The signal received by any remote-sensing instrumentation is influenced by the spatial structure of the observed surface. For example, a rough surface causes scattering. In turn, surface properties such as surface roughness or regular spatial structures may be identified from the remote-sensing observation.
In the first part we introduce approaches to characterization and classification of surface structures, based on spatial structures functions and using deterministic and connectionist association. Examples of surfaces will include seafloor, snow and ice surfaces, sensor types will include radar, sonar, passive microwave and field data. The salient point in providing a range of instrumentation and a range of sensors is to illustrate the dependence of the mathematical difficulty of the problem on scale, on the relationship of sensor support to scale, on the complexity of the surface structure and on the error associated with data collection.
More generally, the problem has two faces: