Shade analysis on facial images for robotic lighting

Abstract

Impression of a facial image is deeply depends on the shades and shadows cast on a face, and smart lighting control is therefore a key technology for obtaining its appealing appearance. The optimal control of lighting is usually done by a skillful photographer or lighting technician in a commercial studio, but the effect of such skillful lighting can be extracted from the features of shading pattern on professional images. The target of our research is to develop a robotized LED lighting system for automatically controlling its positions and orientations to supply expected optical effects. This article focuses on a method of estimating the lighting condition from a sample image, by which a similar shading style is obtained for the facial image taken under our robotic lighting.