Research On Anti-Collusion Digital Fingerprinting Technology

Digital fingerprinting is an effective means for copyright protecting and digital rights management. The publishers will mark each copy of the digital objects with a unique codeword and distribute them to different users, while the illegal users would combine several differently-fingerprinted copies of the same content to attenuate or even remove the fingerprint in order to get unauthorized copy. At this moment, the publishers will be able to report any illegal copying of digital products in terms of the fingerprints marked in the copies, and simultaneously the illegal users will be deterred from releasing an unauthorized copy. This paper focus on the issue of digital fingerprinting, the main contents are organized as follows:1. Firstly, the paper provides an overview of the related knowledge about the digital fingerprint, including the background, basic idea, applications etc. The principle, classifications and performance measures are presented in detail. Then the research and typical algorithms at home and abroad are summarized. Based on that, we have analyzed the digital fingerprint system model and its requirements as a means of copyright protecting.2. Secondly, the characteristics of four typical fingerprint coding schemes are studied as well as their advantages and disadvantages, and then a brief discussion is made on their anti-collusion characters. The research shows that traditional security code has a good anti-collusion capability, but the length of the code is too long to fulfill the need of reality. Furthermore, colluders can easily compare the differences of the copies and find fingerprinting code. As a result, the fingerprinting code is manipulated and the colluders achieve the piracy purpose. If the fingerprinting code words are from continual range, the shortcoming can be resolved. However, the fingerprinting detection is inefficient in the continual fingerprinting schedule.3. A new kind of fingerprint scheme is proposed, which is based on the image perceptive technology and human visual model. Different copies can be abtained by warping the fingerprinted images randomly with different keys, but their perceptual quality is still in good state. On the contrary, the illegal users will get a colluded copy which has a serious decline in quality; as a result, the collusion action is prevented. The experiment shows that the algorithm has an effective anti-collusion quality on a small number of users' piracy.4. Meanwhile, based on the knowledge of concatenated code and Reed-Solomon code, a hybrid encoding fingerprint scheme is designed, which combines the advantages of orthogonal fingerprint and coding fingerprint. The fingerprinting sequence is constructed by concatenating the orthogonal sequence corresponding to the symbols in the IPP codes. The experiment shows that this fingerprint algorithm has a good performance of short codes and high anti-collusion character.