Meta features for remote sensing image content indexing
K. Seidel, Michael Schröder, Hubert Rehrauer, Mihai P. Datcu · 1998
In this paper we present a general formalism to model several levels of meta information on spatial content. In particular, we expand on a two-level procedure of Bayesian inference: model fitting and model selection. We study five cases of extraction of spatial information from remote sensing images using this scheme. In this way, we put together pieces of information obtained from different levels of interpretation of the original image data.