Project Summary

The aim of the project is to develop a multi-criterion semi-automatic authentication method for medieval and pre-modern paintings employing a multi-criterion analysis tool using convergent approach in order to determine the degree of similarity between two or more artworks. The automation of the authentication method is based on the processing of the artwork’s image taken both in the visible and invisible spectra, which will be correlated with the information acquired from the historical, artistic, stylistic and iconographical analysis of the artworks, as well as with the physical-chemical examination of the pigments found in the color layer, where possible. Using image processing a number of geometric key elements of a painting will be identified employing image decomposition (ROI), which will then be searched for in other paintings from the database. Identifying geometric features will be done in several steps and involves the use of image processing, which will be performed by applying various tools such as Edge detection, FFT or Cross Correlation. After identifying the similar geometric elements between two or more paintings, an analysis will be performed using relevant information provided by art historians or using the data obtained from the physical-chemical analysis of the pigments in the color layer and on the basis of a comparison using this data, a correlation score of the studied painting with those in the database will be calculated and the image will be assigned to one of the categories in the database. The final result will be evaluated by an art historian in order to confirm the paternity of the work. The advantage of the method is that both the search by geometric elements, as well as the one making use of other features specific to painting, are done automatically in a database, the required time for authenticating an artwork becoming thus increasingly shorter. The database will be implemented online, so it can be easily accessed through a web browser, which will facilitate access for an increasing number of researchers and even the general public.

Investigation methods




George Oprescu Institute of Art History from Bucharest


Partner 1

Technical University of Cluj-Napoca


Partner 2

Babes-Bolya University


Partner 3

National Institute for Research Development for Isotopic and Molecular Technologies


Partner 4

Romanian Academy Cluj Branch - Institute of Archeology and Art History