Fakultät Informatik

Depth-of-Field Rendering by Pyramidal Image Processing

 Martin Kraus*,  Magnus Strengert

*Computer Graphics and Visualization Group, Technische Universität München, Germany
Visualization and Interactive Systems Group, Universität Stuttgart, Germany


We present an image-based algorithm for interactive rendering depth-of-field effects in images with depth maps. While previously published methods for interactive depth-of-field rendering suffer from various rendering artifacts such as color bleeding and sharpened or darkened silhouettes, our algorithm achieves a significantly improved image quality by employing recently proposed GPU-based pyramid methods for image blurring and pixel disocclusion. Due to the same reason, our algorithm offers an interactive rendering performance on modern GPUs and is suitable for real-time rendering for small circles of confusion. We validate the image quality provided by our algorithm by side-by-side comparisons with results obtained by distributed ray tracing.

Depth-of-field effect added by GPU image postprocessing to a ray-traced pbrt scene

Associated publications

Depth-of-Field Rendering by Pyramidal Image Processing
M. Kraus, M. Strengert, EG 2007 [Bibtex]


accompanying video of the paper [33 MB MOV]



- In collaboration with partners from industry, we have a number of thesis topics available in the area of point-based rendering, geo-localization using public data, scene fusion from different viewpoints. If you are interested, please contact  westermann(at)tum.de


- One PhD position on   Turbulence Visualization is available at the Computer Graphics & Visualization group.