Research On The Method Of Multi-source Spatial Data Integration

Along with wide and deep application of GIS, the demand for spatial data becomes greater. To establish the required GIS, we usually expend large human material and financial resource to collect various data, thus the quantity of spatial data becomes very large. Basing on its application demand, each department chooses its own GIS software platform, data model and projection to collect data. Spatial data varies in model, structure and semantic. In order to share massive data with different formats better, according to the multi-semantic, storage format diversity of spatial data, this paper researches on method of multisource data integration.This paper puts forward a method for multisource data integration based on general spatial data model, builds corresponding relations between general spatial data model and each data model, establishes "data channel" between each two data formats. In this way, any transformation and expression between any two data formats can be realized without losing information at maximum. When building mapping relations between GIS data models and general data model, the method based on middleware technique is adopted to integrate multisource data into a common platform through corresponding conversion algorithm. The main points of this paper are as follows:1, Explained the theories of multisource spatial data integration based on spatial data model integration, geographical feature semantic integration and spatial reference unification. Summarized the mode and new technology applied in spatial data modal integration. On the foundation of analyzing the causes of semantic heterogeneousness, we introduced semantic integration method by resorting, recoding and the other method based on ontology. From coordinate system transformation, projection transformation and altitude datum unification, we realized unification of spatial reference.2, On the foundation of analyzing typical spatial data model characteristics, comparing the merits and shortcomings of common GIS data models, deeply researching on the geographical spatial data of OpenGIS, we designed a general spatial data model and gave elaboration on its framework, spatial data structure and characteristics.3, Taking DLL and middleware technology, we built the mapping relation between each data modal and general data modal. Gave solutions to the problems may appear in the process of data format transform .Realized data integration of different GIS data format.4, Based on general data modal design and data integration technology, independently developed multisource data integration platform system, on which we realized different format data integration.