Parallel Dynamic Cloud Rendering Method Based on Physical Cellular Automata Model

Abstract

SUMMARYDue to the complicated composition of cloud and its disordered transformation, the rendering of cloud does not perfectly meet actual prospect by current methods. Based on physical characteristics of cloud, a physical cellular automata model of Dynamic cloud is designed according to intrinsic factor of cloud, which describes the rules of hydro-movement, deposition and accumulation and diffusion. Then a parallel computing architecture is designed to compute the large-scale data set required by the rendering of dynamical cloud, and a GPU-based ray-casting algorithm is implemented to render the cloud volume data. The experiment shows that cloud rendering method based on physical cellular automata model is very efficient and able to adequately exhibit the detail of cloud.