Source-code + data set added: Illuminant Chromaticity from Image Sequences!

In 2013, VĂ©ronique Prinet, Dani Lischinski and Michael Werman presented their work during the International Conference on Computer Vision on “Illuminant chromaticity from image sequences”. In this paper, they proposed a physically-based method to recover the illuminant from temporal sequences, without the need for user interaction or using strong assumptions and heuristics. The method is based on the assumption that the incident light chromaticity is constant over a short space-time domain, and it can even be applied to scenes illuminated by two global light sources assuming the incident light chromaticity can be modeled by a mixture of the two illuminant colors. Now, they made available the source-code of this method, as well as the data sets they experimented on. For more information on the project, please visit their project website.