Bright Chromagenic Algorithm for Illuminant Estimation

Abstract

This article proposes a new algorithm for illuminant estimation based on the concept of chromagenic color constancy, where two pictures are taken from each scene: A normal one and one where a colored filter is placed in front of the camera. The basic formulation of the chromagenic algorithm has inherent weaknesses, namely, a need for perfectly registered images and occasional large errors in illuminant estimation. Our first contribution is to analyze the algorithm performance with respect to the reflectances present in a scene and demonstrate that fairly bright and de-saturated reflectances (e.g., achromatic and pastel colors) provide significantly better chromagenic illuminant estimation. We thus propose the bright chromagenic algorithm and show that it not only remedies the large error problem but also allows us to relax the image registration constraint. Experiments performed on a variety of synthetic and real data show that the newly designed bright chromagenic algorithm significantly outperforms current illuminant estimation methods, including those having a substantially higher complexity.