The use of forest raw products has increased significantly in the last decades. The search for high productivity has led to the introduction of exotic species, mainly Eucalyptus sp. and Pinus sp. This work evaluated the precision of classification obtained from surveying forest implemented and natural stands in the region of the map area of Cachoeira do Sul - RS, using geoprocessing techniques, remote sensing, GIS (geographic information system) and GPS (global positioning system). It was verified that the area is occupied by natural vegetation (35.54%), Pinus sp. (1.89%) and Eucalyptus sp. (0.77%), with the following values precision in the supervised digital classification: Global precision (85.23%), Kappa (84.90%) and Tau (77.74%). It was concluded that the three accuracy indexes can be used, although Kappa and Tau were more consistent.
Geoprocessing; accuracy indexes; forest surveys
