@InProceedings{SimõesCamaQuei:2017:FiClMe,
               author = "Sim{\~o}es, Rolf Ezequiel de Oliveira and Camara, Gilberto and 
                         Queiroz, Gilberto Ribeiro de",
          affiliation = "{Instituto Nacional de Pesquisas Espaciais (INPE)} and {Instituto 
                         Nacional de Pesquisas Espaciais (INPE)} and {Instituto Nacional de 
                         Pesquisas Espaciais (INPE)}",
                title = "Filtering and Clustering Methods For Satellite Image Time Series",
            booktitle = "Anais...",
                 year = "2017",
         organization = "Workshop dos Cursos de Computa{\c{c}}{\~a}o Aplicada do INPE, 
                         17. (WORCAP)",
             keywords = "Satellite Image Time Series, Filtering, Clustering.",
             abstract = "Using time series derived from big Earth Observation data sets is 
                         one of the leading research trends in Land Use Science and Remote 
                         Sensing. One of the more promising uses of satellite time series 
                         is its application for classification of land use and land cover, 
                         since our growing demand for natural resources has caused major 
                         environmental impacts. Given this motivation, this work provides a 
                         survey of two topics which are relevant for image classification: 
                         noise removal and cluster analysis. In noise removal, we 
                         investigate different techniques for filtering and smoothing time 
                         series. For cluster analysis, we discuss methods that have been 
                         published in the literature for time series clustering and test 
                         their application to SITS. This discussion is illustrated by a 
                         number of examples. In the filtering part, we did two experiments 
                         of using smoothing methods to support the classification of noisy 
                         time series. The smoothing methods applied to the data and a 
                         cross-validation showed that our improvement were insignificant 
                         when compared to the original data experiment. In the clustering 
                         part, we present some preliminary results of using agglomerative 
                         hierarchical clustering with ward linkage as merging criterion. 
                         The result of the experiments suggest that the patterns extraction 
                         by clustering process is a promising technique to improve 
                         prototype extraction from SITS data.",
  conference-location = "S{\~a}o Jos{\'e} dos Campos, SP",
      conference-year = "20-22 nov. 2017",
             language = "en",
           targetfile = "Simoes_filtering.pdf",
        urlaccessdate = "01 maio 2024"
}