%0 Conference Proceedings
%4 sid.inpe.br/marte2/2017/10.27.15.29
%2 sid.inpe.br/marte2/2017/10.27.15.29.01
%@isbn 978-85-17-00088-1
%F 59330
%T Mapeamento Multitemporal de queimadas na bacia do rio Grande-BA: aplicação de uma Rede Neural Artificial em produtos MODIS
%D 2017
%A Pinheiro, Priscila Santos,
%A Borges, Elane Fiuza,
%A Sano, Edson Eyji,
%@electronicmailaddress pinheiro.priscila@hotmail.com
%E Gherardi, Douglas Francisco Marcolino,
%E Aragão, Luiz Eduardo Oliveira e Cruz de,
%B Simpósio Brasileiro de Sensoriamento Remoto, 18 (SBSR)
%C Santos
%8 28-31 maio 2017
%I Instituto Nacional de Pesquisas Espaciais (INPE)
%J São José dos Campos
%P 5591-5598
%S Anais
%1 Instituto Nacional de Pesquisas Espaciais (INPE)
%X Fire in the Cerrado biome is used as a management tool. In agriculture, the fire is used for the cleaning of the pastures, as well as regrowth of the vegetation to serve as food for the herd. However, recurring fire practices in this environment end up causing severe damage to the environment. Thus, remote sensing, combined with other practices, is important in terms of monitoring and conservation of the landscape. In this way, this paper aimed to map the areas of burn scars to the Rio Grande-BA basin, from 2005 to 2014, through the EVI of the sensor MODIS and an artificial neural network. For the collection of input samples from the neural network a Graphical User Interface (GUI) was created, where the user is able to arbitrate which input data and their possible percentages, the number of samples, as well as the window size of scanning of incoming data. The network was trained in the MATLAB, with the backpropagtion algorithm. For the validation of the neural network a manual vetorization was used from data from the Landsat and Resourcesat series for the same period analyzed. Next, from the confusion matrix, errors of omission and commission, global accuracy and Kappa index were generated. Where the data were found in the latter classification with qualities ranging from good to very good and global accuracy ranging from 65% to 82%.
%9 Análise de séries temporais de imagens de satélite
%@language pt
%3 59330.pdf