Research On Digital Audio Watermarking Algorithm Based On Wavelet Transform

With the development of the computer network and multimedia technology, multimedia data has been one of important sources from which people get information,and has been a significant part of our lives.But data copyright protection is facing flinty challenge because of the quick development of the digital signal processing and the technology of computer network transmission. Therefore,research on how to protect the digital product effectively on the condition of open network is a hot topic.Digital watermarking technology is the most active branch in the field of researching on information hiding technology now,which used in copyright protection has heavy demands of imperceptibility and robustness,and the unique character of hiding information is also one of the demands for watermarking technology.The characteristics of digital audio watermarking and human hearing system were analyzed in this dissertation,and three kinds of digital audio watermarking technology were proposed as follows:A kind of self-adaption digital audio watermarking algorithm based on wavlet transform was improved.The self-adaption digital audio watermarking algorithm which confirms the embedded intensity by using the population mask value of mask model was brought forward.Its characteristic is adjusting the intensity value of the embed watermarking according to the change of mask value of audio signal,and the quality of origin signal and robustness were assured.A kind of wavlet field self-adaption digital audio watermarking algorithm based on average value quantify was proposed.The method of quantify coefficient embedding watermarking information was used in this algorithm,and blind watermarking was implemented,practical value of watermarking project was improved.Because the covariance of population is less than the one of sample,agility of embedding method was enhanced by modifying a series of average value of wavlet coefficient.A algorithm combining wavlet transmission and independent component analysis was proposed.By using the method of maximum peak value of ICA, which is a new and available watermarking method assures the heavy embedded capacity,strong robustness and the heavy attack.Algorithms proposed above can be applied in the field of copyright protection and information hiding and used as a method protecting the property right of audio digital product,and used in hiding communication by using audio as carrier.