Multispectral remote sensing image classification algorithm based on rough set theory

Rough set theory is a relatively new mathematical tool to deal with imprecise, incomplete and inconsistent data. A method of multispectral image classification using rough set theory is proposed. First, to decrease computational time and complexity, band reduction of multispectral image using attribute reduct concept in rough set theory and information entropy is performed. Then, mixture model initial parameters of remote sensing image are mapped from crude classes, which are generated using equivalent relation. Finally image cluster is obtained unsupervised with Gaussian mixture model whose parameters are refined by Expectation Maximization algorithm. The proposed method is performed on a multispectral image, and the experimental results show the feasibility and effectiveness of the algorithm by means of comparison and analysis.