Rapid Synchronous Acquisition of Geometry and Appearance of Cultural Heritage Artefacts

Abstract

In order to produce visually appealing digital models of cultural heritage artefacts, a meticulous reconstruction of the 3D geometry alone is often not sufficient, as colour and reflectance information give essential clues of the object s material. Standard texturing methods are often only able to overcome this fact under strict material and lighting condition limitations. The realistic reconstruction of complex yet frequently encountered materials such as fabric, leather, wood or metal is still a challenge. In this paper, we describe a novel system to acquire the 3Dgeometry of an object using its visual hull, recorded in multiple 2D images with a multi-camera array. At the same time, the material properties of the object are measured into Bidirectional Texture Functions (BTF), that faithfully capture the mesostructure of the surface and reconstruct the look-and-feel of its material. The high rendering fidelity of the acquired BTF texture data with respect to reflectance and self-shadowing also alleviates the limited precision of the visual hull approach for 3D geometry acquisition.