Data compression-a possibility for the integration of remote sensing data in GISs

One main goal of processing remote sensing data is the classification of these data to derive thematic maps and to include those in geographical information systems (GIS). In this study the authors focus on RADAR-data. The classification of RADAR-data shows various advantages and disadvantages compared with a classification based on optical data. Main advantages are the independence of atmospheric conditions, the independence of illumination and the penetration depth, which allows an observation of volumetric properties and surface properties. A detailed explanation of the advantages is given elsewhere. Disadvantages arise from the special geometry of RADAR-systems (forshortening, layover) and the ambiguity of the signal interference patterns (speckle). Speckle filters are implemented in various image processing software packages, but an improvement of speckle reduction techniques is necessary to allow an automatic classification. Taking into account that speckle can be regarded as a stochastic process and not as information, it is obvious to treat the processes of data compression and speckle removal as one problem. Therefore the author gives an overview on compression techniques and the coding scheme used in chapter 2. Chapter 3 describes the integration of algorithms in a commercial GIS/remote sensing package. In the final chapter 4 a glance on applications is made.