Early Detection of Coffee Leaf Rust Caused by Hemileia vastatrix Using Multispectral Images

Conventional methodology in the field for sampling of coffee leaf rust, caused by Hemileia vastatrix, has proven to be impractical. This paper proposes a method early detection this disease, which is most significant pathogen plants worldwide, using multispectral images acquired Mapir Survey3W camera and an unmanned aerial vehicle (UAV). For purpose, 160 seedlings cultivar ‘Mundo Novo’ were inoculated with urediniospores H. vastatrix compared control (non-inoculated) determine favorable interval distinguishing healthy infected plants. The 320 placed on dark surface perform imaging flights. In vitro analyses physiological parameters 20 specimens then performed each condition (inoculated/non-inoculated) obtain hyperspectral curves, process was repeated three times at 15, 30, 45 days after inoculation (DAI). Based simulated discrepancy between red near-infrared (NIR) bands identified 15 DAI, showing greater absorption band spectral response NIR band. Thus, able distinguish infection asymptomatic stage (15 DAI) support vector machines (SVM) algorithm. Detection accuracy 80% Kappa index agreement moderate (0.6). low-cost technology can important tool monitoring rust and, consequently, more sustainable management pathogen, causing farmers make applications chemical fungicides only when necessary.