Banca de DEFESA: NAISILA CARVALHO PRESTES

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
DISCENTE : NAISILA CARVALHO PRESTES
DATA : 21/09/2026
HORA: 09:00
LOCAL: Meet
TÍTULO:

SPECTRORADIOMETRY APPLIED TO THE DISCRIMINATION OF IPÊ SPECIES

PALAVRAS-CHAVES:

spectral analysis; machine learning; hyperspectral spectroradiometry; remote sensing; spectral signatures; species discrimination; ipê species

PÁGINAS: 62
GRANDE ÁREA: Ciências Agrárias
ÁREA: Agronomia
RESUMO:

Hyperspectral remote sensing combined with machine learning represents a promising approach for tree species discrimination, as it enables the exploration of subtle differences in spectral responses related to the biochemical and structural characteristics of leaves. In this context, the present study aimed to evaluate the efficiency of spectroradiometry in discriminating four ipê species (Tabebuia rosea, Tabebuia roseoalba, Handroanthus chrysotrichus, and Handroanthus impetiginosus) occurring in the municipality of Sinop, Mato Grosso, Brazil. Leaf samples were collected in an urban environment, and spectral curves were acquired using an Ocean Optics STS-VIS-L-50-400-SMA sensor over the 450–824 nm spectral range. The data were organized into spectral bands and reflectance inflection difference (RID) indices and analyzed using multivariate techniques and machine learning algorithms. The spectral curves revealed differences among species, particularly in the visible and near-infrared regions. PCA, used to assess the conventional discrimination of the species, did not provide a clear separation among the groups. Logistic Regression showed the best overall performance, particularly when associated with the WL dataset, followed by SVM. Classifier performance depended on the set of spectral variables used.

MEMBROS DA BANCA:
Interno - 265126001 - CARLOS ANTONIO DA SILVA JUNIOR
Interno - 131916001 - RIVANILDO DALLACORT
Externo à Instituição - PAULO EDUARDO TEODORO - UFMS