Demarchi, Julio Cesar [UNESP]Piroli, Edson Luís [UNESP]Zimback, Célia Regina Lopes [UNESP]2014-05-272014-05-272011-12-19RA'E GA - O Espaco Geografico em Analise, n. 21, p. 234-271, 2011.1516-4136http://hdl.handle.net/11449/73073This work aims to analyze the land use evolution in the city of Santa Cruz do Rio Pardo - SP through supervised classification of Landsat-5 TM satellite images according to the maximum likelihood (Maxlike), as well as verifying the mapping accuracy through Kappa index, comparing NDVI and SAVI vegetation indexes in different adjustment factors for the canopy substrate and determining the vegetal coverage percentage in all methods used on 2007, May 26 th; 2009, January 7 th and 2009, April 29 th. The Maxlike classification showed several spatial changes in land use over the study period. The most appropriated vegetation indexes were NDVI and SAVI - 0,25 factor, which showed similar values of vegetal coverage percentage, but discrepant from the inferred value for Maxlike classification.234-271porImages classificationLand use mappingVegetation indexesaccuracy assessmentcanopyimage classificationland use changeLandsat thematic mapperlandscape evolutionNDVItemporal analysisvegetation mappingBrazilSanta Cruz do Rio PardoSao Paulo [Brazil]ArtigoAcesso aberto2-s2.0-834551761342-s2.0-83455176134.pdf0000-0002-3350-2651