| Summary: | 碩士 === 國防管理學院 === 國防資訊研究所 === 88 === Today, the Digital Terrain Model (DTM) has become one of the important data resources in many application areas. For military application, DTM not only could offer a ŗD visualization" battle field management, but also could offer a more reality scene for VR-training. Even for the crusing missile, it also uses DTM as a cruse guide. In order to integrate DTM with other artificial satellite images more efficiently, it is inevitable to proceed DTM via an effective interpolation method. Because of a earth surface can be represented by a "fractal signal" or a "multi-frequency signal", and wavelets are easy expression of multi-frequency signals (include high & low frequency). Therefore Wavelets Theory have successfully and efficiently applied to such terrain representation.
The subject of this research is to compare three different interpolation techniques：Haar transform, Daubechies D2 transform and Cubic Convolution for generating DTM. Using relative error method and Root Mean Square error method to evaluate the results. Haar transform is shown to be the best of the three interpolation methods. The result could be used in ROC 3D Information Warfare Management System as a war field representation.
|
|---|
