Banca de QUALIFICAÇÃO: RAUL PIO DE AZEVEDO

Uma banca de QUALIFICAÇÃO de MESTRADO foi cadastrada pelo programa.
DISCENTE : RAUL PIO DE AZEVEDO
DATA : 20/01/2023
HORA: 15:00
LOCAL: Google Meet (videoconferência)
TÍTULO:

REMOTELY SENSING IMAGES AND MACHINE LEARNING IN SESAME CULTURE MAPPING

PALAVRAS-CHAVES:

Sesame, Landsat 8/OLI, Machine Learning, Mapping, Random Forest, SVM.

PÁGINAS: 92
GRANDE ÁREA: Ciências Exatas e da Terra
ÁREA: Geociências
SUBÁREA: Geofísica
ESPECIALIDADE: Sensoriamento Remoto
RESUMO:

The changes in landscapes have been accompanied more intensely in recent decades, thanks to scientific advances, both in the field of technological improvement of satellites as in remote sensing techniques. These alterations have a significant impact on the ecosystem, and to know about them is extremely important from a socio-economic, environmental and scientific from the point of view. To this end, advanced and efficient machine learning techniques have aided professionals in remote sensing to determine these changes, from the simplest to the most complex landscapes, enabling the identification of the most varied uses and occupation of the land, as well at estimating the areas that these uses occupy, allowing a more dynamic management of natural resources, especially in agricultural exploitation, providing reliable information to decision makers. Thus, images from the Landsat 8/OLI satellite have been widely used in land use and land cover classification, especially in the discrimination of agricultural crops, since its temporal and spatial resolution, and free availability, allow tracking the development of summer and winter crops during their life cycle, including the variations that the vegetation indices present according to the stage of vegetative development of the plant. Therefore, the objective of this work is to estimate the cultivated area of the sesame (Sesamum indicum) crop in the crop year 2021/2022, in the municipality of Canarana in the state of Mato Grosso, comparing the performance of Random Forest and Support Vector Machine classifiers, using images from the Landsat 8/OLI satellite. For this purpose, control points in geographic coordinates were collected in the study area, to characterize the samples of the sesame cultivated areas, as a source of information to the supervised classification. The vegetation indices NDVI, EVI, NDBI, PVI and SAVI were used, in addition to Landsat 8/OLI images for the elaboration of thematic maps using machine learning techniques. In the evaluation of the thematic maps, the Global Accuracy and the Kappa index were used as a rule, compared by the Z test, at a significance of α=0.05. The test showed that the Random Forest classifier presented better performance over the SVM in the identification of land uses and land cover, especially for sesame culture. Thus, the use of machine learning techniques on Landsat 8/OLI images proved satisfactory in estimating and mapping sesame cultivated areas.

MEMBROS DA BANCA:
Presidente - 265126001 - CARLOS ANTONIO DA SILVA JUNIOR
Interno - 131916001 - RIVANILDO DALLACORT
Externo à Instituição - CÁCIO LUIZ BOECHAT - UFPI