PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Bayesian Color Constancy Revisited
Peter Gehler, Carsten Rother, Andrew Blake, Tom Minka and Toby Sharp
In: CVPR 2008, 24 Jun - 26 Jun 2008, Anchorage, USA.


Computational color constancy is the task of estimating the true reflectances of visible surfaces in an image. In this paper we follow a line of research that assumes uniform illumination of a scene, and that the principal step in es- timating reflectances is the estimation of the scene illumi- nant. We review recent approaches to illuminant estimation, firstly those based on formulae for normalisation of the re- flectance distribution in an image — so-called grey-world algorithms, and those based on a Bayesian formulation of image formation. In evaluating these previous approaches we introduce a new tool in the form of a database of 568 high-quality, in- door and outdoor images, accurately labelled with illumi- nant, and preserved in their raw form, free of correction or normalisation. This has enabled us to establish sev- eral properties experimentally. Firstly automatic selection of grey-world algorithms according to image properties is not nearly so effective as has been thought. Secondly, it is shown that Bayesian illuminant estimation is significantly improved by the improved accuracy of priors for illuminant and reflectance that are obtained from the new dataset.

PDF - Requires Adobe Acrobat Reader or other PDF viewer.
EPrint Type:Conference or Workshop Item (Paper)
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Machine Vision
ID Code:4727
Deposited By:Peter Gehler
Deposited On:24 March 2009