Banca de QUALIFICAÇÃO: THAÍS MACHADO DE SOUZA

Uma banca de QUALIFICAÇÃO de MESTRADO foi cadastrada pelo programa.
STUDENT : THAÍS MACHADO DE SOUZA
DATE: 14/06/2024
TIME: 14:00
LOCAL: Google Meet - Exame de Qualificação online
TITLE:

WEED CLASSIFICATION USING SPECTRAL RESPONSES OBTAINED FROM PROXIMAL AND SUBORBITAL SENSORS IN CORN CROPS

KEY WORDS:

machine learning, multivariate analysis, remote sensing, image classification.

PAGES: 33
BIG AREA: Ciências Agrárias
AREA: Engenharia Agrícola
SUBÁREA: Máquinas e Implementos Agrícolas
SUMMARY:

Corn is an of the most important crops in Brazil and the world, in addition to being a source of animal and human nutrition, it is also a raw material for the production of by-products in the pharmaceutical, chemical, fuel and beverage industries. Given such importance, it is essential for producers to manage their crops properly, one of the main problems is weeds, which compete with crops and are difficult to identify due to their similarity in color, shape and size. In this project, weed species present in the corn crop will be discriminated using Remote Sensing (RS) techniques, through the acquisition of multispectral images by Remotely Piloted Aircraft (RPA) and obtaining spectral data of weeds and corn by spectroradiometer. The images collected by RPA will be classified using Object Based Image Analysis (OBIA) by Random Forest algorithm (supervised machine learning). Principal Component Analysis (PCA) will be used to identify the most significant bands and the main differentiating bands between the plants in the area in the data obtained by the spectroradiometer. It is hoped that the results of both techniques will be able to differentiate the weed species in the maize crop, making it possible to check whether the use of RPA allows for reliable classification through the resolution of the image obtained, as well as recommending the ideal spectral ranges for distinguishing between the different plant species. In addition, it is hoped that these tools will enable faster decision making, through results obtained in a short period of time and non-destructive methods in crop management, allowing appropriate and localized treatment, through the variability of the area, defining Precision Agriculture (PA).

COMMITTEE MEMBERS:
Presidente - 1050615 - MURILO MACHADO DE BARROS
Interno - 2161955 - ANDERSON GOMIDE COSTA
Externo à Instituição - GABRIEL ARAUJO E SILVA FERRAZ - UFLA