@MastersThesis{Aguiar:1991:UtAtDe,
author = "Aguiar, Ana Paula Dutra de",
title = "Utiliza{\c{c}}{\~a}o de atributos derivados de
propor{\c{c}}{\~o}es de classes dentro de um elemento de
resolu{\c{c}}{\~a}o de imagem ({"}pixel{"}) na
classifica{\c{c}}{\~a}o multiespectral de imagens de
sensoriamento remoto",
school = "Instituto Nacional de Pesquisas Espaciais (INPE)",
year = "1991",
address = "Sao Jose dos Campos",
month = "1991-04-26",
keywords = "reconhecimento de padr{\~o}es, Itapeva (SP)
Mogi-Gua{\c{c}}{\'u}, classifica{\c{c}}{\~a}o autom{\'a}tica,
modelo linear de mistura, mapeador tem{\'a}tico (LANDSAT),
sat{\'e}lites LANDSAT, m{\'a}xima verossimilhan{\c{c}}a, uso da
terra, land use, pixels, spectral energy distribution, targets,
maximum likelihood estimates.",
abstract = "A energia espectral captada por instrumentos de Sensoriamento
Remoto e a integra{\c{c}}{\~a}o, denominada mistura, da energia
espectral refletida ou emitida por todos os objetos, denominados
componentes prim{\'a}rios da mistura, contidos no elemento de
cena. As classes de uso do solo presentes em uma cena podem ser
descritos em termos das propor{\c{c}}{\~o}es destes componentes,
especialmente para alvos florestais. O objetivo desta
disserta{\c{c}}{\~a}o {\'e} analisar o efeito obtido no
processo de classifica{\c{c}}{\~a}o autom{\'a}tica quando
utilizadas bandas sint{\'e}ticas derivadas das
propor{\c{c}}{\~o}es dos componentes em cada pixel. E adotado um
Modelo Linear de Mistura e empregados os m{\'e}todos de
M{\'{\i}}nimos Quadrados com Restri{\c{c}}{\~o}es e
m{\'{\i}}nimos Quadrados Ponderado para estimar as
propor{\c{c}}{\~o}es. As imagens utilizadas (LANDSAT TM)
referem-se a duas {\'a}reas de reflorestamento, denominadas
{"}ITAPEVA{"} e {"}MOGI GUA{\C{C}}U{"}. A an{\'a}lise do
processo de classifica{\c{c}}{\~a}o baseia-se no algoritmo de
Maxima Verossimilhan{\c{c}}a, sob hip{\'o}tese gaussiana, e em
m{\'e}todos de redu{\c{c}}{\~a}o da dimens{\~a}o do
espa{\c{c}}o de atributos frequentemente empregados em
Sensoriamento Remoto. Os resultados obtidos mostram que a partir
de conjuntos substitutos de atributos (formados pela
adi{\c{c}}{\~a}o de bandas sint{\'e}ticas as originais ou
somente pelas bandas sin{\'e}ticas) obt{\'e}m-se, de modo geral,
uma maior compacta{\c{c}}{\~a}o de atributos pelas
transforma{\c{c}}{\~o}es de Componentes Principais e
An{\'a}lise Can{\^o}nica. Contudo, n{\~a}o se obteve melhoria
significativa nas estimativas de desempenho m{\'e}dio e nos
valores de Distancia J-M entre as classes. No entanto, a
an{\'a}lise qualitativa das imagens tem{\'a}ticas forneceu
importantes resultados: a) para as cenas analisadas, concluiu-se
que n{\~a}o se deve utilizar conjuntos substitutos formados pela
adi{\c{c}}{\~a}o de bandas sint{\'e}ticas as originais; b) os
melhores resultados s{\~a}o obtidos pela utiliza{\c{c}}{\~a}o
somente das bandas sint{\'e}ticas, desde que estas sejam geradas
a partir de componentes que representam de forma adequada as
classes da cena e cujas propor{\c{c}}{\~o}es indiquem
diferen{\c{c}}as estruturais dos alvos. Constata-se, desta forma,
a import{\^a}ncia da sombra como componente prim{\'a}rio para
alvos florestais. Utilizar somente bandas sint{\'e}ticas pode ser
visto como um m{\'e}todo de redu{\c{c}}{\~a}o da dimens{\~a}o
do espa{\c{c}}o de atributos compar{\'a}vel aos m{\'e}todos
usualmente empregados em Sensoriamento Remoto. As bandas
sint{\'e}ticas podem tamb{\'e}m ser {\'u}teis para
interpreta{\c{c}}{\~a}o visual, pois, al{\'e}m do excelente
efeito visual obtido pela sua composi{\c{c}}{\~a}o colorida, sua
informa{\c{c}}{\~o}es representam conceitos f{\'{\i}}sicos
(propor{\c{c}}{\~o}es) mais facilmente assimil{\'a}veis do que
as assinaturas espectrais das classes. ABSTRACT: The spectral
energy collected by the Remote Sensing instrumentation is the
integration, called mixture, of the energy reflected or emmited by
the objects, called primary components of the mixture, contained
in a picture element. The land use classes in a scene can be
described in terms of these components proportions, specially for
forest targets. The objective of this issertation is to analyze
the effect obtained in the automatic classification process when
utilizing synthetic bands derived from components proportions in
each pixel. A Linear Mixing Model is adopted and the estimated
proportions are obtained by the use of the Constrained Least
Squares and weighted Least Squares methods. The Landsat TM images
used for the tests correspond to two reforested areas denominated
{"}ITAPEVA{"} and {"}MOGI-GUAU{"}. The analysis of the
classification process is based on the Maximum Likelihood
Algorithm using the gaussian hypothesis, and on methods of
dimensionality reduction frequently employed in Remote Sensing.
The obtained results show that, in general, by using a substitute
attribute set (formed by the addition of the synthetic bands to
the originals or only by the synthetic bands), a greater
compression performance under the Principal Components and
Canonical Analysis transformations is obtained. However, not
significant improvement in the estimation of the average
performance nor in the J-M Distance values between classes was
obtained. In spite of this fact, the qualitative analysis of the
thematic images provided important results: a) for the analyzed
images it is possible to conclude that one should not use the
substitute set composed by the addition of the synthetic bands to
the originals; b) the best results are obtained by utilizing the
synthetic bands only, provided that they are generated from the
components which adequately represent classes in the scene and
such that their proportions indicate the target structural
differences. One observes, in this way, the importance of shade as
a primary component for forest targets. The use of the synthetic
bands only can be seen as a method for reducing the feature space
comparable to the methods usually employed in Remote Sensing. The
synthetic bands can also be useful for manual interpretation, due
to their excellent visual effect produced by colour composites and
also because their information represents physical concepts
(proportions) easier assimilated than the classes spectral
signatures.",
committee = "Mascarenhas, Nelson Delfino d'{\'A}vila (presidente/orientador)
and Shimabukuro, Yosio Edemir (orientador) and Haertel, Vitor
Francisco de Ara{\'u}jo and Banon, Gerald Jean Francis",
copyholder = "SID/SCD",
englishtitle = "Use of features derives from class proportions in a pixel for the
multispectral classification of remote sensing images",
label = "1406",
language = "pt",
pages = "227",
ibi = "6qtX3pFwXQZ3r59YD6/GNVHb",
url = "http://urlib.net/ibi/6qtX3pFwXQZ3r59YD6/GNVHb",
targetfile = "publicacao.pdf",
urlaccessdate = "06 maio 2024"
}