Error Analysis And Representation Of Multi-resolution Modeling

Model simplification and multi-resolution expression as one of the key technologies in the multi-resolution modeling, is also a hot problem of research on DEM recently. In this paper, against the deficiency of multi-resolution modeling, with a series of analysis and data processing methods, such as mathematical statistics, information and visualization theories, I pursued systemic analysis and research from two aspects, which are the analysis methods and the objects of error research. So this paper has its own value both in theory and techniques.(1) In the paper bellow, I difined the concept of multi-resolution modeling error, then put forward the affecting factors, measuring methods and precision indexes of it, and summarized methods and techniques fit to error analysis of multi-resolution modeling.(2) With the DEM data of the scale of 1:50000 in the southeast hill regions, adopting of the simplifying arithmetic based on Wavelet transform theory, I complemented the visualization expression of RSG simplification, and analyzed the spatial distribution featuress and self correlation of the error. In addition, I described the distribution features and information content changes of topographic feature genes, such as elevation, slope and aspect, by the terrain simplification, and analyzed the rules by which the terrain features affect the simplification error.(3) Based on four kinds of typical physiognomy DEM data, such as loess, glacier, medium mountain and hill, adopting sample method, according to the extraction theory of error simplification, I quantized the relationship between elevation error, slope error and the resolutions of simplification modeling in the process of gradual terrain simplification, then I researched the statistical routines and distribution features of slope matrix in series of resolutions. Furthermore, I shortly analyzed the change trend of the aspect and curvature genes, moreover, I pursued some comparative analysis on the differences between DEMs of different scales in terrain simplification.The research in this paper provides theories and technique reference for error analysis based on multi-resolution modeling. It is helpful to guide the multi-resolution modeling anddecision-making based on models, with great significance both in theory and practice.