分布式虚拟驾驶模拟器中对象建模及场景优化技术的研究
|
摘要
虚拟现实技术是二十世纪末兴起的一门综合性信息及计算机应用技术,它融合了计算机图形学、三维实时动画技术、计算机接口技术、传感技术,人工智能等多个信息及计算机应用技术的分支,是人们通过计算机对复杂数据进行可视化操作与交互的一种全新方式。将虚拟现实技术应用于汽车驾驶模拟训练系统,通过计算机生成汽车行驶过程中的虚拟场景、音响效果和运动仿真,可以使操作者沉浸在虚拟驾驶环境中,根据虚拟驾驶环境提供的视觉、听觉、触觉感受,执行相应驾驶动作,体验、认识和学习现实世界中的汽车驾驶,具有节能、安全、经济等优点,而且培训不受时间、气候、场地的限制,训练效率高,培训周期短。
论文结合汽车驾驶模拟器视景系统的开发,对虚拟场景中的几何建模技术、声音建模技术进行了研究。系统采用MultiGen Creator和3DS Max作为建模工具,采用VC++和OSG构建了仿真开发平台。论文对图形绘制流水线的三个阶段进行了研究;分析各种流水线的瓶颈问题,总结查找瓶颈的典型方法,对常用的图形绘制加速原理和方法进行介绍;讨论了基于GPU的图形绘制流水线和它的一些优化。针对实时视景仿真的特点,提出了场景模型建立和组织过程中应该遵循的原则。对场景建立过程进行了描述,指出了其中应该注意的问题。在场景建模过程中,运用了一系列先进的建模技术,如LOD技术、纹理映射技术、实例技术以及外部引用技术等,在实现逼真视觉效果的同时,减少系统资源耗费,提高系统的运行性能。本文重点研究了三维声效建模中的关键技术问题,包括声音的空间定位原理,声音的距离消减模型,声音的多普勒效应等;对虚拟空间声效模拟系统开发工具,如课题中使用到的OpenAL开发库,进行了必要的介绍,并深入探讨了汽车驾驶模拟器中的声效模块设计及其实现方式。
在所开发的驾驶模拟器视景仿真模块中,文中讨论的建模技术和方法已经获得了成功的应用,兼顾了实时仿真中的沉浸性、实时性和交互性的要求,取得了较好的仿真效果;同时,三维空间声音加入到实时场景中,增加了驾驶者的听觉体验,加强了系统沉浸感。
Virtual Reality is a comprehensive technology that rised at the end of the 20~(th) century. It fused many branches of information and computer application technology, such as computer graphics, 3D real-time animation, computer interface, sensor, artificial inteligence and so on. It is a new mode for people to handle and interact with complex data in a visulized way through computer. When Virtual Reality technology is applied in vehicle simulation system to create virtual scene, 3D sound and movement, operators will immerse in the virtual driving environment, as is driving a real car. According to the virtual vision, hearing and touch feelings obtained from the driving environment, operators can conceive driving reactions, experience and study how to drive in the real world. Obviously, it has a lot of advantages to adopt Virtual Reality technology in vehicle simulation system, such as low-costs, safety, no limitations for time, climate and field, and he training period is short and will be very effective too.
The thesis studies the VR modeling techniques of virtual scene and 3D sound according to the development of Visual Vehicle Simulation System. Firstly, the modeling tools is determined by system requirement, and VisualC++ and OpenSceneGraph(OSG) are adopted to construct simulation platform, and then the system development blue print is put forward on the basis of them. The thesis studies the three phases of graphics rendering pipeline; analyses bottlenecks of the pipeline, and give some methods to find which bottleneck it is in a given hardware architecture. The principles of some graphic rendering accelerating methods is introduced. The pipeline based on GPU and it's optimization methods are discussed too. Some important scene modeling and organising rules are brought forward according to real-time simulation system requirement. Scene modeling process, and issues about scene modeling that should be awared are discussed. In the modeling process, a lot of advanced modeling techniques, such as LOD, texture mapping, instance, external references are used to lessen the cost of system resources, improve system performance, and get perfect vision results. The thesis introduces 3D sound modeling thchnology, includes locatin principle of 3D sound, model of sound rolloff model according to distance, Doppler effect. The audio API used in the vehicle simulation system is discusses too. The design and implementation of audio module of the system are discussed in detail in the thesis.
The VR modeling techniques discussed had been used successfully in system development, which fulfilled the "immersion", "real time" and "interactivity" requirement of system, and got effective simulation results. At the same time, 3D sound added into the graphic scene increase the immersion feel as it supplies the driver hearing sense.
论文结合汽车驾驶模拟器视景系统的开发,对虚拟场景中的几何建模技术、声音建模技术进行了研究。系统采用MultiGen Creator和3DS Max作为建模工具,采用VC++和OSG构建了仿真开发平台。论文对图形绘制流水线的三个阶段进行了研究;分析各种流水线的瓶颈问题,总结查找瓶颈的典型方法,对常用的图形绘制加速原理和方法进行介绍;讨论了基于GPU的图形绘制流水线和它的一些优化。针对实时视景仿真的特点,提出了场景模型建立和组织过程中应该遵循的原则。对场景建立过程进行了描述,指出了其中应该注意的问题。在场景建模过程中,运用了一系列先进的建模技术,如LOD技术、纹理映射技术、实例技术以及外部引用技术等,在实现逼真视觉效果的同时,减少系统资源耗费,提高系统的运行性能。本文重点研究了三维声效建模中的关键技术问题,包括声音的空间定位原理,声音的距离消减模型,声音的多普勒效应等;对虚拟空间声效模拟系统开发工具,如课题中使用到的OpenAL开发库,进行了必要的介绍,并深入探讨了汽车驾驶模拟器中的声效模块设计及其实现方式。
在所开发的驾驶模拟器视景仿真模块中,文中讨论的建模技术和方法已经获得了成功的应用,兼顾了实时仿真中的沉浸性、实时性和交互性的要求,取得了较好的仿真效果;同时,三维空间声音加入到实时场景中,增加了驾驶者的听觉体验,加强了系统沉浸感。
Virtual Reality is a comprehensive technology that rised at the end of the 20~(th) century. It fused many branches of information and computer application technology, such as computer graphics, 3D real-time animation, computer interface, sensor, artificial inteligence and so on. It is a new mode for people to handle and interact with complex data in a visulized way through computer. When Virtual Reality technology is applied in vehicle simulation system to create virtual scene, 3D sound and movement, operators will immerse in the virtual driving environment, as is driving a real car. According to the virtual vision, hearing and touch feelings obtained from the driving environment, operators can conceive driving reactions, experience and study how to drive in the real world. Obviously, it has a lot of advantages to adopt Virtual Reality technology in vehicle simulation system, such as low-costs, safety, no limitations for time, climate and field, and he training period is short and will be very effective too.
The thesis studies the VR modeling techniques of virtual scene and 3D sound according to the development of Visual Vehicle Simulation System. Firstly, the modeling tools is determined by system requirement, and VisualC++ and OpenSceneGraph(OSG) are adopted to construct simulation platform, and then the system development blue print is put forward on the basis of them. The thesis studies the three phases of graphics rendering pipeline; analyses bottlenecks of the pipeline, and give some methods to find which bottleneck it is in a given hardware architecture. The principles of some graphic rendering accelerating methods is introduced. The pipeline based on GPU and it's optimization methods are discussed too. Some important scene modeling and organising rules are brought forward according to real-time simulation system requirement. Scene modeling process, and issues about scene modeling that should be awared are discussed. In the modeling process, a lot of advanced modeling techniques, such as LOD, texture mapping, instance, external references are used to lessen the cost of system resources, improve system performance, and get perfect vision results. The thesis introduces 3D sound modeling thchnology, includes locatin principle of 3D sound, model of sound rolloff model according to distance, Doppler effect. The audio API used in the vehicle simulation system is discusses too. The design and implementation of audio module of the system are discussed in detail in the thesis.
The VR modeling techniques discussed had been used successfully in system development, which fulfilled the "immersion", "real time" and "interactivity" requirement of system, and got effective simulation results. At the same time, 3D sound added into the graphic scene increase the immersion feel as it supplies the driver hearing sense.
引文
[1]Sutherland I.The ultimate display.Proc IFIP Congress,1962:506-508
[2]石教英.虚拟现实基础及实用算法.北京:科学出版社,2002
[3]范正伟,贺秀良,姜丁等.汽车驾驶模拟器研究现状与未来展望.汽车运用,2004,10:31-32.
[4]陈定方,李勋祥,李文锋等.基于分布式虚拟现实技术的汽车驾驶模拟器的研究.系统仿真学报,2005,17(2):347-350.
[5]尹念东,陈定方,李安定.基于OpenGVS的分布式虚拟汽车驾驶视景系统设计与实现.武汉理工大学学报(交通科学与工程版),2006,30(6):984-987.
[6]Grigore C.Burdea,Philippe Coiffet.虚拟现实技术(第二版),魏迎梅,栾悉道等译.北京:电子工业出版社,2005.
[7]戴唐云.基于3DSMAX_OGRE的火箭视景仿真系统的研究与实现:[硕士学位论文].四川:电子科技大学机械电子工程系,2007
[8]航线多媒体编著.3DS MAX三维建模专项训练.北京:中国铁道出版社,2002
[9]MultiGen-Paradigm Inc..MultiGen Creator User's Guide(Version3.0).U.S.A.:MultiGenParadigm Inc.,2004
[10]吴家铸,党岗,刘华峰等.视景仿真技术及应用.西安:西安电子科技大学出版社,2001
[11]王焕钢,尹岩青等.3DS max宝典.北京:电子工业出版社,2004
[12]http://www.opengl.org/
[13]http://www.openscenegraph.com
[14]Akeley,Kurt.The Silicon Graphics 4D/240GTX Super workstation.IEEE Computer Graphics and Applications,1989,9(4):71-83
[15]H.Gouraud.Computer Display of Curved Surfaces.IEEE Trans.Computers,1971,20(6):623-628
[16]B.T.Phong.Illumination for Computer Generated Images.Ph.D Dissertation,University of Utah,1973
[17]Tomas Akenine-Moller,Eric Haines.Real Time Rendering.Massachusetts:A.K.Peters,2002
[18]刘姚新.基于GPU的实时绘制算法研究:[硕士学位论文].重庆:重庆大学计算机软件与理论系,2007
[19]Subodh Kumar,Dinesh Manocha,William Grrett.Hierarchical back-face computation.Computers and Graphics,1999,vol.23(5):681-692
[20]Assarsson,Tomas Moller.Optimized view frustum culling algorithms for bounding boxes.Journal of Graphics Tools,2000,5(1):9-22
[21]P.Lindstrom,V.Pascucci.Visualization of Larget Terrain Made Easy.Proc.IEEE Visualization,2001:363-370
[22]Paul S.Heckbert,Michael Garland.Survey of Polygonal Surface Simplification Algorithms.School of Computer Science Carnegie Mellon University,1997
[23]Cignoni P.,et al.BDAM - Batched dynamic adaptive meshes for high performance terrain visualization.Computer Graphics Forum 22,2003:505-514
[24]Wagner,D.Terrain geomorphing in the vertex shader.In ShaderX2:Shader Programming Tips and Tricks with DirectX 9.Wordware Publishing.2004
[25]Davis.ROAM Using Triangle Clusters(RUSTIC).Master' s thesis,U.C CS Dept,2000
[26]Levenberg,J.Fast view-dependent level-of-detail rendering using cached geometry.IEEE Visualization 2002:259-266
[27]李胜.大规模室外地形场景加速绘制技术研究:[博士学位论文].北京:中国科学院软件研究所,2004
[28]胡小强.虚拟现实技术.北京:北京邮电大学出版社,2005
[29]杨宝民,朱一宁.分布式虚拟现实技术及其应用.北京:科学出版社,2000
[30]孙家广等.计算机图形学(第三版).北京:清华大学出版社,1998
[31]蔡士杰,吴春榕,孙正兴等译.计算机图形学.北京:电子工业版社,1998
[32]朱忠祥,毛恩荣.汽车驾驶训练模拟器中的三维建模技术.农业机械学报,2004,35(1):22-25
[33]淮永建,郝重阳等.LOD和多边型简化.系统仿真学报,2001,13(增刊):26-30
[34]Eck M,DeRose T,Duchamp T et al.Multiresolution analysis of arbitrary meshes.In.Proceedings of SIGGRAPH'95[C].Held in LosAngles,California,1995:173-150
[35]周昆,潘志庚,石教英.多细节层次模型间的平滑过渡.计算机辅助设计与图形学学报,2000,12(6):463-466
[36]Sharon Rose Clay.Optimization for Real-Time Graphics Applications.Silicon Graphics Computer System,1996
[37]袁慧,曾黄麟.有真实感图形的纹理映射算法.计算机应用,1998,18(10):29-31
[38]Mourant R R,Refsland D.Developing a 3D Sound Environment For a Driving Simulator.Proceedings of the Ninth International Conference on Virtual System and Multimedia [C].Montreal,Canada,2003:711-719
[39]罗福元,王行仁,彭晓源.声音渲染技术及其在虚拟环境中的应用.系统仿真学报,1999,11(5):364-367
[40]何正伟等.实时三维声音仿真系统的设计与实现.计算机应用研究.2002,1:23-25
[41]McReynolds,Tom,David Blythe,et al.Advanced Graphics Programming Techniques Using OpenGL Course Notes.SIGGRAPH99 Course Notes,1999
[42]尹念东.汽车驾驶模拟器研究现状与技术关键.湖北汽车工业学院学报,2002,16(4):7-10.
[43]李安定.虚拟现实建模技术研究及其在汽车驾驶模拟器中的应用:[硕士学位论文].武汉:武汉理工大学机械电子工程系,2006
[44]陈定方,罗亚波.虚拟设计.北京:机械工业出版社,2002
[45]Durand.R.Begault.3-D sound for Virtual Reality and Multimedia.United Kingdom Edition.Academic Press Inc,1994
[46]杨小冬.三维运动声音在机载告警系统中的应用:[硕士学位论文].成都:电子科技大学通讯与信息技术,2006
[47]魏国柱,李林,杜安.普通情况下的多普勒效应表达式.东北大学学报(自然科学版).2004,25(6):602-605
[48]顾钦.基于纹理映射的快速渲染技术研究:[硕士学位论文].天津:天津大学电子信息工程学院,2004
[49]官巍,蔡晓琳,陈海.细节层次技术在场景建模中的应用.系统仿真学报.2006,18(2):427-429
[50]刘浩.3D图形渲染引擎的设计与实现:[硕士学位论文].成都:电子科技大学计算机应用技术,2007
[51]Arie Kaufman et al.Real-Time Accelerators for Volume Rendering.Proceedings of the 7~(th)IEEE Visualization Conference(VIS'96)
[52]Gabor Szijarto,Jozsef Koloszar.Hardware Accelerated Rendering of Foliage for Real-Time Applications.Budapest University of Technology and Economics,ACM Inc.:141-148
[53]Dave Salvator.3D Pipeline Turorial.ExtremeTech,2001
[54]周洪龑,大规模场景数据的动态调度和布局优化:[硕士学位论文].杭州:杭州大学计算机应用系,2006
[2]石教英.虚拟现实基础及实用算法.北京:科学出版社,2002
[3]范正伟,贺秀良,姜丁等.汽车驾驶模拟器研究现状与未来展望.汽车运用,2004,10:31-32.
[4]陈定方,李勋祥,李文锋等.基于分布式虚拟现实技术的汽车驾驶模拟器的研究.系统仿真学报,2005,17(2):347-350.
[5]尹念东,陈定方,李安定.基于OpenGVS的分布式虚拟汽车驾驶视景系统设计与实现.武汉理工大学学报(交通科学与工程版),2006,30(6):984-987.
[6]Grigore C.Burdea,Philippe Coiffet.虚拟现实技术(第二版),魏迎梅,栾悉道等译.北京:电子工业出版社,2005.
[7]戴唐云.基于3DSMAX_OGRE的火箭视景仿真系统的研究与实现:[硕士学位论文].四川:电子科技大学机械电子工程系,2007
[8]航线多媒体编著.3DS MAX三维建模专项训练.北京:中国铁道出版社,2002
[9]MultiGen-Paradigm Inc..MultiGen Creator User's Guide(Version3.0).U.S.A.:MultiGenParadigm Inc.,2004
[10]吴家铸,党岗,刘华峰等.视景仿真技术及应用.西安:西安电子科技大学出版社,2001
[11]王焕钢,尹岩青等.3DS max宝典.北京:电子工业出版社,2004
[12]http://www.opengl.org/
[13]http://www.openscenegraph.com
[14]Akeley,Kurt.The Silicon Graphics 4D/240GTX Super workstation.IEEE Computer Graphics and Applications,1989,9(4):71-83
[15]H.Gouraud.Computer Display of Curved Surfaces.IEEE Trans.Computers,1971,20(6):623-628
[16]B.T.Phong.Illumination for Computer Generated Images.Ph.D Dissertation,University of Utah,1973
[17]Tomas Akenine-Moller,Eric Haines.Real Time Rendering.Massachusetts:A.K.Peters,2002
[18]刘姚新.基于GPU的实时绘制算法研究:[硕士学位论文].重庆:重庆大学计算机软件与理论系,2007
[19]Subodh Kumar,Dinesh Manocha,William Grrett.Hierarchical back-face computation.Computers and Graphics,1999,vol.23(5):681-692
[20]Assarsson,Tomas Moller.Optimized view frustum culling algorithms for bounding boxes.Journal of Graphics Tools,2000,5(1):9-22
[21]P.Lindstrom,V.Pascucci.Visualization of Larget Terrain Made Easy.Proc.IEEE Visualization,2001:363-370
[22]Paul S.Heckbert,Michael Garland.Survey of Polygonal Surface Simplification Algorithms.School of Computer Science Carnegie Mellon University,1997
[23]Cignoni P.,et al.BDAM - Batched dynamic adaptive meshes for high performance terrain visualization.Computer Graphics Forum 22,2003:505-514
[24]Wagner,D.Terrain geomorphing in the vertex shader.In ShaderX2:Shader Programming Tips and Tricks with DirectX 9.Wordware Publishing.2004
[25]Davis.ROAM Using Triangle Clusters(RUSTIC).Master' s thesis,U.C CS Dept,2000
[26]Levenberg,J.Fast view-dependent level-of-detail rendering using cached geometry.IEEE Visualization 2002:259-266
[27]李胜.大规模室外地形场景加速绘制技术研究:[博士学位论文].北京:中国科学院软件研究所,2004
[28]胡小强.虚拟现实技术.北京:北京邮电大学出版社,2005
[29]杨宝民,朱一宁.分布式虚拟现实技术及其应用.北京:科学出版社,2000
[30]孙家广等.计算机图形学(第三版).北京:清华大学出版社,1998
[31]蔡士杰,吴春榕,孙正兴等译.计算机图形学.北京:电子工业版社,1998
[32]朱忠祥,毛恩荣.汽车驾驶训练模拟器中的三维建模技术.农业机械学报,2004,35(1):22-25
[33]淮永建,郝重阳等.LOD和多边型简化.系统仿真学报,2001,13(增刊):26-30
[34]Eck M,DeRose T,Duchamp T et al.Multiresolution analysis of arbitrary meshes.In.Proceedings of SIGGRAPH'95[C].Held in LosAngles,California,1995:173-150
[35]周昆,潘志庚,石教英.多细节层次模型间的平滑过渡.计算机辅助设计与图形学学报,2000,12(6):463-466
[36]Sharon Rose Clay.Optimization for Real-Time Graphics Applications.Silicon Graphics Computer System,1996
[37]袁慧,曾黄麟.有真实感图形的纹理映射算法.计算机应用,1998,18(10):29-31
[38]Mourant R R,Refsland D.Developing a 3D Sound Environment For a Driving Simulator.Proceedings of the Ninth International Conference on Virtual System and Multimedia [C].Montreal,Canada,2003:711-719
[39]罗福元,王行仁,彭晓源.声音渲染技术及其在虚拟环境中的应用.系统仿真学报,1999,11(5):364-367
[40]何正伟等.实时三维声音仿真系统的设计与实现.计算机应用研究.2002,1:23-25
[41]McReynolds,Tom,David Blythe,et al.Advanced Graphics Programming Techniques Using OpenGL Course Notes.SIGGRAPH99 Course Notes,1999
[42]尹念东.汽车驾驶模拟器研究现状与技术关键.湖北汽车工业学院学报,2002,16(4):7-10.
[43]李安定.虚拟现实建模技术研究及其在汽车驾驶模拟器中的应用:[硕士学位论文].武汉:武汉理工大学机械电子工程系,2006
[44]陈定方,罗亚波.虚拟设计.北京:机械工业出版社,2002
[45]Durand.R.Begault.3-D sound for Virtual Reality and Multimedia.United Kingdom Edition.Academic Press Inc,1994
[46]杨小冬.三维运动声音在机载告警系统中的应用:[硕士学位论文].成都:电子科技大学通讯与信息技术,2006
[47]魏国柱,李林,杜安.普通情况下的多普勒效应表达式.东北大学学报(自然科学版).2004,25(6):602-605
[48]顾钦.基于纹理映射的快速渲染技术研究:[硕士学位论文].天津:天津大学电子信息工程学院,2004
[49]官巍,蔡晓琳,陈海.细节层次技术在场景建模中的应用.系统仿真学报.2006,18(2):427-429
[50]刘浩.3D图形渲染引擎的设计与实现:[硕士学位论文].成都:电子科技大学计算机应用技术,2007
[51]Arie Kaufman et al.Real-Time Accelerators for Volume Rendering.Proceedings of the 7~(th)IEEE Visualization Conference(VIS'96)
[52]Gabor Szijarto,Jozsef Koloszar.Hardware Accelerated Rendering of Foliage for Real-Time Applications.Budapest University of Technology and Economics,ACM Inc.:141-148
[53]Dave Salvator.3D Pipeline Turorial.ExtremeTech,2001
[54]周洪龑,大规模场景数据的动态调度和布局优化:[硕士学位论文].杭州:杭州大学计算机应用系,2006