结合即时控制系统架构之3D介面VR重点课件.ppt

1、2005/06/22NCSLAB1結合即時控制系統架構之3D介面VR 系統分析電機系控制組電機系控制組(D92921003)黃雋博(R93921067)彭詩淵2005/06/22NCSLAB2OutlinenMotivationnSystem ArchitecturenTechniquenA case studynConclusion2005/06/22NCSLAB3MotivationnInteractive game environmentqreal time system which can interact with other peopleqKeeping the system ma

2、intain the display rate with 25 frame/s.2005/06/22NCSLAB4MotivationnIn the on-line interactive VR system,the delay will be occurred in many situations.Therefore,the timing analysis is very important for this project.The delay events that we predict are as bellow:nDelay in server:Calculation time(mod

3、eling,numerical)Separate the whole environment for individual clientnDelay in network time:Conjunction delayCompensationnDelay in client:Rendering(According to hardware of the system)Request for the server2005/06/22NCSLAB5System Architecture2005/06/22NCSLAB6TechniquenControl over Quality,Bandwidth,B

4、uffer Condition,CPUs Utilization,Scheduling Techniques for real time-systems(Ex:Rate Monotonic(RM)Algorithm,Earliest Deadline First(EDF)Algorithm)2005/06/22NCSLAB7Technique2005/06/22NCSLAB8TechniquenTeb:encoder buffering delaynTdb:decoder buffering delaynTe:encoding delaynTd:decoding delaynTc:networ

5、king delaynTnm:numerical method computing delaynTpm:physics model computing delaynTv:visualization delaynTeb+Tdb+Te+Td+Tc+Tnm+Tpm+Tv40ms2005/06/22NCSLAB9TechniquenLevel of Details(LOD)techniqueqLOD rendering techniques reduce the geometric complexity of 3D models,sacrificing visual rendering quality

6、 in order to increase frame rendering rates.nPhysically-based:based on physical lows or mathematical functions-complex but realnNon physically-based game:do not use physical laws-simple but unrealqTry the best to reduce the high computational costs and maintain the reality2005/06/22NCSLAB10A case st

7、udynNon-physical based methodqGrassland by Neyret 1998ntranslates texture to show the animated grassland Neyret 1998 qLOD example-Grassland Perbet and Cani 2001highest LOD modelmiddle LOD modellowest LOD model2005/06/22NCSLAB11qPhysical based models Cantilever beam model Anjyo et al.1992 Mass spring

8、 damped model Miller 1988 Mass-spring modelProvot 1995 321()()62PPLy xxxEI()iiiiiildlFk LlDdtlA case study111iiiiiimmmmimmXXFk LXXXX 2005/06/22NCSLAB12A case studyqSet spring between the massesnAssume the position of i th particle is nthe position of i+1 th particle is nthe internal force between th

9、e i th spring is nEach particle has gravity,wind resistance and buoyancy(,)iiiimmmmXxyz+1+1+1+1(,)iiiimmmmXxyz+1+1+1iiiiiimmmmimmXXFk LXXXX Use Eulers method()Fattm()()*()v ttv tta tt()()*()p ttp ttv tt2005/06/22NCSLAB13A case studyDifferent types of grassNumber of masses of each grassComputational

10、timefor one frameLine visualization2069812.5 msLine visualization417.2 msSpline visualization481.3 ms1000-Grass ENVIRONMENTBird viewLOW LODHIGH LOD2005/06/22NCSLAB14 For(i=0;iGrassNum;i+)IF(ViewerPOS.Z-grassi.Z)HighLODRange)grassiHighLODVisualization Else grassiLowLODVisualization XYZ2005/06/22NCSLA

11、B15A Case StudyDifferent Types of grasslandsSimulation time for one frameLine visualization(Low LOD)15 msSpline visualization(High LOD)80 mscombine Line and Spline visualization50.0 msFor Client/Sever1000-Grass ENVIRONMENT733MHZ Pentium III CPUDifferent Types of grasslandsSimulation time for one fra

12、meLine visualization(Low LOD)17.2 msSpline visualization(High LOD)81.3 ms combine Line and Spline visualization 34.4 ms3GHZ Pentium IV CPU2005/06/22NCSLAB16Scheduling2005/06/22NCSLAB17SchedulingnEDF2005/06/22NCSLAB18ExperimentnParticle:6240 vs.14040 M ean624014040Com put i ng Net worki ng Drawi ng C

13、om put i ng Net worki ng Drawi ngLow Qual i t y9.630412.48235.005149.827817.4196.2772Opt i m al9.691112.341814.307649.705117.284815.0646Hi gh Qual i t y9.744327.379726.277248.973420.172228.7709St d624014040Com put i ng Net worki ng Drawi ng Com put i ng Net worki ng Drawi ngLow Qual i t y7.61066.713

14、57.30166.34386.1417.6786Opt i m al7.59736.65384.38266.1126.06463.2316Hi gh Qual i t y7.597813.68647.32175.351647.804748.81732005/06/22NCSLAB19Experiment-Low Quality050100150050100150200computing 6240050100150050100150200Networking 6240050100150050100150200Drawing 6240050100150050100150200computing 1

15、4040050100150050100150200Networking 14040050100150050100150200Drawing 140402005/06/22NCSLAB20Experiment-High Quality050100150050100150200computing 6240050100150050100150200Networking 6240050100150050100150200Drawing 6240050100150050100150200computing 14040050100150050100150200Networking 140400501001

16、50050100150200Drawing 140402005/06/22NCSLAB21Experiment-Optimal050100150050100150200computing 6240050100150050100150200Networking 6240050100150050100150200Drawing 6240050100150050100150200computing 14040050100150050100150200Networking 14040050100150050100150200Drawing 140402005/06/22NCSLAB22Conclusi

17、onnThe concept of LOD is used to maintain the human impression under real-time requirement.nThe timing analysis for real-time has been established.nthe rate control for large number of the VR scenario have been discussed.2005/06/22NCSLAB23ReferencesnNeyret 1998 F.Neyret,“Modeling,animating,and rende

18、ring complex scenes using volumetric textures,”IEEE Transactions on Visualization and Computer Graphics,Vol.4,No.1,pp.5570,January 1998.nPerbet and Cani 2001 Frank Perbet and Marie-Paule Cani,“Animating praires in real-time,”Proceedings of the 2001 symposium on Interactive 3D graphics,March,2001nAnj

19、yo et al.1992 K.Anjyo,Y.Usami,and T.Kurihara,“A Simple Method for Extracting the Natural Beauty of Hair,”Proceedings of the 19th annual conference on Computer graphics and interactive techniques,Vol.26,No.2,pp.111-120,Chicago,IL,USA,July 1992.nMiller 1988 G.S.P.Miller,“The motion dynamics of snakes

20、and worms,”Proceedings of the 15th annual conference on Computer graphics and interactive techniques,Vol.22,No.4,pp.169-178,Atlanta,GA,USA,August 1988nProvot 1995 X.Provot,“Deformation constraints in a mass-spring model to describe rigid cloth behavior,”Proceedings of the Graphics Interface,pp.147-1

21、54,Qubec,QC,USA,May 1995.nJ.Vieron and C.Guillemot 2004 J.Vieron and C.Guillemot,“Real-time constrained TCP-compatible rate control for video over the Internet,”IEEE Transactions on Multimedia,Vol.6,Issue 4,pp.634-646,2004.nJ.Bai et al.2002 J.Bai,Q.Liao,X.Lin,and X.Zhuang,“Rate-distortion model base

22、d rate control for real-time VBR video coding and low-delay communications,”Signal Processing:Image Communication.Vol.17,No.2,pp.187-199,2002.nD.Wu et al.2000 D.Wu,Y.T.Hou,and Y.Q.Zhang,“Transporting Real-Time Video over the Internet:Challenges and Approaches,”Proceedings of the IEEE,Vol.88,Issue 12

23、,pp.1855-1877,2000.nB.Li and K.Nahrstedt 1999 B.Li and K.Nahrstedt,“A control-based middleware framework for quality-of-service adaptations,”IEEE Journal on Selected Areas in Communications,Vol.17,Issue 9,pp.1632-1650,1999.nC.Lu et al.2002 C.Lu,J.A.Stankovic,G.Tao,and S.H.Son,“Feedback Control Real-

24、Time Scheduling:Framework,Modeling,and Algorithms,”Real-Time Systems Journal,Special Issue on Control-theoretical Approaches to Real-Time Computing,23(1/2):85-126,July/September 2002.nBaldi and Ofek 2000 M.Baldi and Y.Ofek,“End-to-end delay analysis of videoconferencing over packet-switched networks,”IEEE/ACM Transactions on Networking,Vol.8,Issue 4,pp.479-492,Aug.2000