Research on change classification description and identification of corresponding area objects in multi-scale maps

In order to keep geospatial data up-to-date in topographic databases, change detection and data updating is required. Change detection plays a key role in propogating changes from updated larger-scale maps to to-be-updated smaller-scale maps level by level. In order to analyze changes between maps at different scales and time epochs, a new method is proposed for the change classification, description, and identification between corresponding areal objects at two different scales in this paper. First, literatures related to geographical information change are summarized. Then, change classification of area objects is categorized into 9 kinds of types, i.e., appear, disappear, enlarge, shrink, move, rotate, split, merge and merge after split. Third, formal representation and natural language description for each type of changes are presented. Fourth, with consideration of the false changes caused by cartographic generalization operations (e.g., aggregation, typification), the 4-intersection-difference model for topological relations is further used to identify actual changes from geometric discrepancies produced by cartographic generalization operations. Final, some experiments on residential data at scales of 1 : 2, 000 and 1 : 10, 000 are performed. Experiments demonstrate that the proposed method is able to identify different change types effectively, providing a promising solution for distinguishing real-life changes and discrepancies resulted from cartographic generalization operations.