@InProceedings{BendiniGKMTESF:2015:EfImFu,
author = "Bendini, Hugo do Nascimento and Girolamo Neto, Cesare Di and
Korting, Thales Sehn and Marujo, Rennan de Freitas Bezerra and
Trabaquini, Kleber and Eberhardt, Isaque Daniel Rocha and Sanches,
Ieda Del Arco and Fonseca, Leila Maria Garcia",
affiliation = "{Instituto Nacional de Pesquisas Espaciais (INPE)} and {Instituto
Nacional de Pesquisas Espaciais (INPE)} and {Instituto Nacional de
Pesquisas Espaciais (INPE)} and {Instituto Nacional de Pesquisas
Espaciais (INPE)} and {} and {} and {Instituto Nacional de
Pesquisas Espaciais (INPE)} and {Instituto Nacional de Pesquisas
Espaciais (INPE)}",
title = "Effects of Image Fusion Methods on Sugarcane Classification with
Landsat-8 Imagery",
booktitle = "Anais...",
year = "2015",
editor = "Gherardi, Douglas Francisco Marcolino and Arag{\~a}o, Luiz
Eduardo Oliveira e Cruz de",
pages = "2498--2505",
organization = "Simp{\'o}sio Brasileiro de Sensoriamento Remoto, 17. (SBSR)",
publisher = "Instituto Nacional de Pesquisas Espaciais (INPE)",
address = "S{\~a}o Jos{\'e} dos Campos",
abstract = "The culture of sugarcane has great importance in the brazilian
agribusiness. Remote sensing images have been tradicionally used
on manual mapping of sugarcane fields. Manual classification is a
laborious and time-consuming task, especially given the size of
the territory, and it is still necessary to assess the quality of
the maps. Image fusion can improve the identification and mapping
of surface features. The computational data mining methodology
demonstrates high potential for application in areas related to
crop mapping and several classification techniques can be used.
Most studies on fusion of remote sensing images have focused on
the analysis of spectral and spatial quality of the products
obtained by different algorithms, however, once classification is
applied on these products, it is important to analyze the impact
of fusion in the classification. In the literature there are few
studies on this topic, especially considering the Landsat-8. In
this context, we evaluated five pansharpening methods -
Intensity-Hue-Saturation (IHS), Principal Components (PC),
Gran-Schmidt (GS), Discrete Wavelet Transform (DWT) and DWT+IHS
for the classification of sugarcane fields in a Landsat-8 image
(bands 4, 5 and 6). The Support Vector Machine (SVM) algorithm was
used to perform a target detection of sugarcane, using a binary
classification. The samples used were selected based on a field
survey realized on the study area. The best fusion techniques were
the DWT+IHS, DWT and IHS, which obtained higher Universal Image
Quality Index (UIQI) and Spatial Relative Dimensionless Global
Error in Synthesis (SERGAS) values. However, considering the
effects on classification, the GS fusion showed better results
than other methods.",
conference-location = "Jo{\~a}o Pessoa",
conference-year = "25-29 abr. 2015",
isbn = "978-85-17-0076-8",
label = "504",
language = "pt",
organisation = "Instituto Nacional de Pesquisas Espaciais (INPE)",
ibi = "8JMKD3MGP6W34M/3JM4A4E",
url = "http://urlib.net/ibi/8JMKD3MGP6W34M/3JM4A4E",
targetfile = "p0504.pdf",
type = "Processamento de imagens",
urlaccessdate = "27 abr. 2024"
}