应用GPU编程的水声对抗视景系统关键技术  

Visual realization of an acoustic confrontation simulation system using GPU programming

作　　者：马天[1] 黄建国[2] 高伟[2] 张群飞[2] 

机构地区：[1]西安科技大学计算机科学与技术学院,陕西西安710054 [2]西北工业大学航海学院,陕西西安710072

出　　处：《哈尔滨工程大学学报》2012年第3期289-295,共7页Journal of Harbin Engineering University

基　　金：国家自然科学基金资助项目(60972152);陕西省教育厅专项科研计划资助项目(09JK577)

摘　　要：针对水声对抗复杂环境的实时细致再现问题,应用图形处理器(GPU)编程技术,基于开放场景图(OSG)视景引擎研究开发了一套通用的视景仿真系统.设计了系统的体系结构和场景组织,利用投影网格模型和GPU上快速傅里叶变换的方法生成海浪网格;采用纹理映射技术实现了海面的反射和折射效果;并实现了海底的光束、浮帘粒子和模糊效果;依据后续渲染技术模拟了场景的高动态范围(HDR)效果;基于FMOD音频引擎开发了OSG声音节点类,实现了物体三维声音、事件声音和背景声音的播放;基于多播组通信方式,提出了一种同步性较好、可自由拼接和拆分的多通道组合显示方法.实际应用结果表明,该系统具有较好的实时性和更好的显示效果.In order to solve the real-time visualization problem of a complex environment in acoustic confrontation,a new visual simulation system using graphic processing unit（GPU） programming was developed based on the OpenSceneGraph（OSG） framework.Its architecture and scene organization were designed,and the ocean surface grid was generated using a projection-grid model and fast Fourier transformation method based on GPU;its reflection and refraction effect was realized using texture mapping.The underwater beam,floating particles,and blur effect also were realized,and the high dynamic range（HDR） effect was simulated using post-rendering technology.Three OSG sound nodes were developed based on the FMOD audio engine,which could play 3D object sound,event sound,and background sound.Lastly,a better multi-channel display method was proposed based on multicast communication in which sub-channels not only can be put together to display the composite scene,but also can display different angles of the scene independently.The application results show that this system had better performance in real-time and a better 3D display effect.

关 键 词：水声对抗 视景仿真 海洋场景 多通道 图形处理器 高动态范围 音频引擎 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]