Research On 3D Visualization Methods For Dynamic Cloud

In natural world, cloud is one of the most complex research subjects. The emergence and development of cloud are caused by the transition state of atmospheric physics and the changes of energy. The cloud is the indispensable elements of the earth’s weather system. The simulation of cloud is a hot research topic in computer graphics, which is widely applied in the outdoor scene simulation, studio entertainment and the analysis of atmospheric. In the simulation of outdoor scene, it will greatly enhance the fidelity of the whole scene with the simulation of cloud. As the important meteorological elements, cloud has a direct relationship with the change of the weather. The simulation of 3D cloud, which based on real data can not only convey the real weather information, but also has good auxiliary function of meteorological service. In recent years, the cloud simulation based on real data has been paid more and more attention.Research on 3D cloud is mainly at the aspect of model and rendering algorithm. This paper is based on the meteorological data of numerical simulation. It separately improves the methods of cloud modeling and rendering, which enhances the rendering speed with the guarantee of rendering quality. The main research contents are as follows:At the aspect of cloud model, due to the problems of low drawing speed and high memory cost for the traditional ray casting algorithm in 3D cloud simulation, the paper presents a numerical 3D cloud simulation method based on adaptive ray casting algorithm. Firstly, the algorithm using octree to divide the volume data into data blocks which based on traditional ray casting algorithm. Then it optimizes and pre-calculates unit gradient of volume data. Finally, according to the number of blocks and the importance of data blocks are calculated to extract the sampling frequency which can realize the adaptive sampling. The improved method shows that the rendering speed is significantly improved with the guarantee of rendering quality.At the aspect of cloud rendering, currently existing 3D clouds simulation illumination models are confronted with the shortcomings such as complexity of implementation and low render efficiency. Herein a calculation method for illumination model based on importance sampling is presented; Aiming at the problems of the phase function, a new phase function is presented to improve the rendering efficiency and quality. By the design of the importance function, the method does importance sampling in integration interval and combines two phase functions to render the cloud. The results of the simulation show that the presented method enhances the rendering speed with the guarantee of rendering quality. The results of the visualization can display the physical characters of the cloud both in rendering efficiency and quality, ensuring perfect visual effect.