Detección de niveles de infección de Alternaria spp. en Brassica oleracea var. botrytis, utilizando imágenes multiespectrales capturadas a corta distancia

Abstract

Early detection of diseases, such as Alternaria spp in cauliflower, is crucial to avoid losses. Multispectral imaging offers a solution, allowing the identification of damage patterns and improving agricultural productivity in a sustainable way. The present research, carried out the detection of Alternaria spp. infection levels in cauliflower, by capturing multispectral images at close range (2m). The work was carried out at the Faculty of Agricultural Sciences of the University of Cuenca, in which a 4X4 DBCA was implemented with 4 doses of Alternaria inoculum concentration (Dose 1: 1x103; Dose 2: 1x105; Dose 3: 1x107; Dose 4: 1x100 (control)). The images were captured with the Parrot Sequoia sensor at two phenological stages of the plant, and the severity was determined in the field using a conventional method. The results revealed that there were no significant differences in most cases; however, the NDVI, GNDVI and NDRE indices showed sensitivity after application. NDVI and GNDVI discriminated better the areas showing necrosis and chlorosis than NDRE. When comparing the two methods evaluated, the result was discrepancy, generating new doubts due to the limitation in the capacity presented by the camera to offer precision in the determination of the state of health of the plant, thus generating the need to review and adjust the evaluation methods that lead us to adjust different errors.