| Spatial information processing is a multi-disciplinary intercross fusion field, the spatial information has characteristic of numerous data amount, various type, strongly time-effectiveness,rich information quantity and visual presenting mode., and so on. In economic and social research, and many other aspects of spatial information more and more raster data was needed,and demand is also growing and the increase of spatial data processing and application in the accuracy and timeliness and other aspects of the more higher requirements. As an important group of spatial data organization - raster data, raster data of domestic and foreign different organizational structures, satellites and satellite raster data structure is different from the same satellite sensors with different grid structure of the data is not organized the same. A wide range of raster data organization, format, complexity, in practical applications often require the organization of these different work raster data. Reduce the space of data processing systems rely on spatial data format has important theoretical and practical value. In this paper, the unification of the raster data processing as the main line, for more uniform processing of raster data sources that exist in a number of key issues. In this paper, the main contributions to the work were come to the point as following:1, A unified multi-source raster data processing model was proposed. At the first part of this paper ,I analyze the characteristics of raster data structure, encoding and compression schemes, and then for the two currently widely used format:hierarchical label file format and file format, called TIFF and HDF, four kinds of (GeoTIFF and Big TIFF, HDF4 and HDF5) organization of different raster data analysis, with the integration model theory of spatial data, spatial information processing will be required to extract the common elements for spatial information processing system to provide uniform data-processing interface. 2, An improved thin plate spline (Thin Plate Spline referred to as TPS) interpolation algorithm was presented. In this paper, some general interpolation algorithm were studied, focus of the inverse distance to a power interpolation, Kriging and Minimum Curvature method. Thin plate spline interpolation algorithm is an improvement of the minimum curvature interpolation. Among of the interpolation algorithm, TPS interpolation algorithm is the more suitable for spatial interpolation, but at present, this algorithm is not suitable for a wide range of interpolation, block policy based on artificial specified, in this paper, an improved TPS interpolation algorithm was proposed to improve the TPS interpolation algorithm, one aspect of improvement was the blocked policy, that is changed based on block the manner specified by the user into the decision by the system's hardware environment, the other one was a wide range of data interpolation algorithm of main memory and secondary storage to support collaborative work, while no increase in algorithm complexity.3, A prototype system was developed which named by multi-source raster data processing integration prototype system. In the prototype system, the functions were realized the unity of GeoTIFF and HDF data processing, and presented a three-dimensional display of data capacity, in the prototype system has achieved such as the affine transform, resampling, projection transformation etc.4, In the end of this paper,To feed the requirements of infra-red remote sensing images of remote sensing products and the environment subsystem needs, the former work has been used in these projects. The results of system test indicate that spatial information can effectively reduce the raster data processing systems rely on specific format, improved processing system development productivity, better meet the user access to the grid, according to performance requirements. |
PDF Full Text Request