委托机构广州市海珠区城市管理局51

1、Efficient Image-Based Methods for Rendering Soft Shadows alan.heirich,laurent.mollcompaq Maneesh AgrawalaRavi RamamoorthiAlan Heirich Laurent MollPixar Animation StudiosStanford UniversityCompaq Computer CorporationCompaq Computer CorporationHard vs. Soft ShadowsHard ShadowsSoft ShadowsShadow mapsIm

2、age-based hard shadows Williams 78Time, memory depend on image size, not geometric scene complexityDisadvantage: bias and aliasing artifactsSoft shadows Chen and Williams 93View interpolate multiple shadow mapsIBR good for soft shadowsIBR good for secondary effectsArtifacts less perceptibleIBR works

3、 well for nearby viewpointsShadow maps from light source Light source localized areaPoorly sampled regions are also dimly litIBR good for soft shadowsPoorly sampled regions are also dimly litAttenuation onlyWith lightingLightShadow mapContributionsExtend shadow maps to soft shadowsImage-based render

4、ing especially suitableTwo novel image-based algorithms:Layered attenuation maps (LAM) Coherence-based raytracing (CBRT) LAMDisplay: 5-10 fpsSome aliasing artifactsInteractive applicationsGamesPreviewing CBRTRender: 19.83 minSpeedup: 12.96xProduction quality imagesRefresher: LDIsLayered depth images

5、 Shade et al. 98GeometryCameraRefresher: LDIsLayered depth images Shade et al. 98LDIRefresher: LDIsLayered depth images Shade et al. 98LDI(Depth, Color)PrecomputationRender views from points on light (hardware)Create layered attenuation map (software)Warp views into LDI Store (depth, attenuation)Obj

6、ects in LAM visible in at least 1 viewPrecomputation1st viewpointPrecomputation2nd viewpointAttenuation = 1/2Attenuation = 2/2PrecomputationWarped 2nd viewpoint Attenuation = 1/2Attenuation = 2/2Not present DisplayRender scene without shadows (hardware)Project into LAM (software)Read off attenuation

7、 Attenuation modulates shadowless renderingDisplay LAM (center of light)EyeDisplay LAM (center of light)EyeAttenuation = 2/2Color = Color * 2/2Display LAM (center of light)EyeDisplay LAM (center of light)EyeNot in LAMAttenuation = 0Color = Color * 0Previous Interactive MethodsHW per-object textures

8、Herf and Heckbert 97Convolution Soler and Sillion 98Texture intensiveLAM size: 512 x 512Avg num depth layers: 1.5Precomp: 7.7 sec (64 views) 29.4 sec (256 views)Display: 5-10 fpsLAM size: 512 x 512Avg num depth layers: 2Precomp: 6.0 sec (64 views) 22.4 sec (256 views)Display: 5-10 fpsLAM VideoLayere

9、d attenuation maps fast, aliasesCoherence-based raytracing slow, noiseLAMCBRTCoherence-based raytracingHierarchical raytracing through depth imagesTime, memory decoupled from geometric scene complexityCoherence-based samplingLight source visibility changes slowlyReduce number shadow rays tracedAlso

10、usable with geometric raytracerRepresent scene with multiple shadow mapsLightImage-based raytracing1st shadow mapRepresent scene with multiple shadow mapsLightImage-based raytracing2nd shadow map1st shadow mapTrace shadow ray through shadow mapsLightImage-based raytracing2nd shadow map1st shadow map

11、Hierarchical img based raytracingPreviousHeight fields:Musgrave et al. 89New views:Marcato 98 Chang 98Lischinski and Rappoport 98 Shadows:Keating and Max 99Our contributionsAccelerations shadow ray traversalFast methods handling multiple depth imagesSpeedup: 2.20 xLight source visibility imageLightV

12、isibility images1Light source visibility images1s2Vis image for s1LightVisibility imageCoherence-based samplingCompute visibility image at first point s1Loop over following surface points siPredict visibility image at si from si-1Trace rays where prediction confidence lowPredicting visibilityBlocker

13、 ptss1s1s2PredictionPredicting visibilityBlocker ptss1s1s2Prediction Low confidence Light source edges Blocked/unblocked edgesPrediction confidencePredicted visibility Trace rays in all Xed cells High confidence:5 Low confidence:31 Total cells:36 Ratio:5/36 = 0.14 Low confidence Light source edges B

14、locked/unblocked edgesPrediction confidencePredicted visibility Trace rays in all Xed cells High confidence:56 Low confidence:88 Total cells:144 Ratio:56/144 = 0.40Propagating low confidence If traced ray = prediction trace neighbor cells Similar to Hart et al. 99 Prediction correctPropagating low c

15、onfidence If traced ray = prediction trace neighbor cellsPrediction incorrect Similar to Hart et al. 99 Light cells: 16 x 16 (256) Four 1024 x 1024 maps Precomp: 2.33 min Render:19.83 min Rays:79.86 Speedup:12.96x 2.27x due to image-based raytracing accelerations 5.71x due to coherence-based samplin

16、g Light cells: 16 x 16 (256) Four 1024 x 1024 maps Precomp: 3.93 min Render:65.13 min Rays:88.74 Speedup:8.52x 2.16x due to image-based raytracing accelerations 3.94x due to coherence-based samplingLAMCBRTConclusionsTwo efficient image-based methodsLayered attenuation maps Interactive applications C

17、oherence-based raytracingProduction quality imagesIBR ideal for soft shadows secondary effectsFuture workDynamic scenesAntialiasing with deep shadow mapsHardware implementationAcknowledgementsTom LokovicReid Gershbein, Tony Apodaca, Mark VandeWettering, Craig KolbStanford graphics groupPrediction er

18、rorsMissed blockersDependent on surface sampling densityMissed holesDependent on light source sampling densitys1s2missedblockereMbJ8G4D1z-w*t!qYnVjSgOdLaI6F3C0y)v%s#pXlUiRfNcK8H5E2A+x(u$rZnWkThPeMbJ7G4D1z-w&t!qYmVjSgOdL9I6F3B0y)v%s#oXlUiQfNcK8H5D2A+x*u$rZnWkShPeMaJ7G4C1z)w&t!pYmVjRgOcL9I6E3B0y(v%r#o

19、XlTiQfNbK8H5D2A-x*u$qZnWkShPdMaJ7F4C1z)w&s!pYmUjRgOcL9H6E3B+y(v%r#oWlTiQeNbK8G5D1A-x*t$qZnVkSgPdMaI7F4C0z)w&s!pXmUjRfOcL9H6E2B+y(u%r#oWlThQeNbJ8G5D1A-w*t$qYnVkSgPdLaI7F3C0z)v&s#pXmUiRfOcK9H5E2B+x(u%rZoWlThQeMbJ8G4D1A-w*t!qYnVjSgPdLaI6F3C0y)v&s#pXlUiRfNcK9H5E2A+x(u$rZoWkThPaI7F3C0z)v&s#pXmUiRfOcK9H6E

20、2B+x(u%rZoWlThQeMbJ8G4D1A-w*t!qYnVjSgPdLaI6F3C0y)v&s#pXlUiRfNcK9H5E2A+x(u$rZoWkThPeMbJ7G4D1z-w*t!qYmVjSgOdLaI6F3B0y)v%s#pXlUiQfNcK8H5E2A+x*u$rZnWkThPeMaJ7G4C1z-w&t!pYmVjRgOdL9I6E3B0y(v%s#oXlTiQfNbK8H5D2A+x*u$qZnWkShPeMaJ7F4C1z)w&t!pYmUjRgOcL9I6E3B+y(v%r#oXlTiQeNbK8G5D2A-x*t$qZnVkShPdMaI7F4C0z)w&s!pY

21、mUjRfOcL9H6E3B+y(u%r#oWlTiQeNbJ8G5D1A-x*t$qYnVkSgPdMaI7F3C0z)v&s!pXmUiRfOcK9H6E2B+x(u%rZoWlThQeNbJ8G4D1A-w*t$qYnVjSgPdLaI7F3C0y)v&s#pXmUiRfNcK9H5E2B+x(u$rZoWkThQeMbJ7G4D1z-w*t!qYmVjSgOdLaI6F3B0y)v%s#pXlUiRfNcK8H5E2A+x(u$rZnWkThPeMbJ7G4C1z-w&t!qYmVjRgOdL9I6F3B0y(v%s#oXlUiQfNbK8H5D2A+x*u$qZnWkShPeMaJ7

22、G4C1z)w&t!pYmVjRgOcL9I6E3B0y(v%r#oXlTiQfNbK8G5D2A-x*u$qZnVkShPdMaJ7F4C0z)w&s!pYmUjRfOcL9H6E3B+y(u%r#oWlTiQeNbK8G5D1A-x*t$qZnVkSgPdMaI7F4C0z)v&s!pXmUjRfOcK9H6E2B+y(u%rZoWlThQeNbJ8G4D1A-w*t$qYnVjSgPdLaI7F3C0z)v&s#pXmUiRfOcK9H5E2B+x(u%rZoWkThQeMbJ8G4D1z-w*t!qYnVjSgOdLaI6F3C0y)v%s#pXlUiRfNcK8H5E2A+x(u$r

23、ZoWkThPeMbJ7G4D1z-w&t!qYmVjSgOdL9I6F3B0yrZoWkThQeMbJ8G4D1z-w*t!qYnVjSgOdLaI6F3C0y)v%s#pXlUiRfNcK9H5E2A+x(u$rZoWkThPeMbJ7G4D1z-w&t!qYmVjSgOdL9I6F3B0y)v%s#oXlUiQfNcK8H5D2A+x*u$rZnWkShPeMaJ7G4C1z-w&t!pYmVjRgOdL9I6E3B0y(v%s#oXlTiQfNbK8H5D2A-x*u$qZnWkShPdMaJ7F4C1z)w&s!pYmUjRgOcL9H6E3B+y(v%r#oWlTiQeNbK8G5

24、D2A-x*t$qZnVkShPdMaI7F4C0z)w&s!pXmUjRfOcL9H6E2B+y(u%r#oWlThQeNbJ8G5D1A-w*t$qYnVkSgPdLaI7F3C0z)v&s!pXmUiRfOcK9H6E2B+x(u%rZoWlThQeMbJ8G4D1A-w*t!qYnVjSgPdLaI6F3C0y)v&s#pXlUiRfNcK9H5E2A+x(u$rZoWkThQeMbJ7G4D1z-w*t!qYmVjSgOdLaI6F3B0y)v%s#pXlUiQfNcK8H5E2A+x*u$rZnWkThPeMaJ7G4C1z-w&t!pYmVjRgOdL9I6E3B0y(v%s#o

25、XlUiMbJ7G4D1z-w*t!qYmVjSgOdLaI6F3B0y)v%s#pXlUiQfNcK8H5E2A+x*u$rZnWkThPeMaJ7G4C1z-w&t!qYmVjRgOdL9I6F3B0y(v%s#oXlUiQfNbK8H5D2A+x*u$qZnWkShPeMaJ7F4C1z)w&t!pYmUjRgOcL9I6E3B+y(v%r#oXlTiQeNbK8G5D2A-x*u$qZnVkShPdMaJ7F4C0z)w&s!pYmUjRfOcL9H6E3B+y(u%r#oWlTiQeNbJ8G5D1A-x*t$qYnVkSgPdMaI7F3C0z)v&s!pXmUjRfOcK9H6E

26、2B+y(u%rZoWlThQeNbJ8G4D1A-w*t$qYnVjSgPdLaI7F3C0y)v&s#pXmUiRfNcK9H5E2B+x(u$rZoWkThQeMbJ8G4D1z-w*t!qYnVjSgOdLaI6F3C0y)v%s#pXlUiRfNcK8H5E2A+x(u$rZngPdLaI7F3C0y)v&s#pXmUiRfNcK9H5E2B+x(u%rZoWkThQeMbJ8G4D1z-w*t!qYnVjSgOdLaI6F3C0y)v%s#pXlUiRfNcK8H5E2A+x(u$rZnWkThPeMbJ7G4C1z-w&t!qYmVjSgOdL9I6F3B0y)v%s#oXlUi

27、QfNcK8H5D2A+x*u$rZnWkShPeMaJ7G4C1z)w&t!pYmVjRgOcL9I6E3B0y(v%r#oXlTiQfNbK8G5D2A-x*u$qZnWkShPdMaJ7F4C1z)w&s!pYmUjRgOcL9H6E3B+y(v%r#oWlTiQeNbK8G5D1A-x*t$qZnVkSgPdMaI7F4C0z)v&s!pXmUjRfOcL9H6E2B+y(u%r#oWlThQeNbJ8G5D1A-w*t$qYnRgOcL9H6E3B+y(v%r#oWlTiQeNbK8G5D1A-x*t$qZnVkSgPdMaI7F4C0z)w&s!pXmUjRfOcL9H6E2B+y

28、(u%r#oWlThQeNbJ8G5D1A-w*t$qYnVkSgPdLaI7F3C0z)v&s#pXmUiRfOcK9H5E2B+x(u%rZoWlThQeMbJ8G4D1A-w*t!qYnVjSgPdLaI6F3C0y)v&s#pXlUiRfNcK9H5E2A+x(u$rZoWkThPeMbJ7G4D1z-w&t!qYmVjSgOdLaI6F3B0y)v%s#pXlUiQfNcK8H5E2A+x*u$rZnWkThPeMaJ7G4C1z-w&t!pYmVjRgOdL9I6E3B0y(v%s#oXhPeMbJ7G4D1z-w*t!qYmVjSgOdLaI6F3B0y)v%s#pXlUiQfN

29、cK8H5E2A+x*u$rZnWkThPeMaJ7G4C1z-w&t!pYmVjRgOdL9I6E3B0y(v%s#oXlUiQfNbK8H5D2A+x*u$qZnWkShPeMaJ7F4C1z)w&t!pYmUjRgOcL9I6E3B+y(v%r#oXlTiQeNbK8G5D2A-x*t$qZnVkShPdMaI7F4C0z)w&s!pYmUjRfOcL9H6E3B+y(u%r#oWlTiQeNbJ8G5D1A-x*t$qYnVkSgPdMaI7F3C0z)v&s!pXmUiRfOcK9H6E2B+x(u%rZoWlThQeNbJ8G4D1A-w*t$UjRfOcL9H6E3B+y(u%r

30、#oWlTiQeNbJ8G5D1A-x*t$qYnVkSgPdMaI7F3C0z)v&s!pXmUiRfOcK9H6E2B+y(u%rZoWlThQeNbJ8G4D1A-w*t$qYnVjSgPdLaI7F3C0y)v&s#pXmUiRfNcK9H5E2B+x(u$rZoWkThQeMbJ7G4D1z-w*t!qYnVjSgOdLaI6F3C0y)v%s#pXlUiRfNcK8H5E2A+x(u$rZnWkThPeMbJ7G4C1z-w&t!qYmVjRgOdL9I6F3B0y(v%s#oXlUiQfNcK8H5D2A+x*u$rZnWkShPeMaJ7G4C1z)w&t!pYmVjRgOcL9I6E3B0y(v%r#oXlTiQfNbK8G5D2A-x*u$qZnVkShPdMaJ7F4C0z)w&s!pYmUjRgOcL9H6E3B+y(v%r#oWlTiQeNbK8G5D1A-x*t$qZnVkSgPdMaI7F4C0z)v&s!pXmUjRfOcK9H6E2B+y(u%rZoWlThQeNbF4C1z)w&s!pYmUjRgOcL9H6E3B+y(v%r#oWlTiQeNbK8G5D1A