Large-scale Scenes In The Simulation Of Virtual Groups To Simplify

Large-scale groups animation has always been one of the hot research area of virtual reality technology.Group animation is widely used in games, movies, cartoon, architectural planning and disaster simulation (like fire) and assistant conductor .Group behavior through real simulations could be assisted films have a real effect of scenes; verify secure channel can be simulated to determine the reasonableness of such design in public places (shopping malls, sports stadiums, entertainment venues, etc.); simulated environment is made available to the intelligent traffic management (the effective settings of traffic lights, trunk road junctions traffic flow management); guidance and training platform has been provided to the disaster scene. The research has been widely appreciated, the research and the creative tool of group animation has a great demand.At the same time a complete and efficient group of animation design and production tools will also facilitate the connection of digital media industries.As a key technology of animation, model simplification can not be replaced. To achieve large-scale scenes, there will inevitably be involved in drawing and rendering a large number of models, but for real-time scene rendering, the rendering speed and visual fidelity models are two very important indicators. In order to achieve the harmonious unification of two indicators, we must choose a good algorithm to get the trade-off of the rendering speed and visual fidelity.For the static model, the technology has achieved a very good model simplification. The model simplification algorithms can be divided into four main categories: vertex clustering, incremental simplification, the sampling algorithm and adaptive subdivision. Based on these four algorithms, the static techniques has developed very perfect,such as The Reduction Technique of Large-Scale Animated Crowd based on the level of detail models and The Reduction Technique of Large-Scale Animated Crowd based on point sampling. This paper has done a detailed research and analysis of above algorithms,and generate a new algorithm: The Reduction Technique of Large-Scale Animated Crowd Based on Point-Sampling.While the static scene of the simplify large-scale technology has been relatively satisfactory, but the dynamics technology is only just the beginning. In view of this situation, in the new algorithm, we add a new method to simplify the dynamic model. The results show that: in the condition of ensuring the scene rendering speed and visual authenticity, our algorithm can get a real-time animation with 30,000 people.