a layered depth-of-field method for solving partial occlusion
DESCRIPTION
A Layered Depth-of-Field Method for Solving Partial Occlusion. David C. Schedl Michael Wimmer [email protected] [email protected] Institute of Computer Graphics and Algorithms Vienna University of Technology. WSCG – 26 th June 2012. Real life partial occlusion. - PowerPoint PPT PresentationTRANSCRIPT
A Layered Depth-of-Field Method for Solving Partial
OcclusionDavid C. Schedl Michael [email protected] [email protected]
Institute of Computer Graphics and AlgorithmsVienna University of Technology
WSCG – 26th June 2012
David C. Schedl
Real life partial occlusion
f=18 mm, N=4
David C. Schedl
Intro/Overviewdepth-of-field approximation
post processingpartial occlusion in realtime
David C. Schedl
Thin lens
David C. Schedl
Previous workPotmesil and Chakravarty, 1981
CoC - equationfirst post-processing methodblur according to CoCsstill a referenceartifacts
David C. Schedl
Artifacts
color bleeding:
[Demers2004]
[Riguer2005]
depth discontinuity:
David C. Schedl
Partial Occlusionpinhole vs. finite aperture
David C. Schedl
Partial Occlusion
pinhole: finite aperture:
David C. Schedl
Previous work – solve partial occlusion
non-realtime:ray-tracing (Cook et al., 1984)Accumulation B. (Haeberli and Akeley, 1990)
layered methods:Kraus and Strengert, 2007
occluded scene content only interpolated
Lee et al., 2010image composition via ray traversalsimulate more lens effectsmore complex and slower than ours
David C. Schedl
Our Method
David C. Schedl
Our Method
David C. Schedl
Overview
David C. Schedl
Rendering& Depth PeelingIm
ages
David C. Schedl
Matting – functionsw
eigh
t
depth
David C. Schedl
Matting – functionsw
eigh
t
depth
David C. Schedl
Matting – layers In
put
David C. Schedl
Matting – layers In
put
David C. Schedl
Matting – layers In
put
Laye
rs
David C. Schedl
Matting – layers In
put
Laye
rs
David C. Schedl
Matting – layers In
put
Laye
rs
David C. Schedl
Blurringuniformly blur layersGaussian filter
David C. Schedl
Blurring
David C. Schedl
Blurring
David C. Schedl
Composealpha-blend back to front
David C. Schedl
Optimizationreduce filter widthrecursive Gaussians
David C. Schedl
Optimization - front
David C. Schedl
Optimization - front
David C. Schedl
Optimization - front
David C. Schedl
Optimization - Compositing
David C. Schedl
Results - Homunculus
f=100mm, N =1.4, focus=18 500 mm, 17 layers, 3x DP
David C. Schedl
Results - Dragons
f=100mm, N =1.4, focus=3 000 mm, 22 layers, 3x DP
David C. Schedl
Results – BenchmarksIntel Core i7 920, Geforce GTX 480OpenGL and GLSL1024 x 1024px
ours Lee et. al. 2010 Accum. B.
optimized non-rec. 256 rays 256 views
Homunculus(74k tri.) 102 ms 1.4x 13.2x 47x
Dragons(610k tri.) 98 ms 1.3x 14.7x 42x
David C. Schedl
Conclusionlayered DoF methodpartial occlusion solvedcomparison to:
Accumulation Buffer Lee et al., 2010
optimized by recursive Gaussiansefficient composition with alpha blending
David C. Schedl
Outlookscreen-spaced antialiasingavoid empty layers: clusteringinaccurate but faster blurring methodscombine with eye-tracker
Thank you!
slides will be available at: http://www.cg.tuwien.ac.at/research/publications/2012/schedl-2012-dof/
David C. Schedl