Skip to content

Documentation Status Coverage Tests PyPI - Python Version PyPI DOI

Unraphael banner Unraphael banner

Unraphael

Unraphael is a digital workflow tool that uses computer vision to unravel the artistic practice of Raphael (Raffaello Sanzio, 1483-1520), while providing new digital approaches for the study of artistic practice in art history. Dozens of faithful reproductions survive of Raphael's paintings, attesting to the lucrative practice of serial production of paintings within the artist's workshop and to the lasting demand for the master's designs.

Unraphael provides a flexible and easy-to-use GUI to inspect and assess the structural similarity of figure-outlines in images. Photographs of paintings are used as input for the application.

While Unraphael was made for art historians and researchers in the humanities to study the artistic practices of and the process of making copies of paintings, the functionality of Unraphael extends well beyond the study of Raphael's paintings and can be used for a wide range of applications in the digital humanities and beyond.

Features: - Image preprocessing - Background removal - Image alignment - Image clustering based on structural similarity

To install:

pip install unraphael

Try unraphael in your browser!

You can also try unraphael directly from your browser.

Link Description Image
Image similarity Group your images using cluster analysis Image similarity
Image preprocessing Preprocess your images, e.g. background removal, color adjustments, applying image filters, segmentation Image preprocessing
Object detection Quickly and accurately identify and segment figures or objects within an image to analyse the isolated components Object detection
Image comparison Compare your images based on their structural components Image comparison

Using the unraphael dashboard locally

To install and use the dashboard locally:

pip install unraphael[dash]
unraphael-dash

Development

Check out our Contributing Guidelines to get started with development.

Suggestions, improvements, and edits are most welcome.