Particle.news
Download on the App Store
8 ARTICLES
·
yesterday

MIT’s AI-Printed Mask Restores Damaged Paintings in 3.5 Hours

Printed on thin polymer films, the AI-generated masks can be removed without harming the artwork, speeding conservation up to 70 times faster than traditional techniques.

Scans of the painting during various stages in its restoration. At left is the damaged piece, with the middle panel showing a map of the different kinds of damage present; green lines show full splits in the underlying panel support, thin red lines depict major paint craquelure, blue areas correspond to large paint losses, while pink regions show smaller defects like scratches. At right is the restored painting with the applied laminate mask.
Image
Image
Image

Overview

  • MIT graduate student Alex Kachkine developed the method by scanning a cleaned painting, creating a digital restoration and printing it onto a two-layer polymer mask.
  • In tests on a 15th-century oil painting, the technique infilled over 5,600 damaged regions using 57,000 colours in just 3.5 hours.
  • The thin removable films adhere to flat, varnished canvases and can be peeled off or dissolved with conservation-grade solutions.
  • A digital record of each mask preserves exact restoration details for future conservators and helps maintain fidelity to the artist’s original intent.
  • The approach is most effective on smooth oil paintings and requires conservator oversight to address ethical and stylistic considerations.