Use of altimetry and images of satellite in the differentiation of forest types / Uso de altimetria e imagens de satÃlite na diferenciaÃÃo de tipos florestais

AUTOR(ES)
DATA DE PUBLICAÇÃO

2005

RESUMO

This work shows the auxiliary data integration to improve the differentiation of forest types with the digital classification of image of sensor ETM+ (Landsat 7). The auxiliary data are here compared with topography of the place. The study area is into a brejo de altitude and is located among the cities of Belo Jardim, Brejo da Madre de Deus e TacaimbÃ, in the agreste of Pernambuco. The methodology classifies part of the image that recovers the area of study for the supervised method using the algorithm of the maximum likelihood and to before compare this result with the integration of data auxiliary, during and after the process of classification of the image. The classification errors matrix is allowed to analyze the spectral similarities between the classrooms of forest types identified and the accuracy of the classification by means of the global performance of the classification and the coefficient of Kappa contingency. The classification without integration of altimetry presented a global performance of 90,29% and 0,8544 (85,44%) of Kappa and with in the incorporation of altimetry before the classification process have a global performance of 94,74% and a coefficient kappa of 0,9157 (91,57%). For the incorporation of altimetry during of the classification has a global performance of 95,18% and a coefficient kappa of 0,9211 (92,11%). The altimetry incorporation after the classification process has a global performance of 94,57% and a coefficient kappa of 0,9134 (91,34%). The three used techniques had shown that they are efficient to improve the estimate of the areas covered for these types of vegetation

ASSUNTO(S)

tipos florestais integraÃÃo de dados auxiliares geodesia forest types classificaÃÃo classification integration of data auxiliary

ACESSO AO ARTIGO

Documentos Relacionados