Study On Techniques Of Digital Image And Audio Watermarking

Recent years, with the rapid development of computer communication and network technique, the treatment, storage and transmission of multimedia data become more and more convenient, which greatly enriches the lives of people. However, the illegal copy and distribution of multimedia data also become more and more easy, and a large number of authors'and publishers'intellectual property copyrights have suffered from violation, which led to huge damage of their benefits in many applications. There is sufficient evidence to show that various kinds of piracy have become the major obstacle to the healthy development of digital industry. In this situation, digital watermarking used as the countermeasure to resolve the problem of copyright protection and content authentication has attracted a great deal of attention in the field of international information security.Based on the robust and fragile watermarking technique, using image and audio as the host signals, the author has designed some novel watermarking algorithms. The achieved innovative results are listed below.1. The current fragile image watermarking algorithm with discrimination of tampers on watermark and content is analysed, and a conclusion that it is disabled for some tampers is obtained. Based on the characteristics of randomness and extreme sensitive to initial values of chaotic map, a novel fragile image watermarking algorithm capable of tamper detection, tamper location, tamper discrimination and tamper restoration is proposed, which has good invisibility.2. In order to enhance the invisibility and robustness of SVD(singular value decomposition)-based robust image watermarking algorithm, two rules are proposed. The first rule is that only the coefficients in the first column of U component and V component may be modified for embedding watermark bits. The second rule is that when using one of the two components to embed watermark bits, the other component may be adopted for compensating the visible distortion. Theoretical analysis and experimental results have confirmed the correctness and effectiveness of the proposed rules.3. Aiming at the security hole of using binary image as fragile watermark, based on SVD, the fragile watermark is generated from audio itself and a novel fragile audio watermarking algorithm is proposed, which is not only inaudible and sensitive against MP3 compression, noise adding, lowpass filtering, etc, but also fragile to malicious modification. Moreover, it can detect and localize the tampered regions effectively.4. The current public transform domain robust audio watermarking algorithms are vulnerable to malicious attacks. Hence, based on the uncertain angle of discrete fractional transform, a class of robust audio watermarking algorithms using discrete fractional transforms, including DFRFT(discrete fractional fourier transform), DFRCT(discrete fractional cosine transform) and DFRST(discrete fractional sine transform), are proposed. The security of the proposed algorithms is enhanced by using chaotic sequence to select the specific angle of discrete fractional transform. Experimental results show that the watermarks are imperceptible and robust against various common audio signal processing attacks, such as MP3 compression, re-sampling, etc, moreover, the false alarm and false rejection probability of the proposed algorithms are all low.5. The stability against playback speed modification of absolute average value of DFT(discrete fourier transform) coefficients of audio signal is confirmed by theoretical analysis and experiments. Using DFT domain relation among three neighboring audio segments, a novel robust audio watermarking algorithm is presented. Simulation results show the presented algorithm has acceptable imperceptibility and robustness against common audio signal processing attacks, including MP3 compression, re-quantization, re-sampling, etc. Furthermore, the ability of resisting playback speed modification and amplitude variation is also satisfactory.6. The stability against playback speed modification of ZCR(zero-crossing rate) of audio signal is confirmed by theoretical analysis and experiments. On the basis of this time domain statistical feature, a novel robust audio watermarking algorithm is proposed. Experimental results demonstrate that the proposed algorithm is inaudible and robust against various common audio signal processing attacks, including MP3 compression, re-quantization, re-sampling, etc. Moreover, it has superior robustness against playback speed modification.