Study On 3D Visualization Method Of Weather Radar Data Based On Surface Rendering

Meteorological observation has always been an important measure for mankind to improve production quality and living standards.Weather radar,as a common tool to predict the weather,has made important contributions to the social development.In order to make the cloud data detected by weather radar play a more important role,people begin to study the 3D visualization of weather radar data.On the one hand,it is conducive to the intuitive understanding of the meteorological map.On the other hand,it also provides convenience for researchers to carry out in-depth research in the field.According to the feasibility and requirements of weather radar data 3D visualization,this thesis chosen the Marching Cubes algorithm,referred to as MC algorithm to implement surface reconstruction for weather radar data from a number of 3D reconstruction algorithms.Owing to distribution of radar data is uneven and radar data is located on the cone,this thesis carried on the data processing of the original gridded radar data for matching with the MC algorithm.Meantime,this thesis proposed the Cube Weighting Interpolation(CWI)method that is suitable for uneven data transform into even data.The main steps of the CWI method are including:Firstly,normalizing and mapping txt data to 3D space.Next,establishing standard 3D data grid,and using the weighted interpolation method to complement the rest points.Finally,correcting and optimizing interpolation data.In this thesis,Delaunay triangulation and Crust algorithm in the surface mesh reconstruction algorithm are introduced,and the two algorithms are employed to implement 3D reconstruction of weather radar data.In the process of reconstruction,it is necessary to point out that the implementation of the two reconstruction algorithms should be combined with CWI and mathematical morphology to extract the edge of meteorological radar data as their input data.After comparing the reconstruction results with the MC algorithm,result shows that MC algorithm is superior to the other two algorithms both in reconstruction accuracy and reconstruction rate.Finally,for the sake of facilitating the operation and display,this thesis employed MATLAB to achieve a simple 3D reconstruction system with the functions of data reading,data pre-processing and 3D reconstruction display.The system can show the reconstructed cloud features in four levels of horizontal and vertical,different contours,different times and realistic rendering,and provide a reference for the weather forecasting work to a certain extent.