Avaliação do potencial de índices de vegetação para detecção de doença na cana-de-açúcar em imagens hiperespectrais adquiridas por VANT
Technological innovations from Precision Agriculture have contributed the agronomic development optimizing and automating their activities bringing benefits in agricultural production and the environment. The purpose of this study was to evaluate the potential of vegetation indices in hyperspectral images taken by unmanned aerial vehicle (UAV) for detection sugarcane mosaic. Indexes that brought a major contribution to the detection of regions infected with sugarcane mosaic were Normalized difference Vegetation Index (NDVI), Normalized difference Vegetation Index Red Edge (NDVI705), New Vegetation Index (NVI), and Anthocyanin Index Reflectance 2 (ARI2). All indices analyzed in this case study, except the Photochemical Reflectance Index (PRI), Carotenoid reflectance Index 2 (CRI2), and REP were not efficient to show bare soil areas. The PRI, Carotenoid reflectance Index 1 (CRI1) and CRI2 and especially Modified Chlorophyll Absorption Ratio Index (MCARI) showed the noises of hyperspectral imaging, Transformed chlorophyll absorption reflectance index (TCARI) and Red Edge Position Determination (REP) did not present information related to the presence of mosaic in sugarcane. However, among the indexes, analyzed ARI2 was able to highlight areas infected with mosaic; this index is associated with the concentration of anthocyanin in vegetation, the anthocyanin may be indicative of senescence and stress in some species of plants. The importance of such research is justified by its contribution to agriculture with the generation of a methodology that enable the use of sensors designed to take images in appropriate spectral bands to monitoring crops, and assist in agricultural management practices using a approach that cause less environmental impact.