Research On Terrain Data Acquisition & Digital Terrain Analysis In Digital Terrain Modeling

People live on the earth and learn to cope with its terrain, since terrain information plays an important role in many disciplines such as geological, geography, hydrology, disaster monitoring and resource management, et.al., the terrain data acquisition, representation, management and analysis are always the key research points in survey and mapping related disciplines. Digital terrain modeling is an integrated framework for digital terrain model (DIM) related research activities, including terrain data acquisition, management, representation and the terrain analysis, et.al.. Although the research emphasis of digital terrain modeling differed during different period, the terrain data acquisition and digital terrain analysis are always the important research points, since they determined the availability and usability of terrain information respectively.In this dissertation, focused on the aerial image matching based terrain data acquisition and DEMs based digital terrain analysis, some theory, approach and key technical problem of digital terrain modeling are discussed. Based on survey of existing methods, in this dissertation: aimed to the reliability of stereo image matching in digital photogrammetric, a new matching strategy utilized local continuity constraints encoded by triangulation of the well-defined points is proposed; concerned with the inherent efficient, accuracy and size constraint problems of existing topographic structure extraction approach, an efficient, accurate and large scale data adapted approach based on the combination of vector and raster processes are proposed. The dissertation consists the following parts:1) The terrain data acquisition and digital terrain analysis are investigated in the framework of digital terrain modeling: the existing photogrammetry and remote sensing based terrain data acquisition methods are discussed, the aerial image reliable matching method are analyzed; the DEMs based digital terrain analysis approaches are investigated, among which the emphasis is put on the topographic structure extraction based on grid DEMs.2) Aims to demonstrate utilization of the well-defined points information to improving reliability of image matching, this dissertation triangulate the well-defined points and then to consider each triangle as a local continuous area, which is then used as a geometric constraint for image matching. Constrained by the triangle constraint integrated with the conventional epipolar geometric constraint, the local continuity constraint is enforced by changing the constraint areas dynamically according to the local texture density.3) The topographic structure extraction and analysis based on grid DEMs are investigated in this dissertation, the topographic structure extraction approach fills depressions by adopting vector processes combined with traditional neighborhood raster processes, assigns flow direction over flat by applying a neighbor-grouping scan method. In order to improve the result coherence between upper and lower stretch of the watershed, the extraction results are modified according to the Strahler staple of river network.4) Based on the discussion of the management, indexing and dispatch mechanisms for large scale OEMs data, the topographic structure extraction and analysis based on large scale grid OEMs are investigated in this dissertation. The indexing approach based on database-based DEM managementand data dynamic dispatch approach based on data paging are utilized in the topographic structure extraction, the data scale constraint problem of existing file-based DEM management is solved, the inconsistence of extraction result due to block-based processing is avoided.