Characterization of a Multispectral Camera for Abiotic and Biotic Stress Detection in Greenhouse

Nowadays, multispectral sensors are not commonly used in protected crops due to structural features and the presence of an artificial light. This study describes the application of RGB-NIR multispectral camera for greenhouse applications. Different light conditions were evaluated for their influence on the vegetation index on Rosa rugosa. Images were acquired in a darkroom at natural, and light-emitting diodes. The abiotic stress identify was evaluated on a randomized block (5 × 5) trial on 250 plants of Pelargonium spp. LAI and VARI indexes showed to be suitable to highlight the growth stages of plants. Finally, the spreading of an infection was evaluated on Mandevilla spp. The highest correlation was found using image analysis indexes. According to these results, RGB-NIR cameras can be profitably used in the greenhouse environment by I) correcting the incident light, II) combining pixel classification.