The Digital Fingerprinting System Based On LDPC Code

With the expansion of networks all over the world, we see more and more information distribution. And digital media have conquered our live during the latest few years. On one hand there are great advantages in the use of digital imagery, video or sound, especially with the rapid growth of the Internet; on the other hand digital copies are exact and easily made by illegally copying and distributing, which might lead to financial loss or legal problems of the producers. How to protect the original authors' intellectual property rights and integrity of digital products is a realistic problem urgently needing to be solved. Digital fingerprinting has been suggested in the literature as the most effective means for copyright protection and authentication.Digital fingerprinting is the technology to prove of ownership on copyrighted material. Embedding of unique customer identification as a watermark into data is called digital fingerprinting to identify illegal copies of documents and detect originators of illegally made copies. In chapter 1, the background and the state of information hiding techniques are introduced. In chapter2 and chapter3, we discuss the significance, development and theory of fingerprinting technology in a whole. In chapter 4, attackers can compare several fingerprinted copies to find and destroy the embedding identification string by the altering the data in those places where a difference was detected. So we present a new fingerprint anti-collusion codes which combine BIBD code and orthogonal modulation.In the chapter5, based on the analysis of the equival channel in then digital fingerprinting system, a fingerprint embedding algorithm for digital works on the basis of LDPC code is proposed in this paper. The algorithm embeds and retrieves the watermark signature into and out of the digital works by employing the LDPC code technology and the human visual characteristics. The watermark is first LDPC coded and embeds into the host signal, then retrieves from the watermarked works using LDPC iterative methods. In the end, we simlutate thefingerprint system based on the LDPC code as well on the Visual C++6.0 platform. The experiment results demonstrate that the proposed algorithm has greatly decreased the error code rate of the retrieved watermark and improved the attack of defending.