Automatic classification in Landsat images for the mapping of Otacílio Costa –SC

Abstract

This paper aimed to compare digital classification methods(supervised, unsupervised, object - oriented) in Landsatimages in order to map the changes in land use and occupationfor the years 2007 and 2017 for the municipality of OtacílioCosta - SC. For this purpose, images of the Landsat-5 TMsensor and the Landsat-8 OLI sensor were used. After thedigital processing of the images, the classes of use and soilcoverage were defined and the samples generated, dividedinto 60% of training and 40% of validation. Finally, theclassification accuracy statistics for each method werecalculated. The unsupervised methods were inefficient in allanalyzed years, while the supervised ones were superior tothe others. On the other hand, the object-orientedclassification presented a classification considered excellentin 2007 and very good in 2017. The performance of theclassification by the SVM method (Support Vector Machine)was excellent in 2007 and 2017, and it was considered thebest evaluated method. From this, the mapping of the classesof use and coverage revealed a reduction of 4.8% ofagricultural areas and 2.3% of urban areas and an increase of1% for vegetation and 1.5% for water bodies.