Content aware music score pre-processing

In most OMR algorithms, the pre-processing stage consists on a binarization operation followed by the estimation of the staffline thickness and the vertical line distance within the same staff. Subsequent operation can use these values as references. This work presents an improvement on previous conventional estimates for the two reference lengths and consequently a novel binarization method based in the content knowledge of the image. An objective evaluation over a large dataset of handwritten music scores is conducted based in the performance of the subsequent steps in an OMR system.