A MATCHING APPROACH FOR THE INTEGRATION, CHANGE DETECTION AND ADAPTATION OF HETEROGENEOUS VECTOR DATA SETS

Due to different acquisition methods and not synchronized updating periods the topographic content of geoscientific data sets and the German digital topographic map (ATKIS) differs in geometry, accuracy and actuality. In former times these differences between printed analogue maps were not as apparent as today, when different data sets are overlaid in modern GIS-applications. Integrating different data sets – in our case topographic and geoscientific data – allows for a consistent representation and for the propagation of updates from one data set to the other. To enable the integration of these data sets a workflow is in development, based on harmonization, geometric alignment, change detection, and updating. These steps are necessary to ensure consistency, but they are hardly practicable when performed manually. Taking into account that corresponding objects from different data sets have been acquired at different points of time, it is obvious that parts of the geometry have been changed and that the comparison based only on the object geometry using standard algorithms will lead to unsatisfying results. The manual examination of the different data sets shows very obvious similarities in large parts of corresponding objects geometry. A simple matching using overlay for the selection of candidates is recorded as an XML relation set. According to these relations different geometric alignment strategies have been implemented and evaluated which allow the adaptation of the geometry. Based on different parameters and thresholds these strategies are capable of aligning geometries to avoid discrepancies, but they still enable the identification of changes which occurred between different acquisition steps. Beside the automated alignment and change detection, the automatic derivation of an updated geoscientific map is possible, which is performed using a rubber-sheeting transformation based on the different alignment strategies. This paper shows the integration workflow but concentrates on the different alignment methods. BACKGROUND