Research On Anti-collusion Embedding And Extracting In Digital Fingerprinting System

While the network and digitalization has brought lots of convenience to the application and transmission of digital resource, it has also brought many negative effects, such as unauthorized copying, information rectification, etc. Though the traditional encrypt method has to some extent increase the difficulty of pirate, when the content is gotten it can not prevent the product from copying and illegally broadcasting. Digital fingerprinting is put forward to resolve the problem of media copyright protection, by fingerprint encoding, it embeds the different codewords to the digital product to make that the copy is unique so that if an illegal copy is found, it's possible to trace down to the dishonest users by extracting the fingerprints from the illegal copy.Chapter 1 mainly introduces the research background, development history and the application in the copyright protection of the digital fingerprint. Chapter 2 introduces the basic concept, model, requirement and some typical fingerprint schemes during the development of the digital fingerprint. In chapter 3, we introduce the anti-collusion fingerprint encoding algorithm, this method combined the orthogonal code with BIBD code together, it not only easy to create, shorten the fingerprint length to some extent, but also can identify all the illegal users effectively.The fingerprinted digital products may be interferences by various kinds of noise when transmitted, which try to delete the fingerprints or disturb the extraction of the fingerprints. In order to improve the transmission reliability, and exactly extract the fingerprints after the interference from the channel. Chapter 4 proposes a new robust anti-collusion fingerprint embedding and extracting method, namely the embedding and extracting method is based on Turbo code and lifting wavelet transform. This algorithm takes advantage of the excellent error correcting characteristic of Turbo code, the fingerprints are coded by Turbo code before embedding and added to the wavelet domain which is suitable for the visual characteristics of human. In the end, the Turbo iterative decoding is adopted to extract the fingerprints. fingerprint encoding, embedding, attacking, extracting and tracing. During the experiment, we mainly simulate the collusion attack and the dual attack including collusion attack and robust attack, in the end, we do some compared experiment and list the results. The results demonstrate that the embedding method has a good performance on collusion attack, and the proposed algorithm we applied in this paper can reduce the BER (bit error rate) of digital products during transmission, strengthen the robustness of the fingerprints, balance the contradiction between the transparency and robustness and improve the overall collusion resistance.