Artworks in the Spotlight: Characterization with a Multispectral Dome
Irina Ciortan, Tinsae Dulecha, Andrea GIachetti, Ruggero Pintus, Alberto Jaspe-Villanueva, and Enrico Gobbetti
IOP Conference Series: Materials Science and Engineering 364(1): 012025, 2018
Paper PDF Bibtex DOI CRS4 Website
@Article{Ciortan:2018:ASC,] author = {Irina Ciortan and Tinsae Dulecha and Andrea Giachetti and Ruggero Pintus and Alberto Jaspe-Villanueva and Enrico Gobbetti}, title = {Artworks in the Spotlight: Characterization with a Multispectral Dome}, journal = {IOP Conference Series: Materials Science and Engineering}, volume = {364}, number = {1}, pages = {012025}, year = {2018}, url = {http://vic.crs4.it/vic/cgi-bin/bib-page.cgi?id='Ciortan:2018:ASC'}, }
We describe the design and realization of a novel multispectral light dome system and the associated software control and calibration tools used to process the acquired data, in a specialized pipeline geared towards the analysis of shape and appearance properties of cultural heritage items. The current prototype dome, built using easily available electronic and lighting components, can illuminate a target of size 20cm x 20cm from 52 directions uniformly distributed in a hemisphere. From each illumination direction, 3 LED lights cover the visible range of the electromagnetic spectrum, as well as long ultraviolet and near infrared. A dedicated control system implemented on Arduino boards connected to a controlling PC fully manages all lighting and a camera to support automated acquisition. The controlling software also allows real-time adjustment of the LED settings, and provides a live-view of the to-be-captured scene. We approach per-pixel light calibration by placing dedicated targets in the focal plane: four black reflective spheres for back-tracing the position of the LED lamps and a planar full-frame white paper to correct for the non-uniformity of radiance. Once the light calibration is safeguarded, the multispectral acquisition of an artwork can be completed in a matter of minutes, resulting in a spot-wise appearance profile, that stores at pixel level the per-frequency intensity value together with the light direction vector. By performing calibrated acquisition of multispectral Reflectance Transformation Imaging (RTI), with our analysis system it is possible to recover surface normals, to characterize matte and specular behavior of materials, and to explore different surface layers thanks to UV-VIS-IR LED light separation. To demonstrate the system features, we present the outcomes of the on-site capture of metallic artworks at the National Archaeological Museum of Cagliari, Sardinia.