Video based simulation on real-time snow falling and accumulation

Photo-realistic snow simulation is an important topic in computer graphics and computer vision fields, and it is widely utilized in game and film industries. However, existing methods are hardly to be effectively used in reality. Object modeling in terms of computer graphics methods is highly complex and the methods based on computer vision approach are mostly lack of background information. In this paper, we present a video or image based simulation method for realistic snow falling and dynamic accumulation, in terms of depth map image. Also we implement a real-time wind field to achieve realistic physical movement of snowflakes, where motion blurring, image tone adjustment, fog rending methods are introduced for improving realistics. Experimental results show that our method attains good vision effect in real-time.