Research On Digital Audio Watermarking Algorithm

The rapid development of computer network technology enables digital products, such as digital images, digital music and video appear and be widely applied. At the same time, the powerful Internet breed rampant piracy of digital products. The copying, tampering problem represented by digital music is getting worse, so copyright protection of digital products needs to be solved. After years of development, digital watermarking technology is considered to be an effective method to protect the copyright of digital products.In this paper, on the basis of a comprehensive analysis of audio watermarking algorithm, a dual function of watermark scheme in hybrid domain is proposed, which can realize the digital audio copyright protection and content authentication. The algorithm is mainly using the ideas of a variety of signals transformed and combined to realize two functions of digital watermarking, copyright protection and content authentication. Firstly, the original digital audio is decomposed into many segments according to the size of the watermark. Secondly, the DWT were performed on each audio segments. Finally, robust watermark is embedded in the low-frequency component, then the high frequency component are decomposed on QR decomposition at the same level and the fragile watermark is embedded in the first row of QR decomposition. The algorithm uses watermark Arnold scrambling and Logistic chaotic scrambling technology which double encrypt original watermark in order to improve the privacy and security. Blind detection was achieved for both robust and fragile watermarksFor deep exploration of digital audio watermarking technology, this paper combined the low-pass characteristic of the neighborhood average sequence and the energy-gathered feature of the discrete cosine transform, then proposes a robust audio watermarking algorithm based on neighborhood averaging method and DCT. This work proposed a method of embedding synchronization mark into audio signals in the cross of two neighborhood averaging sequences and embedding watermark into the maximum coefficient in DCT domain of the sequences. Embed in this way, not only make the search of synchronous code more efficiently, but also improve the accuracy of the extracted watermark information. Experimental results show that this algorithm can not only deal with digital audio watermarking noise, low-pass filtering, rate-change sampling and other attacks, but also has a good robustness.