Research On The Method Of Terrain Scene Data Organization Based On GRID/TIN Hybrid Structure

Traditional Digital Elevation Model (DEM) is usually represented by GRID, but this kind of data structure cannot represent the terrain surfaces after embedded with vector data. If simply used irregular network to represent the whole DEM and embedded vector data, the most area unaffected by embedded vector data not only did not improve data accuracy, but also lost the advantage of the data structure. This thesis will use an integrated TIN and GRID method, make full use of the simple date structure of GRID and the advantage of the Triangulated Irregular Networks (TIN) about detail characteristic, to research how to organize and express the digital terrain scene embedded with vector data.The main work of this thesis is as follows:1) The integration method of TIN and GRIDResearch how to embed the vector data into DEM to construct a hybrid surface model integrated GRID and TIN, realize the effective integration of multi-source heterogeneous terrain and vector data.2) The organization and management of hybrid terrain data.Research the data structure of hybrid terrain data and the organization methods of vector data, build the spatial index of data in memory, make the scheduling scheme of data, and achieve the effective organization of terrain scene data.3) Construct multi-resolution surface models, based on the hybrid terrain of GRID and TIN.Research the view-dependent and real-time simplify method of GRID-TIN data. Under the limitation of the threshold of simplification error, simply terrain meshes dynamically based on the view-dependent algorithms so as to reduce the geometric complexity of the terrain scene.Based on the research work, this thesis use a hybrid terrain data structure based on GRID-TIN, realize the integration of vector data with DEM data. This method can block the terrain data and storage with blocks, support data extension and the scheduling of large terrain data, and also can support to add embedded vector data information into the hybrid terrain data. During the terrain visualization, use the Real-time Optimal Adaptation Mesh (ROAM) and Progressive Mesh (PM) real-time simplify algorithms to simplify the GRID blocks and the TIN blocks. Realize the multi-resolution of the terrain in real-time and dynamically. This thesis use the DEM in 5 meters resolution of Mount Lu and the vector data in 1:10000 scale for experimental verification and analysis. The experimental results validate the thesis proposed method that can effectively support embedded vector with the GRID-TIN hybrid data and support real-time dynamic multi-resolution modeling of hybrid data.