WG Correspondence

Activities

In the field of multiscale representation of spatial data there are a set of important research challenges and trends. First of all, generalization operations are more and more designed to also operate in a web-environment via web-services. Important issues here are the adequate semantic descriptions of the individual services and their orchestration to solve a complex generalization task. For automatic processes it is of vital importance to evaluate the quality of the results. Methods for quality evaluation but also for integration of quality issues into the whole generalization process have to be studied.

In order to control and trigger generalization operations, not only the objects themselves have to be considered, but also the hierarchical and local context they are embedded in. To this end methods for data interpretation are needed. Such methods are similar to image interpretation techniques, thus there is a close link to working groups in Commissions III and V. A further challenge is the generalization of temporal information, e.g. data collected by moving objects.

City models are important data sets needed for many applications. The derivation of these models, their multiscale-representation e.g. in CityGML, as well as methods for continuously streaming the huge data sets are of enormous importance. Finally, multiscale issues are important for geosensor networks where both local information is collected by individual sensors, which has, however, to be adequately aggregated (generalized) and integrated in order to detect higher level information. These issues are currently being investigated, however, they need deepening and more research.

There is already a well-established cooperation with the ICA Commission on Generalization and Multiple Representation which will be continued.