Classifying Cannabis Sativa Flowers, Stems and Leaves using Statistical Machine Learning with Near-Infrared Hyperspectral Reflectance Imaging

Several jurisdictions around the world have now legalized the production of medicinal cannabis products. Consequently, Cannabis sativa start to emerge as an economically significant agricultural crop. Hyperspectral sensing tools are employed in agriculture to gather just-in-time non-contact and non-destructive health and growth information in high-value crop production. Unfortunately, no data or reports can be found in the literature on the use of hyperspectral sensing to monitor and evaluate the Cannabis sativa plants. This paper investigates the use of hyperspectral near-infrared imaging to identify the Cannabis sativa plant components such as flowers, stems and leaves on the crop. This is the first step towards developing a real-time monitoring system that would support decision-making on the optimal daily growing conditions to improve crop yield and profitability.