The Zeta-image, illuminant estimation, and specularity manipulation

Identification of illumination, the main step in colour constancy processing, is an important problem in imaging for digital images or video, forming a prerequisite for many computer vision applications. In this paper we present a new and effective physics-based colour constancy algorithm which makes use of a novel Log-Relative-Chromaticity planar constraint. We call the new feature the Zeta-image. We show that this new feature makes use of a novel application of the Kullback–Leibler Divergence, here applied to chromaticity values instead of probabilities. The new method requires no training data or tunable parameters. Moreover it is simple to implement and very fast. Our experimental results across datasets of real images show that the proposed method significantly outperforms other unsupervised methods while its estimation accuracy is comparable with more complex, supervised, methods. As well, we show that the new planar constraint can be used as a post-processing stage for any candidate colour constancy method in order to improve its accuracy. Its application in this paper demonstrates its utility, delivering state of the art performance. The Zeta-image is a wholly new representation for understanding highlights in images, and we show as well that it can be used to identify and remove specularities. More generally, since the Zeta-image is intimately bound up with specularities, we show how specular content in the image can be manipulated, either decreasing or increasing highlights.