NDVI imagens for monitoring nutritional deficits in plants

Abstract

New technologies have been successfully introduced into agriculture to speed up decision-making through quick analysis of crops. Among these, the vegetation indices capture the reflectance of the wavelengths of plant structures and allow inferring the quality of the plant mass. Among the various models, the Normalized Difference Vegetation Index (NDVI) is the most used. Aiming to test this functionality in protected environments, this work aimed to evaluate the use of the NDVI index for crops in protected cultivation. For this, weekly images of strawberry and cabbage cultivation were captured, which received different nutritional treatments. The strawberry was grown in a hydroponic system and was subjected to three different solutions: supply water with recommended fertilization (TA), reuse water with complement through commercial fertilizers (TRA), and reuse water only (TR). The cabbage was grown in a bed system with three doses of controlled-release fertilizer (100%, 80%, and 60% of the dose recommended by the IAC 100 bulletin). The images were processed in QGIS software to generate NDVI images. Cabbage cultivation did not present differences in the nutritional status of the plants, therefore, the doses of fertilizers used did not promote physiological changes that would demonstrate symptoms of deficiency, nor a drop in production, consequently, there was no difference in biomass reflectance. Strawberry cultivation, on the other hand, showed visible deficiencies in the treatment using only reused water, with a difference of up to 20% of NDVI observed during monitoring. Thus, the results conclude that NDVI images can monitor nutritional disturbances in the evaluated plants and may be a helpful technology in the automation of agricultural greenhouses.