O uso da terra da bacia do ribeirão Bom Jardim-MG por meio de imagem do satélite Resource
Laís Naiara Gonçalves dos Reis²
Dayane Zandonadi Soares²
Camila Nonato Junqueira³
¹ Universidade de Brasília – UnB/IGD
Campus Darcy Ribeiro, ICC Ala Central - 70919-970 - Brasília-DF, Brasil mireillegeo@gmail.com ²Universidade Federal de Uberlândia – UFU/ IG
Av. João Naves de Ávila, 2121 - Bl. H - 38400-902 - Uberlândia-MG, Brasil lais_ungida@hotmail.com, ecodayane@yahoo.com.br
³ Universidade Federal de Uberlândia – UFU/IB
Campus Umuarama, Bloco 2D, Sala 28 - 38400-902 - Uberlândia-MG, Brasil camilan.j@hotmail.com Abstract. The Indian origin satellite Liss3 – Linear Imaging Self-Scanner Satellite – Resource enables the acquisition of information related to vegetation, crops and characterization of plant species because it has four spectral bands: band 2 (green) from 0.52 to 0.59 µm; band 3 (red) 0.62 to 0.68 µm, band 4 (near infrared) from
0.77 to 0.86 µm band 5 (mid infrared) from 1.55 to 1.70 µm.
The first scene was captured in November 2009. The study aims to verify the accuracy of geospatial technologies by using satellite image to produce a land use map.
We used the software SPRING 5.2 for the Digital Image Processing (DIP) of the LISS3/RESOURCE image, orbit 329, point 091, September 7, 2010 with complete scene of 141x141 km, 23 m of spatial resolution, 24 days of revisit period and 7 bits of radiometric resolution. Currently, images are freely available by the Instituto
Nacional de Pesquisas Espaciais (INPE). The PDI procedures used was radiometric and geometric correction followed by an automatic mapping using Bhatthacharya method. We mapped the following thematic classes: forestry, native vegetation, agriculture, pasture and water resources. To check the mapping accuracy it was used a land truth map of the study area prepared by Resende (2011) and Kappa index. As a result of this work, regarding land use, there