@Article{MoreiraVale:2014:ApEvTo,
author = "Moreira, Eder Paulo and Valeriano, M{\'a}rcio de Morisson",
affiliation = "{Instituto Nacional de Pesquisas Espaciais (INPE)} and {Instituto
Nacional de Pesquisas Espaciais (INPE)}",
title = "Application and evaluation of topographic correction methods to
improve land cover mapping using object-based classification",
journal = "International Journal of Applied Earth Observation and
Geoinformation",
year = "2014",
volume = "32",
pages = "208--217",
keywords = "Topographic effect, Landsat, SRTM, Classification accuracy, Land
cover classification, Radiometric correction.",
abstract = "This study applies and evaluates topographic correction methods to
reduce radiometric variation dueto topography characteristics in
rugged terrain. The aim of this study was to improve the
capability ofsatellite images to generate more reliable land cover
mapping using object-based classification. Severalsemi-empirical
correction methods, which require the estimation of empirically
defined parameters,were selected for this study. Usually, these
parameters are estimated relying on a previous land covermap.
However, in this work the correction methods were applied
considering the unavailability of a pre-vious land cover map and
the ease for implementation, so the main land cover type was used
to estimatecorrection parameters to be applied to correct all land
cover type. Landsat 5 TM image and topographicdata derived from
SRTM (Shuttle Radar Topography Mission) over an area located in an
agriculturalregion of southeastern Brazil were used. Land cover
classification was carried out using an object-basedapproach,
which includes image segmentation and decision tree
classification. The evaluation of topo-graphic correction methods
was based on: spectral characteristics expressed by standard
deviation andmean values of spectral data within land cover
classes; relationship between spectral data and solar
illu-mination angle on the slope (cos i); object (segment) mean
size; decision tree structure; visual analysis;and classification
accuracy. Results show that the standard deviation of spectral
data and the correlationbetween spectral values and cos i
decreased after data correction, but not for all methods for some
of thetested TM bands. The methods herein referred as Cosine, S1,
Ad2S and SCS methods showed to increasethe standard deviation and
the correlation compared to the uncorrected data, mainly for bands
1, 2 and3. Object mean size, in general, decreased after
correction, except for C method. The effect on the objectsize
showed to be related to a calculated standard deviation of
adjacent pixels values. The decision treestructure given by the
number of leaves also decreased after correction. The C, SCS + C
and Minnaertmethods showed the highest performance, followed by S2
and E-Stat, with a general accuracy increasearound 10%. Land cover
classification from uncorrected and corrected data differed in a
large portion ofthe total studied area, with values around 29% for
all correction methods.",
doi = "10.1016/j.jag.2014.04.006",
url = "http://dx.doi.org/10.1016/j.jag.2014.04.006",
issn = "0303-2434",
label = "lattes: 9558008969923427 1 MoreiraVale:2014:ApEvTo",
language = "en",
url = "http://www.sciencedirect.com/science/article/pii/S0303243414000907#",
urlaccessdate = "05 maio 2024"
}