Stochastic Extension Method for Animating Water Flows

We present a simple and efficient method for animating different types of water flows with emphasis on a surging wave. The method employs a stochastic approach of extending 2D simulation data into 3D animation space. For the reality of a wave motion along the direction of its dominant propagation, we utilize a physically-based, particle simulation at the pre-processing step and extract surface and splash particles only. At the animation step, these 2D surface frames are stochastically sampled along the transverse direction in the animation time and slice (called time-slice) domain with the aid of a user-defined noise function. For the surface reconstruction in 3D, we use a geometrical technique with a smoothing filter in order to remedy undesired 2D artifacts that result from the slice sampling. In conjunction with appropriate supplements, we apply our method to other types of water flows, indicating further interesting applications of a ripple wave and a whirlpool rotation.