A HIERARCHICAL QUALITY-DEPENDENT APPROACH TOWARD ESTABLISHING A SEAMLESS NATIONWIDE TOPOGRAPHIC DATABASE

摘要

Nationwide geospatial databases in general and topographic ones in particular are today one of the most common infrastructure for mapping and other geo-related tasks. These databases are designated to establish an adequate, continuous and if possible homogeneous representation of our natural environment. New data acquisition technologies, which present high accuracies and resolution levels that were not known until recently, yield rapid and frequent updating of existing nationwide databases. This enables the generation of a multi-source mosaiced database that is multi-quality as well, i.e., introducing varied accuracies within its coverage area. Simultaneous analysis, such as integration, of two or more of these nationwide databases will evidently present multi-scale spatial inconsistencies. These are a function of various factors, among them the different levels of accuracy within each database. Common height integration mechanisms will not suffice here. This paper presents a framework for dealing with the problems and considerations in utilizing topographic databases that are quality derived while trying to give a solution to the existing geometric ambiguities. A conceptual new algorithmic approach is detailed, which relies on a hierarchical modeling mechanism that is designated for extracting the existing varied-scale discrepancies in order to produce a common geospatial framework. Moreover, designated quality-derived constraints are implemented in the process to ensure that accuracy is preserved. This novel approach proved to be accurate while producing seamless topographic database that retained the level of detailing and accuracies presented in the source databases, as well as local trends and morphology.