DERIVATION AND APLICATION OF ALGORISM OF 5 METERS (IMAGE MASK) FOR MANGROVE CLASSIFICATION

Mangroves' forest are one of the most important coastal ecosystems. Distributed in tropical and sub-tropical regions, mangroves achieve their maximum development in southwest of Asia. Mangroves' forest are under severe stress mainly in sub-tropical countries due to natural tensors and the demographic ones. Along the Brazilian coast, mangroves are subjected to the effects of human growth and economic pressure with their diverse activities affecting mangrove structural development and biodiversity. It is necessary and urgent that mangrove are integrated in the Coastal Planning as a way of preservation and conservation of its natural resources through a sustainable management. Remote sensing techniques and GIS were applied in the actualization of Ratones' mangrove through digital images from Spot-Pan, orbit 714-405, special resolution of 10 meters, 07/11/1995 and from Landsat Thematic Mapper-TM, channels 1 to 5 and 7, orbit 220-078, special resolution of 30 meters, 15/07/1993. In the process a polygon of 5 meters height (mask) was applied with the objective to separate mangrove classes from water. This procedure was important and efficient to evaluate pixels mixture in the occurred classes in the supervised classification in the entire scene and to estimate mangrove's stands with regression equations between them and satellite date.