Classifying algorithm for watermarks based on mathematical morphology

Due to existing digital watermarking scheme barely researching on classification of extracted watermark, we proposed a novel classifying algorithm for watermarks based on mathematical morphology. Applying mathematical morphology operation to a binary image watermark which composed of a certain proportion of black pixels and white pixels, at the same time the different ID corresponding to the different watermarks were generated and composed the database K. Applied the mathematical morphology filter to the watermark which was repaired by the histogram normalization processing utilizing the known parameter and hole filling processing worked on the extracted watermark to get the ID' corresponding to extracted watermark. Retrieval the ID' in database K, if there was not an ID that was the same to ID', the watermark does not be protected, and if there is an ID that is the same to ID', search the watermark that corresponding to the with the ID' to achieve the purpose of distinguishing the ownership. The simulation results show that application of mathematical morphology method to detect the watermark information classification was portable for other watermarking algorithms, which provided a new idea of the wide application of digital watermarking.