| Summary: | 碩士 === 國立臺灣大學 === 土木工程學研究所 === 100 === Map is one of the most intuitive tools in recording and representing the geographic information. Under the consideration of preserving cartographic data in a compact and efficient manner, digital maps composed of vectors are mostly used. Also, users can apply coordinate transformation to analyze data with different datum definitions. A complete coordinate transformation analysis includes model choosing, observation using, parameter estimation, and quality assessment. Typically, control points are used as observables to solve transformation parameters, while check points are used for a quality assessment. As a consequence, poor distribution or an insufficient number of control/check points might lead to a biased transformation solution. In this study, linear features and projective invariant points were both used as observables in a coordinate transformation analysis between multiple vector maps. The goal was to provide a better geometric constraint for the transformation. Furthermore, two numerical indices, namely the absolute consistency and relative similarity, were used for evaluating the quality of a transformation solution. Based on the case studies using both simulated and real datasets, it has been proven that the proposed approach is capable of fully making use the geometric connotation inherent in a vector map and providing a comprehensive quality evaluation on the obtained transformation. Consequently, a more reliable and robust integration analysis for digital vector maps can be achieved when the proposed approach is implemented.
|
|---|
