Art objects conservation or historical analysis necessitates a thorough knowledge of the materials used by the artist and during the subsequent changes, their chemical composition and determination of their preservation state. In the case of paintings this requires the ability to correctly identify the pigments that were used for creation or later restoration of the artwork. This is a challenging problem, as the applied method should be non-contact, robust for the wide variety of chemical substances used and straightforward in the interpretation. Recently, the hyperspectral imaging has emerged as a promising measuring methodology for this kind of the artwork analysis; the combination of acquiring spectral information and planar (photography-like) pixel arrangement provides a lot of potential for material characterization. While initial studies of hyperspectral imaging application to art objects analysis are encouraging, the difficulties of working with its multidimensional data are acknowledged; in many cases complex algorithms are required to fully utilize its potential.

In the paper Automatic pigment identification from hyperspectral data, by B. Grabowski, W. Masarczyk, P. Głomb, A. Mendys, published on Journal of Cultural Heritage, Vol. 31, 2018, Pages 1-12,, the problem of algorithm design for pigment identification based on a hyperspectral image of a painting is investigated.
Previous studies suggested a potential of the hyperspectral imagery as a data source, and proposed several methods that could be used as elements of the pigment identification algorithm. However, an automatic identification scheme transferable to many different art objects requires a multi-stage processing consisting of a number of methods. The authors of this paper argue that the problem of building an automatic multi-stage system for this task has not been previously explored, so they investigate the situation with the need for identification of a pigment component fully mixed with other elements of a painting (e.g. with a binder, priming and canvas). This is an important distinction from many previous works as the identification library contains just pigments and is created independently of the object being studied. Various processing steps are combined to achieve a robust solution requiring minimal user intervention. Using a special set of paintings and a reference pigment database the viability of applying this method in the pigment recognition setting is demonstrated.
The results confirm the potential of using hyperspectral imaging in the art conservation setting, and based on them we discuss the potential construction and elements of such an algorithm. The introduced approach may become a helpful tool in the art objects conservation work, as it allows to automatically process spectrally mixed data from a hyperspectral camera.
The work has shown the use of advanced algorithms to support chemical analysis of paint layers of cultural heritage objects. Hyperspectral analysis provides additional information in a non-invasive way. The proposed approach has shown considerable potential in distinguishing pigments of similar colours and similar elemental composition, therefore it can be a real aid for art conservators. As the results for test samples are promising, the future work could include expanding the spectral database by adding other colours and evaluation of pigment identification on real cultural heritage objects.