A Practical Reflectance Transformation Imaging Pipeline for Surface Characterization in Cultural Heritage

We present a practical acquisition and processing pipeline to characterize the surface structure of cultural heritage objects. Using a free-form Reflectance Transformation Imaging (RTI) approach, we acquire multiple digital photographs of the stud- ied object shot from a stationary camera. In each photograph, a light is freely positioned around the object in order to cover a wide variety of illumination directions. Multiple reflective spheres and white Lambertian surfaces are added to the scene to automatically recover light positions and to compensate for non-uniform illumination. An estimation of geometry and re- flectance parameters (e.g., albedo, normals, polynomial texture maps coefficients) is then performed to locally characterize surface properties. The resulting object description is stable and representative enough of surface features to reliably provide a characterization of measured surfaces. We validate our approach by comparing RTI-acquired data with data acquired with a high-resolution microprofilometer