The Study Of Comprehensive Model Of GIS Irregular Curve Position Uncertainty

Spatial data uncertainty as a basic theoretical issues of the geospatial information。To address this issue, it has many years for its research in Academia,large number ofstudies have been carried out on the point, line, surface, and the curve of the error model.But few take into account the impact of model errors.with the existing error band ofdefects,combined with model error,we studying the uncertainty of the location of theirregular curve model.The thesis includes the following:(1)Summarying the research findings of the uncertainty of the location of the dataspace field.and analysis the location uncertainty of the curve immportantly, explainingthe impact of model error for the uncertainty of the curve,we drawn that it is morereasonable to consider the model error uncertainty band of GIS irregular curve.(2)Studying the location uncertainty of the irregular curve, fitted cubic splineinterpolation method for irregular song,and analysising the connection point and the point ofmaximum curvature of irregular curve.we obtain that when collect points that havespecial meaning in the digital curve which can reduce the impact of the error effectively.(3)Describing the impact and the necessity of existence of model error for locationuncertainty of spatial data,and establishing integrated uncertainty model of taking intoaccount the location uncertainty and model uncertainty at the same time. It is differentfor existing line model error band.(4)Based on uncertain theory and probability theory,by irregular curve of the cubicspline fitting and practical examples,analysising the irregular curve modeluncertainty,finally,programming the example of gived by matlab6.5.We come to that thevisualization of location uncertainty that taking into account the model error and doesnot.In summary,we study the existing uncertainties that are not taken into accountspecially in this paper,considering the impact of model error and its quantitativerepresentation and visualization.This research is negligible for the study of spatial datauncertainty.There are a lot of follow-up work to be done.