry of Agriculture, Food Sovereignty and Forestry, These comprised 36 Italian plus varieties from Sri Lanka, Malaysia, Bangladesh, Philippines, Indonesia, Taiwan, Australia, Brazil, India, and Louisiana. They were characterised by their amylase (“high”, “low” and “waxy”) and by their grain morphology (“Round”, “Medium”, and 3 variations of “Long”).
The researchers built classification models based on both Near Infra-Red Hyperspectral Imaging (NIR-HSI) and on morphological classification. They found that NIR-HSI alone gave relatively poor classification but they recommend it as a quick, cheap screening tool, with follow-up where required using a morphological classification model. They propose a morphogram tool to visualise the results. They report that this workflow gave good results, and recommend it as a routine approach that can be applied at any point where rice is traded..
Photo by Łukasz Rawa on Unsplash
Comments