• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2014, Vol. 36 ›› Issue (09): 1795-1800.

• 论文 • 上一篇    下一篇

可扩展的实时自然景物模拟算法

肖何1,饶云波2,李佳1,邓利平1   

  1. (1.西华师范大学计算机学院,四川 南充 637009;2.电子科技大学信息与软件工程学院,四川 成都 610054)
  • 收稿日期:2013-01-03 修回日期:2013-05-17 出版日期:2014-09-25 发布日期:2014-09-25
  • 基金资助:

    四川省教育厅科研项目(11ZB274)

A scalable real-time natural
scenery simulation algorithm           

XIAO He1,RAO Yunbo2,LI Jia1,DENG Liping1   

  1. (1.School of Computer,China West Normal University,Nanchong 637009;
    2.School of Information and Software Engineering,University of
    Electronic Science and Technology of China,Chengdu 610054,China)
  • Received:2013-01-03 Revised:2013-05-17 Online:2014-09-25 Published:2014-09-25

摘要:

针对传统的粒子系统实时仿真存在只能针对单一自然景物模拟、计算耗时、图像不真实、算法复杂等问题,提出了一种基于粒子系统和图形处理器(GPU)加速通用可扩展的自然景物模拟算法。在该算法中,粒子的物理运动计算过程和渲染阶段完全由CPU转移至GPU,可以增加粒子数量和提高渲染速度;同时,在渲染过程中,可以较好地利用硬件支持的粒子图技术来改善渲染中粒子的外表,选择不同纹理,从而能够较方便地模拟不同的自然景物。最后,在GPU上实现了雪花、喷泉、烟花、瀑布等模拟,算法充分利用了GPU的多通道并行处理性和可编程性,提高了自然景物模拟的实时性,可运用于虚拟现实系统。

关键词: 硬件加速, 粒子系统, 可编程, 实时绘制, 纹理映射

Abstract:

Conventional particle simulation methods have some limitations, such as singleness, timeconsuming calculation, low reality, algorithm complexity, etc. In view of this problem, a novel scalable realtime natural scenery simulation algorithm is proposed, which system, which is based on particle system and Graphics Processing Units (GPU). In the proposed algorithm, the computation process and rendering of particles’ physical motion is on GPU rather than CPU, thus increasing the number of particles and improving the rendering speed largely. Meanwhile, in rendering process, the appearances of particles are improved by the particle graph technique with hardware support, and different textures are selected to facilitate simulating different sceneries (rain and snow, fountain, firework, waterfall). Finally, these sceneries simulation on GPU shows that our algorithm makes use of the parallelism and programmability of GPU, improves the realtime requirement of natural scenery simulation and can be applied in virtual reality systems.

Key words: hardware acceleration;particle system;programmable;real-time rendering;texture mapping