3D Terrain Dynamic Multi-resolution Modeling And The Key Algorithm

With the rapid development of information society, the geographic information system (GIS) is becoming more and more important. It has become the focus to process the large-scale3D terrain data and real-time visualization in the scope of GIS. In order to satisfy people’s vision in the virtual scene and to experience a virtual world better, the realistic simulations and the frame rate are considered. It is difficult to meet the requirements of these two aspects, although the hardware has been rapid development. It is necessary to study model and reduce the time and space complexity of the algorithm from the perspective of software.In view of the key problems of enhancing the efficiency of data processing and real-time visualization, adaptive dynamic multi-resolution LOD model and based on wavelet analysis dynamic multi-resolution modeling is studied by using three-dimensional terrain rendering theoretical knowledge and related technology. Evaluation strategy of nodes importance and multi-buffering mechanisms based on wavelet transform are proposed respectively, which can provide a scientific basis and guarantee for choosing level of detail scheduling model and scheduling large-scale three-dimensional terrain data.The main research work and conclusions of this thesis are as follows:(1) studying the theoretical basis of three-dimensional space and the relevant technology of the virtual to achieve. Summarizing the current dynamic multi-resolution three-dimensional terrain modeling methods and theories. The advantages and the disadvantages of the different modeling methods are analyzed and compared.(2) Because the size of spatial data is large.In order to solve the problem of low efficiency of data access,blocking data by according to the observer’s line of sight. Reasonable internal and external memory are scheduled through the priority of different data blocks and scheduling policies. Terrain data is scheduled reasonably in real time for requiring, in order to improve the speed of reading data and reduce to access the external memory.(3) In order to smooth transition between levels of different detail.The paper proposes a new method for determining the importance of the node display, thereby establishing a corresponding terrain model details. When the process of building the triangular, triangular segments and simplify in real time is discussed.(4) Using the good local characteristics of the wavelet analysis through telescopic pan and transform operations,the signal can be progressively refined into different resolution model accuracy.The terrain data is two-dimensional discrete signals essentially, so that we can do to deal with similar terrain data signals.Then the terrain data will get a different set of wavelet coefficients through transforming the wavelet.They not only have coefficients that t may reflect the low-frequency part of the overall trend of terrain data, but also have coefficients that may reflect frequency component of partial minutiae of terrain data.Then Wavelet coefficients are filtered by the different threshold filter, we can get the corresponding wavelet with different levels of detail and get the terrain data from the different levels of detail.Then choosing a better wavelet function through the experimental method, discussing the method of conditions filtering and select evaluation model from the level of detail terrain.Finally, a multi-level caching mechanism is proposed so as to achieve a reasonable scheduling data and improve reading speed terrain data.We also discuss structuring real-time quad-tree.The terrain is smooth realistic simulation through experiments.