Recovering registered geometry and high dynamic range texture with coded structured light

Abstract

In the last two decades the problem of accurately capturing the geometry of an object was extensively studied, while the acquisition of high-quality textures only in recent years has become subject of research. In most cases, geometry and texture are acquired separately and a registration step has to be carried out to correlate them. In this work we describe the usage of a color coded structured light. The traditional use of color in coding was revised in [Sa02b], and we had shown that, by projecting complementary slides, the reflectivity restrictions on objects can be eliminated. In this paper, by varying exposure between pairs of complementary slides, we are able to recover high dynamic range images and geometry at the same time, already registered to each other.