Hyperspectral imaging to oversee the status of baby leaf vegetable crops: the Agrofiliere Project results

This communication reports a compendium of vegetation indices and hyperspectral band ranges relevant for monitoring the main cryptogamic diseases of wild rocket based on an overview of the recently published papers on the results of the AGROFILIERE project, aimed at studying highly integrated systems for supporting the management of baby leaf vegetable crops. In detail, the phytosanitary framework of wild rocket includes some disease taken as reference for characterizing the hyperspectral response in the reviewed studies, such as downy and powdery mildews, parenchymatic soil-borne diseases, tracheofusariosis and their exception respect to water stress. Thirteen major literature vegetable indices were selected among those that proved capable of discriminating plant health status from disease and some early stages of plant interaction with a pathogen. On the other hand, a few studies have helped to highlight the spectral ranges relevant to a specific disease in wild rocket through the application of machine learning models on hyperspectral data. This overall presentation contributes useful hyperspectral information on the whole available disease frame of wild rocket, with a view to building non-invasive and contactless digital diagnostic tools for disease detection in baby leaf vegetable cropping systems to support their control.