@InProceedings{GirolamoNetoFKSEBMT:2015:ClAuÁr,
author = "Girolamo Neto, Cesare Di and Fonseca, Leila Maria Garcia and
Korting, Thales Sehn and Sanches, Ieda Del Arco and Eberhardt,
Isaque Daniel Rocha and Bendini, Hugo do Nascimento and Marujo,
Rennan de Freitas Bezerra and Trabaquini, Kleber",
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 {Instituto Nacional de Pesquisas
Espaciais (INPE)} and {Instituto Nacional de Pesquisas Espaciais
(INPE)}",
title = "Classifica{\c{c}}{\~a}o autom{\'a}tica de {\'a}reas cafeeiras
utilizando imagens de sensoriamento remoto e t{\'e}cnicas de
minera{\c{c}}{\~a}o de dados",
booktitle = "Anais...",
year = "2015",
editor = "Gherardi, Douglas Francisco Marcolino and Arag{\~a}o, Luiz
Eduardo Oliveira e Cruz de",
pages = "1609--1616",
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 = "Coffee is the main crop produced in the southern of Minas Gerais
state, Brazil, and techniques for estimating the area used for
this crop are being intensely investigated in order to produce
reliable yield estimates. Coffee trees have a similar spectral
pattern to forest, making it difficult to automatically
distinguish these land use types. This study evaluated the Random
Forest and Decision Tree algorithms for an automatic
classification of coffee areas in municipality of Machado, Minas
Gerais, Brazil. First, the data were preprocessed by creating gray
level masks in each of the 11 bands of a Landsat-8 image. Then the
Random Forest and Decision Trees were trained and applied on the
image in order to verify its potential for discriminating coffee
areas. Lastly, the analysis and validation of the results were
conducted using as reference one map manually classified. The
Kappa index and the overall accuracy were used to assess the
quality of the models tested. The Random Forest classifier
presented better results than the Decision Trees, with an accuracy
of 84.13% and a Kappa index of 0.6, which is more accurate when
compared to previous studies. We also provide a list of bands that
were not suitable for this type of classification.",
conference-location = "Jo{\~a}o Pessoa",
conference-year = "25-29 abr. 2015",
isbn = "978-85-17-0076-8",
label = "303",
language = "pt",
organisation = "Instituto Nacional de Pesquisas Espaciais (INPE)",
ibi = "8JMKD3MGP6W34M/3JM495M",
url = "http://urlib.net/ibi/8JMKD3MGP6W34M/3JM495M",
targetfile = "p0303.pdf",
type = "Classifica{\c{c}}{\~a}o e minera{\c{c}}{\~a}o de dados",
urlaccessdate = "15 jun. 2024"
}