Predicting soil texture using image analysis
Nenhuma Miniatura disponível
Data
2019
Título da Revista
ISSN da Revista
Título de Volume
Editor
Resumo
Laboratory analysis of soil texture is laborious and not environmentally friendly. After sampling, another 56 h are required for the final report and the laboratory procedure employs hydrogen peroxide and sodium hydroxide as chemical dispersion agents. Therefore a new analytical method to predict and classify soil texture is proposed using digital image processing of soil samples (image segmentation) and multivariate image analysis (MIA). Digital images of 63 soil samples, sieved to <2 mm, were acquired. Clay and sand contents determined by the pipette method were used as standard values and, after image processing, particle contents in the measured size fractions were correlated to image data using PLS2 multivariate regression. In order to statistically account for the sampling and validation dataset multivariate statistics was evaluated in conjunction with bootstrapping analysis. The computer vision approach adopted for the recognition of soil textures based on soil images matched 100% of the classification predicted according to the standard method. The new method is low-cost, environment-friendly, nondestructive, and faster than the standard method.
Descrição
Palavras-chave
Citação
MORAIS, Pedro Augusto de Oliveira et al. Predicting soil texture using image analysis. Microchemical Journal, Amsterdam, v. 146, p. 455-463, 2019. DOI: 10.1016/j.microc.2019.01.009. Disponível em: https://www.sciencedirect.com/science/article/abs/pii/S0026265X18313626?via%3Dihub. Acesso em: 14 jun. 2023.