Restoration and Validation of Image Data

Abstract

Images supply large amounts of data, that need appropriate statistical and numerical techniques, in order to achieve their restoration and validation. In this work some procedures of data processing are presented; they combine suitably optimality from the statistical point of view and practicability from the numerical one. Furthermore they have been applied, since a relatively long time, to surface reconstruction and deformation monitoring, but they are now specialized and applied to preprocessing of image data (i.e. data assessment, image quality control). A pilot experiment using a SPOT image has been done and its results are reported.