The Research Of Audio Watermarking Technologies Based On Vector Quantization

A lot of multimedia data like audio, video, image and electronic publications spread on the internet, which make people can get all kinds of information through the network. But challenges of information security work come out because of the unauthorized spread, replica, malicious tamper and attack. Digital watermarking technology was thought to be the effective mean for intellectual property protection and digital multimedia anti-counterfeiting and has become a focus of research in the field of information security.Vector quantization technology which is rooted in shannon rate distortion theory is a lossy compression technology and is promotion of the scalar quantification. VQ technique has been widely used because practice has proved that vector coding instead of scalar coding can obtain better performance. As a optimal solution of signal quantification, VQ technique has high compression rate and good quality of signal recovery. Academics have widely develop research based on VQ technique for its advantages on watermark application like large amount of embedded, easy to realize and good robustness. But this research work more based on image carrier, with audio for carrier rarely seen. This paper combines digital audio watermarking technique with vector quantization techniques and audio watermarking algorithm are explored and researched. The main work includes the following aspects:Firstly, an audio watermarking scheme based on VQ codebook pairing is proposed. Codebook is very important in vector quantization process. Since many watermarking algorithms have used relevant codebooks with all carriers as embedding points. This algorithm segments the codebook with the codeword paired off methods. Then making the subcodebooks represent different information and using the corresponding relation of it with watermarks to determine the output vectors for modifying coefficients of middle frequency, achieving the watermarking purpose. Watermark extraction only needs to judge which subcodebook the matching codeword with MDCT middle frequency coefficients belongs to and the represented information can be extracted as the watermark without original audio signal. The algorithm has good robustness.Secondly, combined with zero-watermarking technology, a novel robust audio multi-watermarking algorithm based on index characteristics of the efficient VQ compression technique is proposed. We make use of the relativity of the neighboring indices to calculate the variance and mean, then generate the polarities by considering these properties. The permuted watermarking information loads on the corresponding polarity sequences to generate the secret keys. In this paper, we realize the zero-embedding of multiple watermarks and guarantee the watermarks'invisibility thoroughly. The watermarks can be extracted without the original digital audio signal and watermarking images, only depending on the private keys. Besides, algorithm can embed different copyright information into one audio work synchronously without considering the embedding order and the stable correlation features between adjacent index guarantee robustness.Thirdly, unlike normal vector quantization, index constrained vector quantization can limit the range of searching codewords by select an constrained position in the index in the process of quantization. This paper makes use of this special searching method to propose an audio watermarking scheme. The original audio signal is first segmented into several frames, DCT transform is performed on every frame and the middle-frequency coefficients are extracted to form vectors. The embedding procedure can be described as finding matching codewords to modify the DCT middle-frequency coefficient vectors in accordance with the watermarking bits and the constrained bits. Normal vector quantization is adopted in the extracting procedure. The constrained bits in the same position are extracted as the watermarking information. This scheme is real-time because the watermarks can be embedded in the quantization process. The experimental results show that the watermarking is fragile and can be used for authentication.