Banca de QUALIFICAÇÃO: JONATTHAN ALVES FAGUNDES

Uma banca de QUALIFICAÇÃO de MESTRADO foi cadastrada pelo programa.
DISCENTE : JONATTHAN ALVES FAGUNDES
DATA : 05/11/2019
HORA: 09:30
LOCAL: Departamento de Engenharia
TÍTULO:

PREDICTION OF COFFEE CROP PRODUCTIVITY USING DIGITAL IMAGES OBTAINED BY UNMANNED AERIAL VEHICLES

PALAVRAS-CHAVES:

Precision agriculture, NDVI, leaf area index, agrometeorological-spectral model

PÁGINAS: 20
GRANDE ÁREA: Ciências Agrárias
ÁREA: Engenharia Agrícola
SUBÁREA: Máquinas e Implementos Agrícolas
RESUMO:

Coffee is the second most traded commodity in the world. Brazil is responsible for approximately one third of world production, according to the International Coffee Organization, being the country with the largest production in the world 2017-2018. In the context of precision agriculture, yield maps play an essential role within the flow of precision agriculture activities, and are one of the main means for determining spatial variability within an agricultural crop. With the advancement of technologies the use of Unmanned Aerial Vehicles (UAVs) has grown around the world. These are increasingly being used in agriculture, among other applications to obtain productivity maps. This project aims to predict the productivity of coffee crops using an agrometeorological-spectral model and digital images obtained with the UAV. The experiment will be carried out in a field located in the city of Poço, in the geographic mesoregion of southern Minas Gerais State, from an agrometeorological-spectral model. The methodology will consist of adapting an agrometeorological-spectral model, whose input spectral variable is the leaf area index, which will be obtained through the NDVI spectral index calculated from the multispectral images of a sensor coupled to a UAV. The component meteorological variables of the agrometeorological-spectral models will be obtained from a weather station of the National Institute of Meteorology and for the determination of the actual evapotranspiration the Penman-Monteith / FAO model will be used. The images will be collected monthly from December 2019 to July 2020 and correlated with yield data of three plots composed of different cultivars. Thus, the estimated productivity will be evaluated at each moment (month) by the agrometeorological-spectral model which will be compared with the actual yield obtained in the field. It is expected at the end of the project, demonstrate the ability to predict the productivity of coffee crops from multispectral images obtained by VANTS, assisting in the planning of yield management and harvesting practices.

MEMBROS DA BANCA:
Presidente - 2161955 - ANDERSON GOMIDE COSTA
Interno - 2161784 - JOAO PAULO BARRETO CUNHA
Externo à Instituição - MINELLA ALVES MARTINS