Omr Research And Prototype System Development

Optical Music Recognition (OMR) is the automatic recognition of a scanned page of printed music notation by a computer program. It scans original paper musical score by digital equipments, such as scanner, and generates a bitmap of a scanned sheet of music, then processes and recognizes the music information in the bitmap file by using a special software system and at last turns the music information into general digital music format (such as MIDI format).In this diploma thesis, the statistic and structural characteristic of musical score image is analyzed and synthesized by relevant technology of image project, pattern recognition, mathematical morphology, software engineering, music knowledge, MIDI and so on. The concept of direction number has been defined, and then the mathematical morphology theory is used to process musical score image and recognize musical information. Specialized direction number algorithms are firstly used to preprocess a musical score image and then recognize stafflines, barlines, pitch, note values, clef, etc. Finally the musical information of the musical score image is automatically stored in the MIDI format. By using our proposed algorithm, multiple staves per system and oblique musical score can also be processed.An OMR prototype has been developed and designed according to the algorithm established in this thesis. Meanwhile, the prototype is used to recognize and process music information from 50 samples of digital music image .The result is that the recognition rate of our OMR system is more than 94%.The thought of establishing music statistics by digital music original database has also been put forward. Based on our OMR prototype system, music software can be developed and designed.