Image Processing Algorithms for Art Conservation (2017-2018)

Background

Imaging technology is increasingly being used by museums to digitally preserve artwork, as well as to analyze and nondestructively reconstruct paintings. A major accomplishment in this field was the digitization of the Ghent Altarpiece in high-resolution images, and x-ray and infrared modalities.

Technologies developed at Duke University include software to remove the wood cradle that was used to reinforce aging canvas and separate the texture of canvas from paint.

This project will extend the work accomplished by the 2016-2017 Bass Connections team and a Summer 2017 Data+ project.

Project Description

The scope of this project is wide. The collection of paintings that can be better understood through digital analysis is always expanding, and each painting presents a unique challenge. Much of the work will be done in collaboration with the North Carolina Museum of Art and team members will work alongside engineers and art historians.

Researchers at Duke previously digitally rejuvenated a 14th century altarpiece by Francescuccio Ghissi depicting scenes from St. John’s life. The process involved detection and painting of cracks formed due to the expansion and contraction of the wood panel of the painting, restoration of aged color pigments and animation of the gold gilding. Even though the results were good, the parts of the process were executed manually under an expert’s supervision. The Data+ team streamlined and automated the rejuvenation process with minimal user inputs by building a prototype for an app in open source software. Work will continue on automating the rejuvenation process to make it feasible for distribution to a wider audience in the form of a website, mobile app and crowdsourcing.

Additional work will be done on the Peruzzi Altarpiece painted in the 14th century by Giotto di Bondone, studies of Van Gogh paintings and the Ghent Altarpiece.

Anticipated Outcomes

Further development of a user interface for the rejuvenation of artwork; products in the form of code and images to art collectors and historians

Timing

Fall 2017 – Spring 2018

Crediting

Course credit available for fall and spring semesters

See earlier related projects, Image Processing Algorithms for Art Conservation (2016-2017) and Smartphone-assisted Digital Rejuvenation of Medieval Paintings (Data+ 2017).

Faculty/Staff Team Members

Ingrid Daubechies, Trinity - Mathematics*
Ed Triplett, Wired! Lab*

Graduate Team Members

Gilad Amitai, Statistical Science - MS

Undergraduate Team Members

Lizzet Clifton, Visual and Media Studies (AB)
Michael Gattas, Economics (BS), Computer Science (AB2)
Nazli Gungor, Electrical & Computer Egr(BSE), Computer Science (BS2)
Hojung (Ashley) Kwon, Art History (AB)
Mitchell Parekh, Computer Science (BS)
Jia Chen (Ivy) Shi
Thomas Tilton, Computer Science (BS), Philosophy (AB2)

Community Team Members

North Carolina Museum of Art

* denotes team leader

Status

Active