A Survey on Digital Watermarking Technologies

During the past decade, with the development of information digitalization and internet, digital media increasingly predominate over traditional analog media. However, as one of the concomitant side-effects, it is also becoming easier for some individual or group to copy and transmit digital products without the permission of the owner. The digital watermark is then introduced to solve this problem. Covering many subjects such as signal processing, communication theory and Encryption, the research in digital watermark is to provide copyright protection to digital products, and to prevent and track illegal copying and transmission of them. Watermarking is embedding information, which is able to show the ownership or track copyright intrusion, into the digital image, video or audio. Its purpose determines that the watermark should be indivisible and robust to common processing and attack. Currently the digital watermarking technologies can be divided into two categories by the embedding position——spatial domain and transform domain watermark. Spatial domain techniques developed earlier and is easier to implement, but is limited in robustness, while transform domain techniques, which embed watermark in the host’s transform domain, is more sophisticated and robust. With the development of digital watermarking, spatial techniques, due to their weakness in robustness, are generally abandoned, and frequency algorithm based on DCT or DWT becomes the research focus. Another tendency in watermarking is blind extraction, which means the host is not need when extracting the watermark; otherwise it is hard to avoid the multiple claims of ownerships.