Music Score Recognition with Waves

Purnima Weliwitigoda, Ruvan Weerasinghe, Dulan Wathugala, Anuja Dharmaratne


This research is an attempt to automate musicians’ work involved in composing music. When it comes to music, the most powerful instruments, which generate beautiful music, are not electronic. Classic examples are piano, violin, clarinet, flute, oboe and other orchestral instruments. In this paper, music from a non-electronic instrument is captured from a simple microphone and subsequently recognized with the aid of a computer. It has proved that without MIDI, music can still be recognized. The core of the research, involves the segmentation of notes and frequency identification. It is done using energy distribution and Fourier transformation. The system is implemented in MatLab, which provides facilities for Fourier transformation. The research’s initial goal is to come up with the graphical score of the music played. As composers generating sheet music find it a difficult task, automating such a task is a great achievement. At the end of the recognition, the set of frequencies and their durations are identified. This has provided the ability to generate MIDI form of the music played. In this paper, the methodology used to do the recognition of music scores, is explained in detail.

Citation Info :

In Conference Proceedings - 6th International Information Technology Conference on From Research to Reality, Infotel Lanka Society Colombo, Sri Lanka, 29 Nov- 01 Dec 2004, pp. 40-48, ISBN 955-8974-01-3.