A 3D Hyperspectral tool for preventive conservation

With palace-museums facing the oddities of quick weather variations in a context of climate change, like sudden degradation of ceiling paintings or strong warping of tapestries, a need for monitoring tools and damage precursor measurements is emerging to help preventive conservation strategies. This can be achieved with new high-end multidimensional processing tools as decision-support system. To that end, the coupling of a hyperspectral imaging (HSI) system – covering both visible and short-wave infrared areas – and 3D LiDAR data was set up in the Queen’s Bedroom and Mercury Room of the Versailles Palace for seasonal takes. A sample of representative elements was first selected according to the EPICO (European Protocol In preventive COnservation) [1] method for each room, then followed thoroughly as seasons changed for further investigation. Once per season, three sets of data are collected for each designated work of art: visible, infrared and surface data. This allows us to monitor their evolution in regards of their nature and the weather conditions. To this day, three seasons are already covered: Summer (July 2025), Autumn (November 2025) and Winter (February 2026). The following example (fig. 1) showcases the Queen’s “serre-bijoux” (jewel displaying case). This specific work of art is interesting as humidity levels can variate in materials such as wood and could be related to hygrometry fluctuations inside the rooms thanks to the infrared data [2]. For this purpose, the different takes of the serre-bijoux are compared, which upbrings technical levers: the angles can be different, the resolution of each camera and the distance too. This research aims at aligning these data in order to make them comparable and extract valuable content for preventive conservation purposes. References: [1] D. Forleo and N. Francaviglia, 2020: “Conserver les collections des demeures historiques : application de la méthode d’évaluation EPICO au château de Maintenon” [2] Z. Yuan et al., 2025: “Rapid and non-destructive detection of wood density based on NIR hyperspectral imaging technology and moisture correction methods,” Spectrochim Acta A Mol Biomol Spectrosc, vol. 341, doi: 10.1016/j.saa.2025.126410. This HypErPICO work has benefited from French State aid managed by the Agence Nationale de la Recherche under the future investment program integrated into France 2030, bearing the reference ANR 21-ESRE-0050 EquipEx+ ESPADON and École Universitaire de Recherche Paris Seine Humanités, Création, Patrimoine - Fondation des sciences du patrimoine ANR-17-EURE-0021.