Capacity improvement of a blind symbolic music score watermarking technique

Current music score watermarking techniques either lack in robustness or in capacity: While embedding data into staff lines guarantees a certain payload, modulation of the staff lines' thickness can be attacked using image processing techniques. Changing properties of musical symbols is more robust. But, the number of musical symbols limits the payload. Therefore, we propose the following method: The staff lines' distances are used as a carrier signal for embedding information. Changes are applied by using an image warping technique to reduce artifacts. Image warping techniques can also be applied locally to music symbols to change their properties, e.g. their horizontal distance or their width. Thus, quality improves. Furthermore, musical symbols doesn't have to be recognized. It is sufficient to cluster the symbols according to their properties and to treat symbols of the same cluster equally. The described method guarantees a minimum payload and increased robust. The application of image warping techniques reduces visible distortion. The results are visual appealing music scores. Capacity improvement can be achieved by applying the proposed methods to music symbols.