Premium
Nonlinear Unmixing of Hyperspectral Datasets for the Study of Painted Works of Art
Author(s) -
Rohani Neda,
Pouyet Emeline,
Walton Marc,
Cossairt Oliver,
Katsaggelos Aggelos K.
Publication year - 2018
Publication title -
angewandte chemie international edition
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 5.831
H-Index - 550
eISSN - 1521-3773
pISSN - 1433-7851
DOI - 10.1002/anie.201805135
Subject(s) - hyperspectral imaging , basis (linear algebra) , pixel , artificial intelligence , pigment , set (abstract data type) , nonlinear system , reflectivity , data set , computer science , cube (algebra) , remote sensing , pattern recognition (psychology) , biological system , mathematics , computer vision , chemistry , optics , geology , physics , geometry , organic chemistry , quantum mechanics , biology , programming language
Abstract Nonlinear unmixing of hyperspectral reflectance data is one of the key problems in quantitative imaging of painted works of art. The approach presented is to interrogate a hyperspectral image cube by first decomposing it into a set of reflectance curves representing pure basis pigments and second to estimate the scattering and absorption coefficients of each pigment in a given pixel to produce estimates of the component fractions. This two‐step algorithm uses a deep neural network to qualitatively identify the constituent pigments in any unknown spectrum and, based on the pigment(s) present and Kubelka–Munk theory to estimate the pigment concentration on a per‐pixel basis. Using hyperspectral data acquired on a set of mock‐up paintings and a well‐characterized illuminated folio from the 15th century, the performance of the proposed algorithm is demonstrated for pigment recognition and quantitative estimation of concentration.