Research On Smilarity Degrees Between Two Individual Building In Multi-scale Map Space

Map data multi-scale spatial expression is recognized as a problem in the community of geographic information science. In recent decades, geological experts at home and abroad have put a lot more efforts into this issue and get a great many achievements in this field. At present, the technology of map automatic comprehensive still cannot fully support the automatical-construction of multi-scale map databases.The theory of space similarity not only provide important theoretical foundation for automated quality assessment of generalized maps, effective method for calculation space similarity degree before and after map generalization, but also the key conditions for program termination of map generalization algorithm.Residents are the important element on the map. There are much research on buildings' map generalization algorithm, and the algorithm of individual residents is more effective than the group's. At present, researchers have only started to explore the similarity calculation of polygon element before and after map generalization. Therefore, individual resident is the research object in this paper. On the basis of researches on this field, deep analysis of space similarity theory and buildings'map generalization algorithms, this paper describes and calculates the similarity degree between arbitrary two individual buildings. The specific tasks include the following several parts:1. Systematically analyze four constraints and algorithm for multi-scale expressions of individual buildings, and then research their effect on the space data.2. In the study of arbitrary two individual buildings'graphic similarity, this article selects related description parameters including shape factor, contour graphic center of gravity and narrow degrees of graph, to measure graphic similarity by using mathematical morphological method.3. Through in-depth analysis of five calculation method for individual buildings'internal direction, this paper finally adopt minimum binding rectangular method to describe and calculate the similarity degree of internal direction between arbitrary two individual buildings.4. In the study of arbitrary two individual buildings'structure similarity degree, by the use of reduce dimension theory, skeleton line replaces polygon element. Through overlaps detection, calculation area between two skeleton lines, numbers of skeleton line's branches, this paper use these parameters to calculate structure similarity degree. 5. Part Four is the demonstration of the method. The experiment indicates that the result of the method is in accord with human cognition in map space, and it can express similarity degrees of buildings in graphics, structure and internal extension direction before and after map generalization.Finally, i.e. part five, this paper summarizes the research achievements, and discusses further research problems.