| Image simulation is an effective method of validating the technical feasibility of Earth observation based on the new sensor, whose hot and difficult points are 3D ground simulation modeling scene. Therefore, the top priority of image simulation is to build 3D modeling scene rapidly and accurately. This thesis is aimed at finding a way to build 3D modeling scene and realize visualization effectively.Author builds 3D modeling scene taking houses, trees, grass, roads and other features in the west of Institutes of Remote Sensing Applications, Chinese Academy of Sciences of the Beijing Olympics Science Park as study objects. Application of color theory is to build the functional relationship between the color values and spectral reflectivity and obtain the feature RGB values as input parameters for POV-ray photons tracing.(1)Program based on MATLAB to build 3D modeling scene. Since different features with different radiation characteristics, 3D scene should be stored as different triangular text files according to different features. In order to solve problems of low computational efficiency caused by the huge amount of data, author writes the programs of scene extension.(2)The introduction of color theory and building the converse relationship of color coordinate systems is to obtain the feature RGB values as input parameters for visualization.(3)Software of POV-ray photons tracing is applied to simulate the imaging process of remote sensing sensors. Through the coordinate transformation, information read by POV-ray is stored to triangular text files. In order to realize the 3D scene visualization, color theory is utilized to get the feature RGB values and validate the rationality of scene simulation with analysis of thermal radiation characteristics.(4)Describe the building and utilization of feature model toolbox, which contains methods of modeling acquired by this study. The main aim is to generate units of different feature models as soon as possible. |
PDF Full Text Request