Project Overview
The Adair Chant Book project, conducted in two phases, centers on the study of a handwritten codex owned by the University of Missouri-Kansas City Libraries. The project utilizes multispectral analysis to examine the book's physical aspects and uncover its history.
Phase 1 (2014-2016) focused on using multispectral analysis to study rare handwritten books and developing an affordable mechanical apparatus for this type of analysis. A key deliverable was the pilot investigation of the palimpsests in the Adair Chant Book
Phase 2 (2020-2024) aimed to study palimpsests in the Adair Chant Book in detail using a home-built, multispectral imaging system. This phase also focused on developing a deep learning model to derive visible light multispectral images from normal RGB images, with the goal of making this technology more accessible. The project also incorporated traditional codicology to understand the book's form and context.
Key Deliverables
Phase 1 (2014-2016) Key Deliverables:
- A pilot investigation of the Adair Chant Book.
- Analyses of the Prague Missal and the Menaion Service Book.
- A pilot hand-built system for multispectral analysis.
- Publications and presentations about the work.
Phase 2 (2020-2024) Key Deliverables:
- A deep learning model to derive visible light multispectral images from normal RGB images.
- A refined home-built, multispectral imaging system.
- A complete inventory of the chants in the Adair Chant Book in the CANTUS database.
- An image gallery of the Adair Chant Book containing RGB, MSI, and SSR images with a zoom feature for comparing the images.
- A white paper describing the development, testing and evaluation of the SSR deep-learning model.