US EPA

8802684 

Journal Article 

Gerais, Brazil 

Teles, LSB; Campos, JEG; Almeida, T; Cicerelli, RE; Yokoyama, E 

2020 

43 

1 

300-310 

Portuguese 

10.11137/2020_1_300_310 

Due to the development of large agricultural provinces, the human need for phosphated inputs has been growing exponentially, stimulating the creation and development of new tools to help identify and prospect new exploration targets. The objective of this work is to propose a methodological development and application of the Object-Based Supervised Classification technique to identify the outcrops of phosphated rocks near to the municipality of Coromandel, Minas Gerais, Brazil. For this purpose, the input data used in the classification were the Digital Elevation Model (ALOS / PALSAR-12.5m) and quantitative gamma spectrometric thematic maps related to the concentrations of the Uranium and Thorium radioelements in ppm, which resulted in a binary map capable of differentiating two main units in the study area: The Phosphated Unit and the Unphosphated Unit. The obtained result was compared with the filed data and showed an excellent quality (82%) by the coefficient Kappa means. © 2020, Universidade Federal do Rio de Janeiro. All rights reserved. 

Remote Sensing; Sedimentary Phosphate; Supervised Classification