GPU Spectral Viewer: Analysing Paintings from a Colorimetric Perspective

Abstract

Over the last fifteen years, multipsectral imaging has gained in importance and interest, especially in the field of Cultural Heritage, art investigation and conservation. Extending the concept of scientific imagery such as colorimetric, infrared reflectography (IRR), ultraviolet (UV) and X-ray imaging applied to the study of paintings, multispectral imaging, coupled with high resolution and HDR (high dynamic range) has significantly improved the scope and accuracy of the non-invasive scientific analysis that is possible. In order to exploit and study such multispectral data, a special GPU-based application using a custom color management process has been developed. In this paper we will present its innovative capabilities in image processing and visualization which enhance the study of works of art.