Linking Geospatial Datasets with Different Scales by Conflation

The comprehensive consistent administration of geographic data of different scales has been under scientific investigation for several years. A database which keeps datasets with different scales and links between corresponding geospatial features is called multiple representation database (MRDB). Linking datasets is motivated by the need for propagating updates from larger scale datasets to smaller scale ones in MRDB, which is the key step to build a MRDB. Because the data are captured by different organizations, one object of the landscape is stored in several datasets at different acquisition times, with different quality characteristics and in different scales. Crucial for linking is certainty of equivalence of different object representations. This paper introduces a general structure of a generic conflation approach for linking geospatial datasets with different scale in MRDB, discusses the topological algorithms used in data matching process in detail.