A new approach based on hyperspectral imaging for farming system authentication of seeds

Abstract

Seeds are an exceptional source of nutrients, so their consumption favors consumer health. In fact, the health concern has caused the market to express an increase in the consumption of organic products. In South America, between the seeds with organic production and large export volumes for Europe and the United States of America, there are Quinoa (Chenopodium quinoa) and Chia (Salvia hispanica). Quinoa is considered a superfood due to its high nutritional value, since it has high quality protein gluten-free. Chia seeds are vegetable source rich in essential fatty acids, phenolic compounds, protein and dietary fiber. To identify the farming system of different foods, isotopic techniques and high-efficiency spectrometry are traditionally used to identify specific markers (e.g. fatty acid profile). Generally, these analytical techniques are expensive, time-consuming and destructive to the sample. Therefore, despite their accuracy, they turn out to be unattractive to the food industry. Hyperspectral imaging system has several advantages over traditional analytical methods (e.g. mass spectrometry or high performance liquid chromatography), such as speed, chemical free, low cost of use and possible application for online processes for quality control and food authentication. Hyperspectral imaging provides spatial and spectral information, which is subject to multivariate analysis, to generate mathematical models capable of discriminating between foods based on their differences in chemical composition. For these reasons, this project aims to develop mathematical models based on information obtained from hyperspectral imaging to identify the farming system of Quinoa and Chia seeds. Thus, a new non-invasive and non-destructive authentication method for these seeds will be created and validated, which will undoubtedly favor the traceability of these products. (AU)