A new approach of staff lines removal for manuscript documents

Abstract

Staff lines detection process facilitates handwriting recognition and segmentation, especially for music symbols. It took part in various research works that allow manuscripts' recognition. In another way, staff lines removal remains a challenge for research by the fact that most document symbols are superimposed on staff lines and their removal causes the alteration of important notes and prevents the recognition process. In this paper, we present a new approach of staff lines removal, keeping intact the manuscript symbols and preserve the overall information shape. The proposed technique has been supported by various experiments where results show its efficiency and endorsed its competition against similar works. © 2020 IEEE.