English
 
Help Privacy Policy Disclaimer
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT

Released

Conference Paper

Bayesian Relighting

MPS-Authors
/persons/resource/persons44457

Fuchs,  Martin
Computer Graphics, MPI for Informatics, Max Planck Society;
Computer Graphics, MPI for Informatics, Max Planck Society;

/persons/resource/persons44144

Blanz,  Volker
Computer Graphics, MPI for Informatics, Max Planck Society;

/persons/resource/persons45449

Seidel,  Hans-Peter       
Computer Graphics, MPI for Informatics, Max Planck Society;

/persons/resource/persons44753

Keller,  Alexander
Computer Graphics, MPI for Informatics, Max Planck Society;

External Resource
No external resources are shared
Fulltext (restricted access)
There are currently no full texts shared for your IP range.
Fulltext (public)
There are no public fulltexts stored in PuRe
Supplementary Material (public)
There is no public supplementary material available
Citation

Fuchs, M., Blanz, V., & Seidel, H.-P. (2005). Bayesian Relighting. In Rendering Techniques 2005: Eurographics Symposium on Rendering (pp. 157-164). Aire-la-Ville, Switzerland: Eurographics Association.


Cite as: https://hdl.handle.net/11858/00-001M-0000-000F-25ED-0
Abstract
We present a simple method for relighting real objects viewed from a fixed camera position. Instead of setting up a calibrated measurement device, such as a light stage, we manually sweep a spotlight over the walls of a white room, illuminating the object indirectly. In contrast to previous methods, we use arbitrary and unknown angular distributions of incoming light. Neither the incident light nor the reflectance function need to be represented explicitly in our approach.\\ % The new method relies on images of a probe object, for instance a black snooker ball, placed near the target object. Pictures of the probe in a novel illumination are decomposed into a linear combination of measured images of the probe. Then, a linear combination of images of the target object with the same coefficients produces a synthetic image with the new illumination. We use a simple Bayesian approach to find the most plausible output image, given the picture of the probe and the statistics observed in the dataset of samples.\\ % Our results for a variety of novel illuminations, including synthetic lighting by relatively narrow light sources as well as natural illuminations, demonstrate that the new technique is a useful, low cost alternative to existing techniques for a broad range of objects and materials.