Estimation of multiple directional illuminants from a single image

We present a new method for the detection and estimation of multiple directional illuminants, using only one single image of an object with known geometry. It obviates the need to modify the imaged scene by inserting calibration objects of any particular geometry, relying instead on partial knowledge of the geometry of the scene. We first develop our method for the case of a Lambertian sphere with known size, illuminated by a set of directional light sources. A novel and robust way is proposed to segment the surface into regions, with each region illuminated by a different set of sources. Our region-based least-squares method is impervious to noise and missing data, which is crucial to extending the method to arbitrary smooth geometry and to surfaces having both Lambertian and specular properties. We propose a novel methodology that integrates information from shadows and shading in the presence of strong directional sources of illumination, even when significant non-directional sources exist in the scene and the object surface is not Lambertian. We demonstrate experimentally the accuracy of our method, both in detecting the number of light sources and in estimating their directions, by testing on images of a variety of synthetic and real objects.