A statistical approach to the fusion of spectral and spatio-temporal contextual information for the classification of remote-sensing images

In this paper, an iterative statistical approach is proposed to fuse spectral information with spatial and temporal contextual information for the classification of multitemporal, multisensor remote-sensing images. Experimental results show, in terms of classification accuracy, the improvement that can be reached by exploiting the contextual information. (C) 2002 Elsevier Science B.V. All rights reserved.