CSE 274 Fall 2015 Lecture 6 ImageBased Rendering and Light Fields http wwwcsucsdedu ravir To Do Project Milestone Reports Due Oct 27 12 page PDF or weblink with at least one image or video to show current results ID: 557304
Download Presentation The PPT/PDF document "Real-Time High Quality Rendering" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Real-Time High Quality Rendering
CSE 274 [Fall 2015], Lecture 6Image-Based Rendering and Light Fields
http://
www.cs.ucsd.edu
/~
ravirSlide2
To Do
Project Milestone Reports Due Oct 271-2 page PDF or weblink with at least one image or video to show current results Describe project, progress to date, and final proposal based on results Slide3
Motivation for Lecture
IBR is tangentially related to course (more pertinent to acquisition of measured materials and scenes)But many of the rendering methods, especially precomputed techniques borrow from itAnd many methods use measured data
Also, images are an important source for renderingLater: Signal-Processing techniques very relevantSlide4
Next few slides courtesy Paul Debevec; SIGGRAPH 99 course notesSlide5Slide6Slide7Slide8
IBR: Pros and Cons
AdvantagesEasy to capture images: photorealistic by definitionSimple, universal representationOften bypass geometry estimation?
Independent of scene complexity?DisadvantagesWYSIWYG but also WYSIAYGExplosion of data as flexibility increasedOften discards intrinsic structure of model?Today, IBR-type methods also often used in synthetic rendering (e.g. real-time rendering PRT) General concept of data-driven graphics, appearance
Also, data-driven geometry, animation, simulationSpawned light field cameras for image capture Slide9Slide10
IBR: A brief history
Texture maps, bump maps, environment maps [70s]Poggio MIT 90s: Faces, image-based analysis/synthesisMid-Late 90s Chen and Williams 93, View Interpolation [Images+depth]Chen 95 Quicktime VR [Images from many viewpoints]
McMillan and Bishop 95 Plenoptic Modeling [Images w disparity]Gortler et al, Levoy and Hanrahan 96 Light Fields [4D]Shade et al. 98 Layered Depth Images [2.5D]Debevec et al. 00 Reflectance Field [4D]Inverse rendering (Marschner,Sato,Yu,Boivin,…)Today: IBR hasn’t replaced conventional rendering, but has brought sampled and data-driven representations to graphicsSlide11Slide12Slide13
Outline
Overview of IBRBasic approachesImage Warping [2D + depth. Requires correspondence/disparity]Light Fields [4D]Survey of some
early workSlide14
Warping slides courtesy Leonard McMillan, SIGGRAPH 99 course notesSlide15Slide16Slide17Slide18Slide19Slide20Slide21
Demo: Lytro Perspective Shift
See demos at http://pictures.lytro.com/Notice image is everywhere in focus
Only small motions, interpolate in apertureSlide22Slide23Slide24Slide25Slide26
Outline
Overview of IBRBasic approachesImage Warping [2D + depth. Requires correspondence/disparity]Light Fields [4D]Survey of some
early workSlide27Slide28Slide29Slide30Slide31Slide32
Lytro
CameraSlide33Slide34Slide35Slide36Slide37
Outline
Overview of IBRBasic approachesImage Warping [2D + depth. Requires correspondence/disparity]Light Fields [4D]Survey of some
early workSlide38
Layered Depth Images [Shade 98]
Geometry
Camera
Slide from Agrawala, Ramamoorthi, Heirich, Moll, SIGGRAPH 2000Slide39
Layered Depth Images [Shade 98]
LDISlide40
Layered Depth Images [Shade 98]
LDI
(Depth, Color)Slide41Slide42
Miller 98, Nishino 99, Wood 00
Reflected light field (lumisphere) on surface
Explicit geometry as against light fields. Easier compressSurface Light FieldsSlide43
Acquiring Reflectance Field of Human Face [Debevec et al. SIGGRAPH 00]
Illuminate subject from many incident directionsSlide44
Example ImagesSlide45
Outline
Overview of IBRBasic approachesImage Warping [2D + depth. Requires correspondence/disparity]Light Fields [4D]Survey of some recent work
Sampled data representationsSlide46
Conclusion (my views)
IBR initially spurred great excitement: revolutionize pipelineBut, IBR in pure form not really practicalWYSIAYGExplosion as increase dimensions (8D transfer function)Good compression, flexibility needs at least implicit geometry/BRDF
Real future is sampled representations, data-driven methodAcquire (synthetic or real) data Good representations for interpolation, fast renderingMuch of visual appearance, graphics moving in this directionUnderstand from Signal-Processing ViewpointSampling rates, reconstruction filtersFactored representations, Fourier analysisLight Fields fundamental in many ways, including imaging