Dimension reduction of remote sensing images by incorporating spatial and spectral properties

A new and efficient dimension reduction method is introduced in this paper. The proposed method, almost the same as the well-known principal component analysis (PCA) method, enjoys the properties of uncorrelatedness of resulting components and orthogonality of transform coefficients. In addition, by incorporating spatial and spectral properties among image pixels, the method obtains more accurate classification results with less computational cost. (C) 2010 Published by Elsevier GmbH.