| The current academe has carried out a number of years for research on spatial data uncertainty as a theoretical issue in geospatial information. The error model of point, line, surface and curve has already been fully researched, but which hardly involved in model error. This paper which aims at the existing shortcomings of error band,discusses this theme around the model error.The thesis has the following main aspects:(1) Summarizing the results achieved in positional uncertainty of spatial data, analyzing main units on positional uncertainty of curve elements and obtaining uncertainty band which only takes into account the model error is relatively tight uncertainty band.(2) Based on the basic line of geographical entities, classifing the line segment specialy, pointing out that the article is to explore for the deterministic linear features. In view of the the error band of existing linear error model takes on two-side big and mid-side small, putting forward the concept of model error, analyzing the inevitability and importance of model error.(3) Based on the uncertainty theory and probability theory, discussing model error and its specific parameters of circular curve and secondary curve in the regular curve, and analyzing the model uncertainty of irregular curve .In the end, realizing the visual model of regular curve and irregular curve with model error using MATLAB.In summary, this paper proposes the concept of model error,which is expressed in visual form, through analyzing the existing shortcomings of uncertainty model.The research has not only settled the problem of uncertainty presentation, but also has enriched the uncertainty theory frame of spatial data. All in all, the work which has been done is limited . The author believes in the facts that all the works accomplished in this paper will promote the research area of spatial data uncertainty theory. Finally, the limitetions of the current study are pointed out and future works are outlined.
|
PDF Full Text Request