Computational methods for artwork authentication using hyperspectral imaging

Authentication of physical artworks is essential for safeguarding cultural heritage, enforcing artist’s intellectual property rights, ensuring consumer protection and the credibility of the art market. The creation of certificates of authenticity based on digital signatures is an innovative response to shortcomings in current authentication practices.

This research project will use hyperspectral imaging and computational methods to create unique digital signatures for physical artworks. New developments in hyperspectral imaging technology have made it possible to use a portable front facing hyperspectral camera to record the spatial and spectral features of an artwork, in a non-invasive manner.

Each hyperspectral image produces a ‘hypercube’ of approximately 5.3 million data points, most of which are in the non-visible spectrum. We will investigate algorithms, such as manifold learning, to condense the hypercube to a lower dimensional representation to find a unique digital signature for the given artwork. This signature can then be used to authenticate .

It will be necessary to allow for and define an acceptable deviation range or ‘sphere of authenticity’ to account for minor variations in the image-capture process and possible age-related deterioration of the artwork. The digital signature and sphere of authenticity will then be stored on an authentication registry supported by blockchain infrastructure. An associated non-fungible token (NFT) will be minted as a certificate of authenticity.

This ability to authenticate an artwork in a non-invasive manner would be valuable to artists, commercial art galleries, auction houses, art collectors, museums, and other cultural institutions. The project’s scope could be expanded to include other cultural objects, such as Indigenous artifacts. We could also use similar computational methods to create a ‘spectral library’ of an artist’s body of work. This could protect against unknown or unauthenticated artworks that may appear on the art market after the artist’s death.

Who's involved

Chief Investigator

Dr Guy Morrow (CI) & Dr David Challis (PL), Faculty of Arts, School of Culture and Communication

MDAP team

Karen Thompson, Kabir Manandhar Shrestha, Aleks Michalewicz, Simon Mutch