ABSTRACT
Traditionally, the severity of diseases, such as Asian rust, is assessed through visual inspections of infected tissues based on diagrammatic scales. However, these methods involve high costs and subjective analyses due to potential errors by the evaluators. Thus, the aim of the present study was to analyze the image data obtained with a NIR camera, using NDVI to correlate the image data with severity and defoliation obtained in the field. Three trials were conducted at the School Farm of the State University of Londrina, during the 2016/2017 crop season, at geographical coordinates S: 23°20’34.4” W: 51°12’41.3” and 570 m altitude, in the municipality of Londrina, Paraná State, Brazil. The used cultivar was DM6563 IPRO, and the population density was 260 thousand plants per hectare. Experimental design was in randomized blocks with four replicates, each replicate consisting of plots with six 6m-width rows spaced 0.45 meters apart. The trials included one control and several treatments with different applications of fungicides that showed the best efficacy for controlling Asian soybean rust during the study period. The trials were named Consfit, Expert and STCK, and rust intensity gradient was generated by applying different fungicides at diverse phenological stages and varying numbers of applications. Good correlation indices (above 90%) were observed with 3.4 cm pixels, and as the pixel size increased, the obtained correlation with both severity and defoliation decreased. The results suggest that it is possible to use aerial images with NIR cameras on UAVs to estimate severity and defoliation by Phakopsora pachyrhizi in soybean crops. This could make the disease monitoring more effective and the decision-making more precise regarding the crop phytosanitary management, thus contributing to reduced use of agrochemicals and optimized yields.
Keywords
NDVI; multispectral; Asian rust; remote sensing
Thumbnail
Thumbnail
Thumbnail
Thumbnail
Thumbnail
Thumbnail
