Skip to content

Staff detection for computer generated and handwritten scores

Notifications You must be signed in to change notification settings

ferasboulala/staff-detect

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

staff-detect

Implementation of the paper bellow. Staff line detection can be used to get a grounding in OMR (Optical Music Recognition). It can handle computer generated scores aswell as most handwritten scores.

Datasets and papers

MUSCIMA++

DeepScores

@article{   su2012,
  author =  {Bolan Su , Shijian Lu , Umapada Pal and Chew Lim Tan},
  title =   {An Effective Staff Detection and Removal Technique for Musical Documents},
  journal = {2012 10th IAPR International Workshop on Document Analysis Systems},
  year =    {2012}
}

Installation

(Requires OpenCV and Boost libraries).

  1. $ git clone https://github.com/ferasboulala/staff-detect.git
  2. $ cd staff-detect && mkdir bin build
  3. $ cd build && cmake ..
  4. $ cd ../bin && ./stav --help

There are many options to the stav binary. Just follow the instructions on the screen. For datascience/machine learning practitioners, it is possible to do batch detection of staves. All staves will be store in a convenient xml format.

Examples

Computer generated score deepscores

Computer generated scores are easy to process. The inputed image is scanned for staff lines. The program detects that staff lines are straight.

Handwritten score muscima

The input image is scanned for staff lines. The program detects that there is a curvature to the staff lines. It estimates the staff line model, corrects the curvature and removes staffs.

About

Staff detection for computer generated and handwritten scores

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published