Análise comparativa entre imagens de satélites cbers2 e landsat5 na classificação da cobertura vegetal na região de jaboticabal – sp

próximo artigo

93 4

Anais XIV Simpósio Brasileiro de Sensoriamento Remoto, Natal, Brasil, 25-30 abril 2009, INPE, p. 1943-1949.

Análise comparativa entre imagens de satélites CBERS2 e LANDSAT5 na classificação da cobertura vegetal na região de Jaboticabal – SP Christiano Luna Arraes1 Marcos Sales Rodrigues1 Tatiane Pereira Santos Morais1 Célia Regina Paes Bueno1 Teresa Cristina Tarlé Pissarra1
1

Abstract. Remote sensing images present potential for mapping and estimating crop yield. This paper details a technique for estimating percent vegetation cover based on land use classify on satellite images using a geographic information system (GIS)-IDRISI and CBERS2 and LANDSAT5 satellites images. The experiment was conducted at the experimental area of the Agricultural Sciences and Veterinary College – FCAV – UNESP, Jaboticabal, São Paulo State, Brazil. The GIS used for geoprocessing the images was IDRISI v.15.01 and the software CartaLinx was used for vetorizing and cutting the images and layers. The software used for sampling a land use in a statistically valid way to arrive at an estimate of the percent land cover. Five classes of the land use at the experimental area were classified as soil uncovered, sugar cane (“raw cane”), forest (including native forest, riparian forest and area with rubber tree planting), mechanically harvested cane and straw (up to the areas of tillage). The image of the satellite CBERS2 presented the five classes described, and at the image from satellite LANDSAT5 was classified just four of those classes. The image of the satellite CBERS2 showed greater efficiency for classifying the land use than the image LANDSAT5, mainly due to its spatial resolution finest. The study indicated the suitability of using CBERS satellite images for precise estimation of land use and the scope for making valuable management decisions on crop cultivation. Palavras-chave: remote sensing, geographic information system (GIS), IDRISI, geoprocessing,