Abstract
The detection of staff lines is the first step of most Optical Music Recognition (OMR) systems. Its great significance derives from the ease with which we can then proceed with the extraction of musical symbols. All OMR tasks are usually achieved using binary images by setting thresholds that can be local or global. These techniques however, may remove relevant information of the music sheet and introduce artifacts which will degrade results in the later stages of the process. It arises therefore a need to create a method that reduces the loss of information due to the binarization. The baseline for the methodology proposed in this paper follows the shortest path algorithm proposed in [CardosoTPAMI08]. The concept of strong staff pixels (SSP's), which is a set of pixels with a high probability of belonging to a staff line, is proposed to guide the cost function. The SSP allows to overcome the results of the binary based detection and to generalize the binary framework to grayscale music scores. The proposed methodology achieves good results.
Highlights
Over the last years, several researchers have been trying to overcome the lack of symbolically-represented music
In [1] this concept was developed for binary images, stating that if we seek the shortest path between the two opposite vertical margins of a music score and consider that the cost of a path through black pixels is lower than a path through white pixels, the shortest paths will be along staff lines
When each column of a binary image is encoded with runlength encoding (RLE), the most frequent sum of two consecutive vertical runs is selected as the estimation of staffspace height+staffline height
Summary
Several researchers have been trying to overcome the lack of symbolically-represented music. Such representations of music scores enable operations such as search, retrieval and analysis. After some image preprocessing (which may include several techniques, e.g. binarization, noise removal, blurring, deskewing, amongst others, to make the recognition process more robust and efficient), an OMR system typically starts with a module for staff lines detection and removal to obtain an image containing only the musical symbols, where the recognition of the symbols is facilitated. Low paper quality or gradient effect on the illumination, very common in handwritten music scores, causes unsatisfactory binarization results. The line detection and removal methods in this paper are tailored for grayscale images of handwritten music scores. Luminance levels between staff lines, symbols and background present in the images
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
