节点文献

虚拟现实中大规模粒子系统的研究

Research on Large Particle System in Virtual Reality

【作者】 田苏昕

【导师】 石祥滨;

【作者基本信息】 沈阳航空工业学院 , 计算机应用技术, 2009, 硕士

【摘要】 在计算机虚拟仿真领域,应用粒子系统模拟不规则模糊物体的方法已经得到了广泛应用。随着游戏玩家对场景环境要求的日益提高,粒子系统已经成为计算机游戏领域不可或缺的关键技术。然而,传统的粒子系统在整个场景使用单一的运动方程来控制粒子运动,对于和视点距离不同的粒子而言,增加了很多不必要的计算量,影响了绘制速度。因此,研究出一种对粒子运动比较灵活的控制方法,对于提高绘制速度和模拟真实度方面具有积极的意义。本文通过对粒子系统的研究,把基于空间感知的AM模型应用于粒子系统中,用来解决整个场景中粒子运动方程单一而带来的运算量大和模拟效果不好等问题。本文主要完成了以下几个方面的工作:1.将基于空间感知的AM模型引入粒子系统绘制中,对全部粒子进行层次划分,克服了传统粒子系统对粒子运动的控制所采用单一运动方程,带来了很多不必要的计算量等问题。本文按照感知强度的强弱将全部粒子进行层次划分,针对不同的层次来采用不同的运动方程控制粒子的运动。2.在按照层次关系优化粒子运动方程的基础上,本文依据GPU特点设计了二级缓冲,解决由于粒子数目巨大而给CPU带了的负担,同时,将对粒子属性和运动方程的计算于GPU内完成,这样做可以更高效地对系统资源进行利用。同时还采用公告板技术(BillBoarding)对粒子系统的绘制效率进行优化。在Windows XP操作系统中,使用Visual C++、OpenGL和Direct3D编程对本文方法进行实验,分别针对大规模场景中的雨和雪两种不同模型进行模拟。实验证明,本方法对粒子进行合理的层次划分,根据被模拟现象的不同设计出适合各个层次粒子的运动方程,并在GPU上采用二级缓冲技术并行计算,提高了系统的拟真度,减少了CPU的运算负担,最终克服了传统粒子系统粒子数目过少不能适应大场景下的应用以及传统粒子系统绘制效率低等问题。

【Abstract】 The application of Particle System Simulation irregular fuzzy objects has been widely applied in the field of computer Virtual Simulation. While the increasing requirements of the game, Particle System has become the key technology. But for the particle which has different distance from viewpoint, traditional Particle System in the entire scene using a single particle equation of motion to control particles` movement, which increases the calculation and impact the rendering speed. So the development of a more flexible controlling the movement of the particle method, is for improving the rendering speed and virtual reality in terms of positive significance.Through the research of Particle System, the AM model, which is applied to Particle System, used to solve the problem of redundant computation which is produced by single equation of motion in the whole scene. It also can bring a great simulation result. We work about the following aspects:First, by bringing the AM model in Particle System, and designing the level of particles, we solve the problem of redundant computation which is produced by single equation of motion in the whole scence.Second, We design a kind of double buffer in GPU, so that we can improve the using of system resources. We also use a technology named BillBoarding to improve rendering efficiency of Particle System.We use Visual C++, OpenGL and Direct3D to complete the experiment in Windows XP. By rendering rain and snow, the experiment proves that this method can improve the rendering efficiency and the reality of Particle System.

【关键词】 粒子系统AM模型GPU纹理映射二级缓冲
【Key words】 Particle SystemAM ModelGPUTexture MappingDouble-Buffer
节点文献中: 

本文链接的文献网络图示:

本文的引文网络