Integrating Innovative Spatial and Spectral Data Fusion Strategies in Hyperspectral Imaging for Cultural Heritage
Alessia Di Benedetto (Politecnico di Milano)
Elisabetta Martinelli (Politecnico di Milano)
Sabrina Samela (Politecnico di Milano)
Paulina Guzmán García Lascurain (Politecnico di Milano)
Cristian Manzoni (Istituto di Fotonica e Nanotecnologie, Consiglio Nazionale delle Ricerche)
M.W.E.M. Alfeld (TU Delft - Team Matthias Alfeld)
Daniela Comelli (Politecnico di Milano)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
The study of cultural heritage (CH) objects benefits greatly from non-invasive techniques like hyperspectral imaging (HSI), which enables material identification and spatial mapping. Due to the heterogeneous composition of CH artifacts, combining complementary techniques is essential for comprehensive analysis. However, handling such high-dimensional datasets remains a challenge. We present a computational protocol that combines spatial and spectral dimensionality reduction to enable early-stage fusion and efficient analysis of fused data, through multivariate methods, with a focus on Uniform Manifold Approximation and Projection (UMAP). We introduce an open-source plugin for Napari viewer, which allows for UMAP-based exploration of fused multimodal datasets. Our approach is demonstrated in case studies involving reflectance and photoluminescence data fusion, showcasing its effectiveness in detecting degradation phenomena and revealing material complexity in both plastic artifacts and historical paintings.
help_outline