@InProceedings{JijonCentMach:2017:ClDaHi,
               author = "Jijon, Mario Ernesto and Centeno, Jorge Antonio Silva and Machado, 
                         Alvaro Muriel Lima",
                title = "Classifica{\c{c}}{\~a}o de dados hiperespectrais AVIRIS baseada 
                         na codifica{\c{c}}{\~a}o bin{\'a}ria",
            booktitle = "Anais...",
                 year = "2017",
               editor = "Gherardi, Douglas Francisco Marcolino and Arag{\~a}o, Luiz 
                         Eduardo Oliveira e Cruz de",
                pages = "1177--1185",
         organization = "Simp{\'o}sio Brasileiro de Sensoriamento Remoto, 18. (SBSR)",
            publisher = "Instituto Nacional de Pesquisas Espaciais (INPE)",
              address = "S{\~a}o Jos{\'e} dos Campos",
             abstract = "Hyperspectral sensors provide images in hundreds of spectral bands 
                         that allows to discriminate objects with more details. But, the 
                         number of available training samples is limited. Thus, the 
                         dimensionality reduction is very important for the classification 
                         of high dimensional data. The approach developed in this work was 
                         the binary coding that was applied in hyperspectral data for 
                         dimensionality reduction. This encoding is based on a simple code 
                         and applied to a spectrum of reflectance pixel by pixel. 
                         Furthermore it seeks to develop a spectral representation that 
                         facilitates the identification of classes and their separability 
                         through the establishment of spectral libraries that stocks a 
                         number of spectra. For that, several experiments that allow the 
                         comparison of land cover classifications were tested. The 
                         alternatives that were performed on binary encoding were applied 
                         to a number of spectral regions (spectral libraries). Each 
                         alternative has been tested to a binary code through one to 
                         various thresholds. The results of these experiments show that the 
                         use of the binary encoding based on three thresholds and by 
                         regions allow the thematic mapping image classification and also 
                         reduce the dimensionality of hyperspectral data, being then more 
                         efficient than the use of one threshold for all the bands.",
  conference-location = "Santos",
      conference-year = "28-31 maio 2017",
                 isbn = "978-85-17-00088-1",
                label = "59229",
             language = "pt",
         organisation = "Instituto Nacional de Pesquisas Espaciais (INPE)",
                  ibi = "8JMKD3MGP6W34M/3PS4G9N",
                  url = "http://urlib.net/ibi/8JMKD3MGP6W34M/3PS4G9N",
           targetfile = "59229.pdf",
                 type = "Geoprocessamento e aplica{\c{c}}{\~o}es",
        urlaccessdate = "09 maio 2024"
}