The Optical Neume Recognition Project
A tool to investigate early staff-less music notation
This project is to design and develop a historical document analysis system as a computer assistant for musicologists and historians to help them in their work in studying early western musical manuscripts. Further we intend to use the resulting system to explore open questions in historical musicology.
The guiding premise is to construct a system that experts in the field of musicology can use to study the subject documents and explore research hypotheses without requiring advanced computer science expertise. For this reason, features and interfaces of the system are being designed in close cooperation between computer scientists and research musicologists.
From a computer science perspective, the system will include the latest technology in historical document analysis, such as specialised noise reduction, binarisation, layout analysis, feature extraction and symbol recognition, word spotting and more. Statistical tools for clustering and classification will also be provided.