超大型三维场景分布式渲染系统体系结构与技术研究  

作　　者：郭阳 李昆昆 于春雨 卢石磊 李文博[1] 刘甜[1] 

机构地区：[1]青岛理工大学复杂网络与可视化研究所,山东青岛266520

出　　处：《科技创新与应用》2022年第17期23-26,共4页Technology Innovation and Application

摘　　要：虚拟仿真、科学计算可视化等领域针对三维场景渲染的需求趋于精细化,导致场景模型体量庞大,单台PC机无法快速渲染超大型三维场景。文章采用分布式渲染方法,设计服务节点+主渲染节点+从渲染节点的分布式渲染系统体系结构,并加入并行机制和子任务划分策略完成整个系统的渲染工作。以500米口径球面射电望远镜(FAST)模型作为测试用例,结果表明,文章所设计的分布式渲染系统在保证实时渲染流畅度的前提下,能够有效减少三维模型的整体渲染时间,提高渲染效率。The demand for 3D scene rendering in virtual simulation,scientific computing visualization and other fields tends to be refined,resulting in the huge volume of scene model and the inability of a single PC to render super-large 3D scenes quickly.In this paper,the distributed rendering system architecture of service node+main rendering node+slave rendering node is designed by using distributed rendering method,and the parallel mechanism and sub-task division strategy are added to complete the rendering work of the whole system.Taking the Five-hundred-meter Aperture Spherical Radio Telescope(FAST)model as a test case,the results show that the distributed rendering system designed in this paper can effectively reduce the overall rendering time of 3D models and improve the rendering efficiency under the premise of ensuring the real-time rendering fluency.

关 键 词：分布式渲染 主从结构 并行机制 子任务划分 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]