Avaliando a fertilidade do solo com imagens digitais

Abstract

A set of 177 soil samples from three Brazilian regions were digitally scanned and its images were correlated to soil’s fertility parameters using multivariate calibration methods. The LSSVMR presented the best sand, clay, and organic carbon content estimates. The textural classification using digital images matched 90.6%. The proposed method can determine sand, clay, and silt contents simultaneously in a few minutes. The method does not require the use of chemicals and it is in accordance with green chemistry lab analysis.