Analysts of multi-resolution geographic databases are particularly interested in detecting changes (or inconsistencies) among the different representations for the same geographic area, or they are interested in detecting qualitatively similar configurations at different geographic locations. We have chosen a method that formalizes the semantics of geographic data from which spatial similarity can be derived. The concepts of the 4-intersection and its component invariants provide a solid framework for a systematic assessment of qualitative spatial similarity. We will extend this model to account for qualitative metric properties of spatial relations, and will develop formal models for assessing spatial changes. We will develop a model for assessing similarity of spatial scenes and study which of the concepts developed for 2-dimensional space apply to 3-dimensional models. We will develop software prototypes to demonstrate the validity of our approach and test it with data obtained from university campus maps.
Last updated on May 10, 1997.
[ Geographic Databases | Spatial Database Research Group | NCGIA Maine ]